You are on page 1of 15

Research Article

Lexical Acquisition and Phonological Development


in Minimally Verbal Children With Autism
Spectrum Disorders
Maysoon F. Billera and Kayleigh A. Yeagera
a
Department of Communication Sciences and Disorders, University of Central Arkansas, Conway

ARTICLE INFO ABSTRACT


Article History: Purpose: This study examines two components of lexical acquisition and pho-
Received December 9, 2021 nological development that occur during the first 50-word stage of language
Revision received March 1, 2022 development in neurotypical (NT) children. One component is how children learn
Accepted May 10, 2022 words based on their existing speech sound inventories (i.e., in-phonology and
out-of-phonology word learning). The other component is the relationship
Editor-in-Chief: Holly L. Storkel between the children’s number of words and the number of phonemes in their
Editor: Toby Macrae speech sound inventories. The goal of this study is to determine if the same
two components occur in children with autism spectrum disorders (ASDs) who
https://doi.org/10.1044/2022_LSHSS-21-00184 are older than their NT peers but are in the same stage of lexical development.
Method: This study involved 20 minimally verbal children with ASDs, ages 28–
72 months, who produced five to 50 spoken words. The children’s spoken
words were obtained from the MacArthur–Bates Communicative Development
Inventories. The speech sound inventories were obtained from the utterances
produced during assessment/play sessions with the children. The children’s
spoken words from the Communicative Development Inventories (CDI) were cat-
egorized as either in-phonology or out-of-phonology based on whether the
words began with a phoneme in the child’s existing speech sound inventory.
Additionally, the children’s number of spoken words on the CDI was compared
to the number of phonemes in their speech sound inventories.
Results: The children in this study produced in-phonology words more often
than out-of-phonology words (z = −3.922; p < .001). Moreover, there was a mod-
erate positive correlation between the children’s number of spoken words and
the number of phonemes in their speech sound inventories (r = .534, p = .019).
Conclusions: The relationship between lexical acquisition and phonological
development appears to exist for the children in this study, who are in the first
50-word stage of development. Clinical implications for increasing the expres-
sive language of children with ASDs were discussed.

Children with autism spectrum disorders (ASDs) areas of semantics, phonology, syntax, and morphology.
vary greatly in their language abilities (Tager-Flusberg They have adequate verbal communication; however, they
et al., 2005). Along the continuum of spoken language, still demonstrate deficits in pragmatic skills (Rapin &
there are children with ASDs who are fully verbal, mini- Dunn, 1997). Minimally verbal children use some words,
mally verbal, and nonverbal. Children with autism who are but “significantly fewer than expected levels relative to
fully verbal have mastered various language skills in the age” (Koegel et al., 2020, p. 2968). Koegel et al. viewed
“significantly fewer” to mean an expressive spoken vocabu-
lary below the 10th percentile for the child’s age. In this
Correspondence to Maysoon F. Biller: mbiller@uca.edu. Disclosure:
study, that would equate to fewer than 50 words for a child
The authors have declared that no competing financial or nonfinancial who is 30 months of age or older (Fenson et al., 2007;
interests existed at the time of publication. Koegel et al., 2020). Approximately 25%–30% of children

1074 Language, Speech, and Hearing Services in Schools • Vol. 53 • 1074–1087 • October 2022 • Copyright © 2022 American Speech-Language-Hearing Association
with ASDs fail to develop spoken language beyond a mini- The Early 8 consonants are /m, n, j, b, w, d, p, and h/,
mal level by the age of 5 years (Kasari et al., 2014). Nonver- which emerge between the ages of 12–36 months and are
bal children over the age of 18 months do not demonstrate consistently produced by around 36 months. The Middle
any consistent spoken words (intelligible or word approxi- 8 consonants are /t, ŋ, k, g, f, v, tʃ, and dʒ/, which emerge
mations) across all contexts and all speakers (Koegel et al., between the ages of 36–78 months and are consistently
2020). They often communicate using unconventional and produced around 66 months. Finally, the Late 8 conso-
conventional gestures. nants are /ʃ, ʒ, l, r, s, z, θ, ð, and clusters/, which emerge
As fully verbal children with autism have less con- between the ages of 60–90 months and are consistently
cerns with phonology, semantics, syntax, and morphology produced around 90 months.
and nonverbal children do not demonstrate enough speech
sound production or recognizable utterances to support Phonological Development
spoken language, this article primarily addresses the lan-
guage development of minimally verbal children with The role that phonological development plays in the
autism. At some point in their language development, acquisition of a child’s lexicon has been well documented
many minimally verbal children with ASDs pass through (Ingram, 1989; Schwartz & Leonard, 1982; Sosa & Stoel-
the single word or first 50-word stage of language devel- Gammon, 2006; Stoel-Gammon, 2011). The phonological
opment. The research in neurotypical (NT) language system appears to have a greater impact on lexical acquisi-
development highlights the interdependence between lexi- tion than lexical factors have on phonological development
cal acquisition and phonological development and their (Stoel-Gammon, 2011). First, children more easily acquire
direct influence on production variability within a child’s new words that begin with phonemes (e.g., vowels or con-
spoken vocabulary (Ingram, 1989; Leonard et al., 1982; sonants) that are in their speech sound inventory (Brady
Stoel-Gammon, 2011). et al., 2015; Leonard et al., 1982; MacRoy-Higgins et al.,
2013; Schwartz & Leonard, 1982; Stoel-Gammon, 2011).
Some researchers focused solely on consonants (Leonard
Lexical Acquisition and Phonological et al., 1982; Schwartz & Leonard, 1982), whereas other
Development in NT Children researchers considered vowels and consonants to be a part
of a child’s speech sound inventory (Brady et al., 2015;
Lexical Acquisition MacRoy-Higgins et al., 2013; Stoel-Gammon, 2011). Sec-
ond, lexical acquisition and phonological development are
During the first 50-word stage of development, com- proportional and bidirectional. Children with larger lexi-
prehension of words develops earlier and more rapidly cons have a greater need for more developed phonological
than production, with comprehension being attained skills. Conversely, children with smaller lexicons have less
5 months prior to production in classic studies of vocabu- developed phonological abilities (Stoel-Gammon 2011).
lary development (Bates et al., 1988; Benedict, 1979).
Children begin saying their first words at approximately Lexical Acquisition and Phonological
12 months of age, and from that point on, their vocabu- Development
lary begins to expand rapidly. Children have well over
200 spoken words by the age of 2 years and 1,000 words A pivotal study was performed on the influence of
by age 3 years (Nicolosi et al., 2006). Nevertheless, there phonological selection and avoidance during word learn-
is variability in children’s word learning based on cross- ing (Schwartz & Leonard, 1982). When children learn new
cultural and socioeconomic factors (Fenson et al., 1994; words, they tend to imitate and spontaneously produce
Hart & Risley, 1975; Hoff, 2003; Rowe, 2012). Further- words with certain phonological characteristics that they
more, variation in an individual child’s development could can say (i.e., phonological selection). Conversely, children
reflect differences in the age the first word was spoken or avoid words with sounds that they find difficult to say (i.e.,
differences in the rate of vocabulary growth (Fenson phonological avoidance). Based on this concept, Schwartz
et al., 1994). During phonological development, children and Leonard (1982) introduced the notion of in-phonology
by 24 months of age tend to attain more consonants in and out-of-phonology word learning. In-phonology words
the initial position than the final position of words (Bleile, contained consonants that had been present in the child’s
2004; Dyson, 1988; Stoel-Gammon, 1985). Stoel-Gammon spontaneous productions. Conversely, out-of-phonology
(1985) reports 11 word-initial consonants present in chil- words contained consonants that had not been present in
dren’s speech sound repertoires by the age of 24 months: / the child’s spontaneous productions. In their study, 12 tod-
b, t, d, k, g, m, n, f, s, h, and w/. Another model of dlers were presented with 16 nonsense lexical concepts,
speech sound acquisition divides the 24 English conso- labeled eight nonsense words that were in-phonology and
nants into the Early, Middle, and Late 8 (Shriberg, 1993). eight nonsense words that were out-of-phonology for the

Biller & Yeager: Lexical and Phonological Development in ASDs 1075


child. The researchers found that the children produced in- Lord & Paul, 1997). Barring any other deficits that may
phonology nonsense words more often and earlier than affect speech production, such as abnormalities of the oral
out-of-phonology nonsense words. They concluded that structures or childhood apraxia of speech, children with
children’s developing vocabularies are based on their speech autism follow a typical sequence in phonological develop-
production abilities and preferences. These preferences can ment (Rapin et al., 2009; Tager-Flusberg et al., 1997).
be observed in the phonemes that constitute the words that For the subset of children with ASDs who experience
children produce (Schwartz & Leonard, 1982). Recognizing phonological deficits, their differences are noticeable, start-
the relationship between lexical acquisition and phonolo- ing in infancy. Infants who were later diagnosed with ASDs
gical development in NT children will guide our research in produced fewer speech-related vocalizations (Warlaumont
the same areas of development for children with ASDs. et al., 2014) and demonstrated delayed onset of canonical
babbling, as well as lower rates of canonical babbling com-
pared to their NT peers (Patten et al., 2014). Additionally,
Lexical and Phonological Development in they produced a significantly higher percentage of atypical
Children With ASDs vocalizations such as squeals, growls, raspberries, car
sounds, and yelling than children with NT language devel-
Lexical Development opment (Plumb & Wetherby, 2013).
Older children with ASDs who are nonverbal or
Previous studies have examined lexical acquisition in minimally verbal may struggle with motor movement and
children with ASDs who range from being nonverbal to motor planning for speech production (Adams, 1998; Bhat
fully verbal (Charman et al., 2003; Jimenez et al., 2021; et al., 2011; De Giacomo et al., 2009; Gernsbacher et al.,
Luyster et al., 2007; Rescorla & Safyer, 2013; Tek et al., 2008; Page & Boucher, 1998; Plumb & Wetherby, 2013;
2008). For those children with ASDs who eventually Rogers et al., 2006; Stone & Yoder, 2001). Children with
develop spoken language, the onset is often delayed: They ASDs were found to have poor oral motor and manual
produce their first words at approximately 38 months of motor abilities compared to NT children (Adams, 1998;
age, compared to 8–14 months of age in NT children Gernsbacher et al., 2008) and poor production of complex
(Howlin, 2003). Once children with ASDs expand their phonemes/syllables, indicative of severe motor planning
spoken language, their lexical development has similarities problems (Adams, 1998). Furthermore, there are researchers
and differences to that of NT children. Children with ASDs who believe that there is a subgroup of children with autism
demonstrate a noun bias, learn words from various seman- who has suspected childhood apraxia of speech (Chenausky
tic categories, and acquire verb meaning using syntactic et al., 2019; Page & Boucher, 1998; Shriberg et al., 2011).
bootstrapping and sentence structure comparable to their
NT peers (Charman et al., 2003; Horvath et al., 2018; Lexical Acquisition and Phonological
Luyster et al., 2007; Rescorla & Safyer, 2013; Swensen Development
et al., 2007). However, there are differences in lexical devel-
opment between children with ASDs and their NT peers There is only one study to date that has examined
such as fewer words in their lexicon and fewer social verbs the relationship between lexical acquisition and phonolo-
such as smile and share (Charman et al., 2003; Jimenez gical development in a clinical population. Even though
et al., 2021; Tager-Flusberg et al., 2005). Another difference the study focused on word learning in children with spe-
between children with ASDs and NT children is their cific language impairment (Leonard et al., 1982) and not
understanding of the meaning of verbs that express some- on children with ASDs, it was important to review as
one’s mental state (i.e., theory of mind), such as remember, another example of in-phonology and out-of-phonology
think, and know (Kazak et al., 1997; Kelley et al., 2006). word learning. The methodology in this study was similar
to the one used in the study of Schwartz and Leonard
Phonological Development (1982); however, the participants in the study of Leonard
et al. (1982) included children with specific language
Phonological development in children with ASDs impairment, as well as NT peers. Another prominent dif-
appears to be one of the least affected domains in children ference between the two studies was the position of the
with ASDs in that these children usually follow a typical consonant in the target words. Whereas Leonard et al.
developmental sequence in their speech sound acquisition (1982) used an existing consonant in the word-initial posi-
(cf. Eigsti et al., 2011). Whereas most individuals with ASDs tion of their in-phonology target words, Schwartz and
do not have specific impairments in articulation and phonol- Leonard used existing consonants in all word positions of
ogy, problems in these areas can be found in low- their in-phonology target words. Leonard et al. discovered
functioning individuals with ASDs, and when children with that NT children and children with SLI were more likely
ASDs are in their early childhood years (Eigsti et al., 2011; to learn the novel words containing initial consonants in

1076 Language, Speech, and Hearing Services in Schools • Vol. 53 • 1074–1087 • October 2022
the children’s existing inventory than words containing ini- recruitment and participant consent were approved by the
tial consonants outside of their inventories. institutional review board at the University of Illinois at
Urbana–Champaign and the University of Central Arkansas.
Potential participants were excluded from the study if
This Study they were reported by their parent/caregiver to have any
other any sensory impairments (e.g., hearing loss or
The notion that children acquire vocabulary based visual impairment) or if the children had too many or
on their speech sound inventories (i.e., sounds they are too few spoken words, according to the inclusionary cri-
currently producing) has been studied in NT children teria. The participants in the study included 20 children
(Schwartz & Leonard, 1982) and children with specific with ASDs, consisting of 16 boys and four girls. They
language impairments (Leonard et al., 1982). There is a ranged in age from 28 to 72 months, with a mean age
paucity of studies that have applied this line of research to of 55 months. Regarding race, 18 children were reported to
children with ASDs. The lexical–phonological association be White, one child to be Black, and one child to be bira-
during the first 50-word stage of development, which cial (White and Black). The children met the following
occurs between an NT child’s first and second birthdays, inclusionary criteria: (a) adequate hearing acuity or detec-
has not been proven to exist in children with ASDs who tion of sound; (b) a diagnosis of ASD by medical or educa-
are older yet are still in the same developmental stage. tional personnel using a standardized measure for ASD
The aim of this study is to examine two components of the and meeting the criteria as determined by the Diagnostic
lexical acquisition and phonological development in chil- and Statistical Manual of Mental Disorders: Fifth Edition
dren with ASDs to determine if they follow the same pat- (DSM-V; American Psychiatric Association, 2013); and (c)
tern of development as their NT peers. First, we want to a spontaneous spoken vocabulary of five to 50 words, as
determine if children with ASDs use more in-phonology measured by the MacArthur–Bates Communicative Devel-
than out-of-phonology words. Second, we want to deter- opment Inventories (Fenson et al., 2007). The children in
mine if the children’s spoken lexicon and speech sound this study would be considered minimally verbal as defined
inventories are commensurate. Determining the relationship by Koegel et al. (2020). Additionally, the Mullen Scales of
between these two areas of development could guide treat- Early Learning (MSEL; Mullen, 1995) was given as a
ment goals for expanding spoken language in children with descriptive measure of each child’s visual reception, fine
ASDs. The research questions are as follows. motor, receptive language, and expressive language skills.
The standardized tests for diagnosing the children with
1. Do minimally verbal children with ASDs produce sig- ASDs were performed by medical professional or school
nificantly more in-phonology than out-of-phonology psychologists who were familiar with administering the
words? assessments. The parents completed the Communicative
2. Are lexicon and phonological development commen- Development Inventories (CDI) and the first author,
surate in minimally verbal children with ASDs? known as the examiner, administered the MSEL. (See
Table 1 for the children’s demographic information and
Our hypothesis for the first question is that mini- inclusionary data.)
mally verbal children with ASDs will produce more in- All of the children passed either a formal audiologi-
phonology than out-of-phonology words, thus demonstrat- cal pure-tone screening test or an informal sound detec-
ing phonological selection and avoidance. Our hypothesis tion task. The task consisted of three different noise-
for the second question is that minimally verbal children makers (e.g., a piece of cellophane, a rattle, and a buzzer)
with ASDs will have a spoken vocabulary proportional to being placed out the child’s field of vision at a distance of
their consonant inventory. 2 ft behind the right and the left ear. If the child turned
toward the sound, it confirmed his or her ability to
respond to acoustic stimuli at minimal to moderate levels.
Method All of the children met the DSM-V criteria for ASD:
Eleven were assessed using the Autism Diagnostic Obser-
Participants vation Schedule–Second Edition (Lord et al., 2012), and
nine were assessed using the Childhood Autism Rating
The children were recruited from preschools and Scale–Second Edition (Schopler et al., 2010). On the CDI,
schools in two regions of the Midwest. A flyer about the the children produced five to 50 spoken words (M =
study were sent to parents or caregivers via e-mail. If a 27.55, SD = 13.17). On the MSEL (Mullen, 1995), the
parent/caregiver indicated an interest in the study, an inter- children’s scale scores were reported as development quo-
view was scheduled, and written consent to participate in tients (DQs) instead of standard scores (Psychology
the study was obtained. The procedures for participant Research and Reference, 2022). The only exception was

Biller & Yeager: Lexical and Phonological Development in ASDs 1077


Table 1. Demographic information and descriptive measures for the participants.

Mullen Scales of Early Learning


Test for Vis. Rec. Composite
Child Age (Mo.) Gender Race autism DQa FM DQ RL DQ EL DQ score

1 28 M White ADOS-2 75 79 36 36 59
2 45 M White CARS2 38 24 29 27 49
3 70 M White ADOS-2 66 60 24 21 49
4 48 M White ADOS-2 54 56 46 33 49
5 53 M White CARS2 51 45 32 34 49
6 71 M White CARS2 35 28 21 28 49
7 45 F White CARS2 64 53 60 44 50
8 67 M White CARS2 31 31 24 22 49
9 52 M White CARS2 40 38 29 29 49
10 68 F White CARS2 37 38 20 22 49
11 59 M White ADOS-2 41 46 39 44 49
12 43 M White ADOS-2 56 79 35 30 52
13 46 M White ADOS-2 59 57 37 39 49
14 40 M White ADOS-2 60 55 63 53 49
15 71 F Black CARS2 38 38 42 41 49
16 42 F White ADOS-2 64 55 38 43 50
17 69 M Mixed ADOS-2 65 83 39 38 59
18 72 M White CARS2 72 71 74 39 80
19 53 M White ADOS-2 49 42 45 38 49
20 62 M White ADOS-2 58 60 45 44 49

Note. The children are listed in order from top to bottom by increasing the number of spoken words. Mo. = month; Vis. Rec. = Visual
Reception; FM = Fine Motor; RL = Receptive Language; EL = Expressive Language; DQ = developmental quotient; M = male; ADOS-2 =
Autism Diagnostic Observation Scale–Second Edition; CARS2 = Childhood Autism Rating Scale–Second Edition; F = female.
a
DQ was obtained for each scale score by dividing the child’s developmental age by his or her chronological age.

the composite score, which is only reported in standard the session. During the children’s assessment/play sessions,
scores. A DQ is obtained by dividing the child’s develop- the examiner administered the MSEL. The eight communi-
mental age by his or her chronological age. Since most of cation temptation activities from the Communication and
the children obtained the lowest possible T score of 49 on Symbolic Behavior Scales (Wetherby et al., 2003) were per-
all of the scales, developmental quotients allowed for vari- formed in an attempt to elicit spontaneous speech from the
ability in the children’s abilities to be revealed. On the children. These activities consisted of blowing bubbles,
Visual Reception scale, the children’s quotients ranged blowing up a balloon and letting it go, using a wind-up
from 31 to 75 (M = 52.65, SD = 12.97). On the Fine toy, reading a book, opening a small jar with food inside
Motor scale, their quotients ranged from 24 to 83 (M = it, playing with blocks, playing with a bag of toys that was
51.90, SD = 16.64). On the Receptive Language scale, the difficult to open, and playing a finger game. The same toys
children’s quotients ranged from 21 to 74 (M = 38.90, and adult interaction style were used each session in an
SD = 13.76). On the Expressive Language scale, their attempt to make the content of the assessment/play sessions
quotients ranged from 21 to 53 (M = 35.25, SD = 8.46). uniform across participants.
Overall, the children’s Early Learning composite standard
scores ranged from 49 to 80 (M = 51.80, SD = 7.31). Measures

Procedure Speech Sound Inventories


The children’s speech sound inventories (e.g., all
The data for the study were collected during the vowels and consonants) were derived from the utterances
parent/caregiver session and the two, 30-min assessment/ spoken by the children during the two 30-min assessment/
play sessions with the child. The children and their parents play sessions, which were video- and audio-recorded. Cap-
were seen either at home or at the child’s preschool. The tured on the videos were the administration of the MSEL,
parent/caregiver completed a social history and the CDI. the communication temptation activities, all interactions
During the parent/caregiver session, the examiner was given between the examiner and the child, and any other utter-
a copy of the medical report containing the child’s diagno- ances produced by the child. Every utterance produced
sis of ASD. The assessment/play sessions were video- throughout the session was transcribed phonetically, which
recorded with an Apple iPad Pro and audio-recorded with included complete words, word approximations, syllables,
a Sony digital voice recorder, which ran the entire length of and phonemes (e.g., vowels and consonants). Additionally,

1078 Language, Speech, and Hearing Services in Schools • Vol. 53 • 1074–1087 • October 2022
utterances that were imitated, elicited, spontaneous, or phonology words. The first group contained the number
produced during vocal play were included in the transcrip- of words from the child’s CDI that began with word-
tion in an effort to obtain a representative and thorough initial phonemes from the child’s existing consonant reper-
sample since it was difficult to elicit language from chil- toire known as in-phonology words. The second group
dren with such limited verbal output (Broome et al., 2017; contained the number of words that did not begin with
Saul & Norbury, 2020). The examiner transcribed 15 out word-initial phonemes from the child’s existing phoneme
of the 20 children’s assessment/play sessions and two grad- repertoire or were misarticulations of a phoneme and were
uate students in communication sciences and disorders known as out-of-phonology words.
transcribed the other five children’s sessions. The number
of utterances among the children varied greatly depending Reliability
upon their level of spoken language.
The examiner scored all the children’s data for in-
Lexicon phonology and out-of-phonology words. An undergradu-
The children’s spoken vocabulary was derived from ate student in communication sciences and disorders was
the words reported by the parents on CDI that the chil- trained to code in-phonology and out-of-phonology words
dren “comprehended and said.” Upon completion of the based on the word-initial phonemes of the spoken words
CDI, the examiner analyzed the results with three consid- on the CDI and the children’s speech sound inventories.
erations in mind. First, the child’s spontaneous production Prior to coding the data, the examiner and undergraduate
of the word on the CDI had to meet a certain criterion to student agreed on the coding rules for the three consider-
be counted. An utterance was considered to be a word if ations listed in the Lexicon section. Training of the under-
it contained at least 50% of the consonants from the target graduate student continued until the student reached at
word with at least one consonant and one vowel from the least 70% agreement with the examiner over three consec-
original word and the consonants produced in the correct utive sessions. For computing reliability, the student’s
order (Leonard et al., 1982). For example, the ba was scoring decisions were compared to those of the examiner.
counted as an acceptable production of ball, but u was Reliability was calculated by dividing the total number of
not accounted as an acceptable production of you. Sec- agreements by the total number of agreements and dis-
ond, misarticulation of the initial phoneme was consid- agreements, and multiplying by 100, to arrive at the per-
ered. If a child misarticulated a phoneme during the centage of agreement. An acceptable level of agreement
assessment/play session (e.g., t/k), it was noted on his or was 80%. Reliability on agreement of in-phonology words
her speech sound inventory form. Furthermore, the child was 85% and that of out-of-phonology words was 100%.
was not given credit for having that phoneme in his or her
speech sound inventory. When transcribing the session, if Data Analysis
the word-initial sound was misarticulated, the word was
included in the list of spoken vocabulary on the CDI. Due to the small sample size in this study (N = 20), a
However, the production was categorized as being out-of- normal distribution of scores could not be assumed. In this
phonology because the child misarticulated the word-initial case, nonparametric statistics were used because there is
phoneme. For instance, a child consistently substituted not the assumption that data are drawn from a normal dis-
t/k. If the child said tomb for comb, then the word comb tribution. The nonparametric Wilcoxon signed-ranks test
would appear on his total list of words, but it would count was chosen to determine if there was a significant difference
as being out-of-phonology. Third, dialectical variation of between in-phonology and out-of-phonology words. The
the initial phoneme was examined. Two of the children in Wilcoxon is more sensitive to a smaller sample size, and by
the study used an acceptable African American English ranking the pairs of scores, it eliminates the effects of out-
dialectal variation of d/θ and d/ð. Consequently, their liers. A Spearman rank-order correlation was performed to
spoken words on the CDI using those variations were determine the relationship between the total number of spo-
counted as being in-phonology. For example, the chil- ken words and the number of phonemes in the children’s
dren said dat for that. They were given credit for having speech sound inventory.
that phoneme in their speech sound inventory and for it
being an in-phonology word.
Results
In-Phonology and Out-of-Phonology Words
The examiner compared the words produced on In-Phonology and Out-of-Phonology Words
each child’s CDI to the list of phonemes that comprised
each child’s speech sound inventory, thereby creating two Statistical analysis revealed that the minimally ver-
lists of words that represented in-phonology and out-of- bal children with ASDs in this study produced

Biller & Yeager: Lexical and Phonological Development in ASDs 1079


significantly more in-phonology than out-of-phonology Child 19 had 42 in-phonology words and six out-of-
words (see Figure 1). The Wilcoxon signed-ranks test was phonology words for a total of 48 words. Child 20 had 42
performed using IBM SPSS Statistics for Windows, Version in-phonology words and eight out-of-phonology words for
27, and revealed that the children produced significantly a total of 50 spoken words.
more in-phonology words than out-of-phonology words
(z = −3.922, p = < .001) with a large effect size (r = −.62). Lexical Size and Phonological Development
The children produced from four to 42 in-phonology words
(M = 24.30, SD = 12.14) and from zero to eight out-of- A Spearman’s rank-order correlation was performed
phonology words (M = 3.20, SD = 2.57). Given the fact on the relationship between the total number of phonemes
that Leonard et al. (1982) used word-initial consonants (i.e., vowels and consonants) and the total number of
from the children’s existing consonant repertoires for their words produced by children in this study. As shown in
target words, a Wilcoxon signed-ranks test was performed Figure 2, there was a moderate positive correlation
on in-phonology words and out-of-phonology words con- between phonemes and words (r = .534, p = .019), which
taining consonants only. The consonant-only analysis was significant at the .05 level. Of the 44 English pho-
revealed very similar results (z = −3.921, p < .001). The nemes, the number of phonemes ranged from 21 to 40
children produced from two to 39 in-phonology words (M = 28.45, SD = 5.25) and the total number of words
(M = 21.75, SD = 11.15) and from zero to eight out-of- ranged from five to 50 (M = 27.60, SD = 13.09). When
phonology words (M = 2.90, SD = 2.40). In-phonology examining Table 2, the relationship between the number
and out-of-phonology word learning was evidenced by the of words and the number of phonemes is more evident for
children with the fewest number of spoken words and the the children who had the least and the greatest number of
children with the greatest number of words. In fact, the gap words (i.e., the children at the lowest and the highest end
between in-phonology and out-of-phonology words grew of the table). For example, Child 2, who produced close
wider as the children’s spoken vocabulary grew larger. As to the fewest number of words (eight words), had the few-
the children’s total number of words increased, their in- est number of phonemes (21 phonemes). Likewise, Child
phonology words increased disproportionately to their 19 and Child 20 who had the greatest number of words
number of out-of-phonology words. For example, the chil- (48 and 50, respectively) had close to the greatest number
dren with the fewest number of spoken words were Child 1 of phonemes (35 and 36, respectively).
and Child 2. Child 1 had four in-phonology words and one
out-of-phonology word for a total of five words. Likewise, Speech Sound Inventories
Child 2 had eight in-phonology words and zero out-of-
phonology words for a total of eight words. The children After examining the association between lexical acquisi-
with the greatest number of words were Child 19 and 20. tion and phonological acquisition, it is worthwhile to examine

Figure 1. Number of in-phonology (In-phono) and out-of-phonology (Out-phono) words used by the participants. The children are listed in
order from left to right by increasing the number of spoken words. The children’s spoken words were obtained from the MacArthur–Bates
Communicative Development Inventories (Fenson et al., 2007). The speech sound inventories were obtained from the utterances produced
during assessment/play sessions with the children. The children’s spoken words from the Communicative Development Inventories (CDI)
were categorized as being either in-phonology or out-of-phonology based on whether the words began with a phoneme in the child’s exist-
ing speech sound inventory.

1080 Language, Speech, and Hearing Services in Schools • Vol. 53 • 1074–1087 • October 2022
Figure 2. A Spearman rank-order correlation was performed to determine the relationship between the number of spoken words and the
number of phonemes in the children’ speech sound inventory. The plot indicated a moderate positive correlation between the number of
words on the Communicative Development Inventories (CDI) and the number of phonemes produced by the children.

the children’s speech sound inventories, as displayed in Our hypothesis regarding our first question was confirmed
Table 2. Recall that the children ranged in age from 28 to in that the children with ASDs in this study overwhelmingly
72 months with a mean age of 55 months. The children produced spoken words that began with consonants from
produced seven to 20 out of the 20 vowels, which included their existing speech sound inventory. According to the the-
monothongs and diphthongs (M = 13.25, SD = 3.62) and ory set forth by Schwartz and Leonard (1982), the children
10–21 of the 24 consonants (M = 15.10, SD = 2.32). Many selected and avoided words according to their individual
of the children in this study (19 out of 20) produced seven phonological preferences. The children were heavily influ-
or eight of the Early 8 consonants (/m, n, j, b, w, d, p, and enced by their existing speech sound inventories as opposed
h/), which are consistently produced by their NT peers at to speech sounds, which they had yet to develop. Moreover,
36 months (Leonard et al., 1982). However, the children this study appears to confirm the notion that the children
were inconsistent in their production of the Middle 8 con- with ASDs have similar word learning styles as NT chil-
sonants (/t, ŋ, k, g, f, v, tʃ, and dʒ/), which are consistently dren in the same stage of development. We posit that these
produced by the age of 66 months. In fact, not one child children were able to use their auditory–articulatory feed-
produced all of the Middle or Late 8 consonants. It is inter- back loop, which is a foundational skill of speech produc-
esting to note that approximately three quarters of the chil- tion, during in-phonology and out-of-phonology word
dren (15–17 children) produced /t, k, g, and ʃ/, but only learning (Stoel-Gammon, 2011). The children may have
approximately half of the children (nine to 10 children) become aware of the tactile and kinesthetic sensations
produced /ŋ, f, and tʃ/ and less than a quarter of the chil- related to speech production of their preferred phonemes,
dren (two to four children) produced /v and dʒ/. Of the and “heard the acoustic output creating an auditory articu-
Late 8 consonants (/ʃ, ʒ, l, r, s, z, θ, ð, and clusters/), which latory feedback loop” (Stoel-Gammon, 2011, p. 9).
are consistently produced by the age of 90 months, almost The second component we explored was a relation-
all the children (19 of the 20) produced /ʃ/, /s/, or /z/; half ship between the number of words and the number of
of the children (10 of the 20) produced /l/ or /r/; and less phonemes produced by the children in this study. Our
than a quarter of the children (two children) produced /ʒ/ hypothesis for the second question was somewhat con-
or /θ and ð/. The children exhibited typical sound substitu- firmed. The children’s lexical acquisition and phonological
tions such as p/f, d/ θ, d/ð, t/k, d/g, w/l, and w/r. development seemed to be proportional only among the
children who had the least and the greatest number of
words. The child in the study (Child 2), who produced
Discussion close to the fewest number of words, had the fewest num-
ber of phonemes, perhaps because the child’s limited con-
In this study, we examined two components of lexi- sonant inventory could not support an expanded reper-
cal acquisition and phonological development in children toire (Stoel-Gammon, 2011). Conversely, the two children
with ASDs who were in the first 50-word stage of lan- who produced the most words (Child 19 and Child 20)
guage development. The first component we investigated had close to the greatest number of phonemes, suggesting
was in-phonology and out-of-phonology word learning. that the demands of their more developed vocabulary

Biller & Yeager: Lexical and Phonological Development in ASDs 1081


Table 2. Number of words, number of consonants, and consonant inventories for the participants.

Number Number of
Child Age (mo.) of words phonemes Speech sound inventory (vowels and consonants)

1 28 5 26 10 V /I, i, æ, u, ɔ, ɑ, ə, ʌ, aʊ, eɪ/


16 C /p, b, t, d, k, g, m, n, s, ʃ, ʒ, tʃ, w, j, h/a
2 45 8 21 11 V /i, ɛ, æ, u, ʊ, ɔ, ɑ, ə, ʌ, eɪ, oʊ/
10 C /b, d, g, m, n, f, v, w, j, h/a
3 70 17 28 11 V /I, i, ɛ, æ, u, ɔ, ɑ, ʌ, aɪ, eɪ, oʊ /
17 C /b, t, d, k, m, n, ŋ, f, s, z, ʃ, tʃ, dʒ, w, j, r, h/a
4 48 18 28 12 V /I, i, ɛ, æ, u, ʊ, ɔ, ɑ, ʌ, aɪ, eɪ, oʊ/
14 C /p, b, t, d, k, m, n, f, s, tʃ, dʒ, w, j, h/a
5 53 19 22 7V /i, ʊ, ɑ, ə, ʌ, aɪ, oʊ/
15 C /p, b, t, d, k, g, m, n, f, s, ʃ, tʃ, w, j, h/a
6 71 19 31 16 V /I, i, ɛ, æ, u, ʊ, ɔ, ɑ, ə, ʌ, aʊ, aɪ, eɪ, oʊ, ɛr, ar/
15 C /b, t, d, k, g, m, n, s, z, tʃ, dʒ, w, j, l, h/a
7 45 19 27 11 V /I, i, ɛ, æ, u, ɑ, ə, ʌ, aɪ, eɪ, oʊ/
16 C /p, b, t, d, k, g, m, n, f, s, z, ʃ, w, j, r, h/a
8 67 22 25 11 V /I, i, u, ɔ, ɑ, ə, ʌ, aʊ, aɪ, eɪ, oʊ/
14 C /p, b, t, d, k, g, m, n, ŋ, s, tʃ, w, j, h/a
9 52 23 24 12 V /I, i, æ, u, ʊ, ɔ, ɑ, ə, ʌ, aɪ, eɪ, oʊ/
12 C /p, b, t, d, k, m, n, s, ʃ, w, j, h/a
10 68 24 31 16 V /I, i, ɛ, æ, u, ʊ, ɑ, ə, ʌ, ɚ, ɝ, aɪ, eɪ, oʊ, ɛr, ar/
15 C /p, b, t, d, k, g, m, ŋ, s, z, ʃ, w, j, r, h/a
11 59 25 24 11 V /i, ɛ, æ, u, ɔ, ɑ, ʌ, aɪ, ɔɪ, eɪ, oʊ/
13 C /p, b, t, d, k, g, m, n, s, z, ʃ, l, h/
12 43 26 21 8V /i, æ, ɑ, ʌ, aʊ, aɪ, eɪ, oʊ/
13 C /b, t, d, g, m, n, f, s, z, w, j, l, h/a
13 46 30 22 10 V /I, i, æ, u, ʊ, ɑ, ʌ, aʊ, aɪ, eɪ/
13 C /p, b, t, d, k, g, m, n, ŋ, s, ʃ, w, h/a
14 40 30 28 13 V /I, i, ɛ, æ, u, ɔ, ɑ, ʌ, aʊ, aɪ, eɪ, oʊ, ɪr/
15 C /p, b, t, d, k, g, m, n, ŋ, s, ʃ, tʃ, w, j, h/a
15 71 35 40 19 V /I, i, ɛ, æ, u, ʊ, ɔ, ɑ, ə, ʌ, ɚ, ɝ, aʊ, aɪ, eɪ, oʊ, ɪr, ɛr, ɔr/
21 C /p, b, t, d, k, g, m, n, ŋ, f, v, θ, ð, s, z, ʃ, w, j, l, r, h/a
16 42 44 33 16 V /I, i, ɛ, æ, u, ʊ, ɑ, ə, ʌ, ɚ, ɝ, aʊ, aɪ, eɪ, oʊ, ɛr/
17 C /p, b, t, d, k, g, m, n, ŋ, s, z, ʃ, tʃ, w, j, l, h/a
17 69 44 35 17 V /I, i, ɛ, æ, u, ʊ, ɔ, ɑ, ə, ʌ, ɚ, ɝ, aʊ, aɪ, eɪ, oʊ, ɛr/
18 C /p, b, t, d, k, g, m, n, ŋ, θ, ð, s, z, ʃ, tʃ, w, j, h/a
18 72 46 31 16 V /I, i, ɛ, æ, u, ʊ, ɔ, ɑ, ə, ʌ, ɝ, aʊ, aɪ, ɔɪ, eɪ, oʊ/
15 C /p, b, t, d, m, n, s, z, ʃ, ʒ, tʃ, w, j, r, h/a
19 53 48 35 18 V /I, i, ɛ, æ, u, ʊ, ɔ, ɑ, ə, ʌ, ɚ, aɪ, ɔɪ, eɪ, oʊ, ɪr, ɛr, ɔr/
17 C /p, b, t, d, k, m, n, ŋ, f, s, ʃ, tʃ, dʒ, w, j, r, h/a
20 62 50 36 20 V /I, i, ɛ, æ, u, ʊ, ɔ, ɑ, ə, ʌ, ɚ, ɝ, aʊ, aɪ, ɔɪ, eɪ, oʊ, ɪr, ɛr, ɔr/
16 C /p, b, t, d, k, g, m, n, ŋ, f, s, z, ʃ, w, j, h/a

Note. The children are listed in order from top to bottom by increasing the number of spoken words. Listed are the consonants produced
correctly. V = vowels; C = consonants.
a
At least seven out of the Early 8 consonants are present (Shriberg, 1993).

created a demand for a larger and more complex phonolo- seemed to avoid certain sound or found them harder to
gical system. This notion is in keeping with the bidirec- produce (/ŋ, f, v, tʃ, and dʒ/). Their selection and avoidance
tional relationship between phonological skills and lexical of these sounds were mirrored in their in-phonology and
development (Stoel-Gammon, 2011). out-of-phonology words. Perhaps a study with a larger
Besides looking at the relationships between phonol- number of participants could determine the sequence of
ogical development and lexical acquisition, the children’s phoneme development of children with autism as compared
individual consonants should be examined for a trajectory to their NT peers. Additionally, there needs to be studies
of development. The children followed a typical sequence that examine the individual phonological characteristics of
by producing the age-appropriate Early 8 phonemes. minimally verbal children with ASDs (Leonard et al., 1982;
However, they were inconsistent in production of the Mid- Saul & Norbury, 2020 Yoder et al., 2015). Whereas this
dle and Late 8 phonemes. Their production of phonemes study contains a description of the children’s number of
tended to coincide with their number of words and not vowels and consonants and the children’s consonant pro-
with their chronological age. Interestingly, the children duction in relation to the Early, Middle, and Late 8 classifi-
seemed to either prefer certain phonemes or found them cation system, there was no statistical analysis of the chil-
easier to produce (/t, k, g, and ʃ/). Conversely, they also dren’s phonological characteristics.

1082 Language, Speech, and Hearing Services in Schools • Vol. 53 • 1074–1087 • October 2022
Overall, the children in this study had small lexicons with ASDs (Yoder et al., 2015). The children in our study
for their chronological age. The average age of the children demonstrated the Early 8 phonemes, which are consis-
was 55 months, and yet the average number of words they tently produced at approximately 36 months. However,
produced was 28. The small lexicon could have been due beyond that measure, it is difficult to ascertain if they
to lack of practice. They may not have been speaking long have developed a large enough phonetic inventories to
enough to use a variety of phonemes and syllable shapes support lexical growth. Besides phonetic inventory, the
and mastering a variety of articulatory movements to pro- remaining three predictive factors are in question. With-
duce an assortment of words. However, we did not gather out testing for the other three factors, we do not know
information about each child’s date of onset of spoken lan- if or how they contributed to the success or failure of
guage; hence, we cannot determine how long each child in the children in this study to expand their spoken lan-
the study had been talking and practicing spoken language. guage. Although not a predictive model for spoken lan-
Another possibility for such small lexicons is that guage, Brady et al. (2021) suggests using an altogether
the children were lacking the predictive factors that con- different form of measuring children’s progress in speech
tribute to expansion of expressive language (Yoder et al., production—a phonemic scoring method. This method
2015). Two of these factors have been the child’s phonetic measures the production of every consonant and every
repertoire and consonant inventory (Chenausky et al., vowel in a target consonant–vowel–consonant word. Con-
2018; Saul & Norbury, 2020; Wetherby et al., 2007; sonants are measured on three features, which are place,
Yoder et al., 2015). A phonetic repertoire has been manner, and voicing. Vowels are measured on four fea-
defined as the number of consonants imitated correctly on tures, which are tongue height, advancement, rounding and
an experimental task (Chenausky et al., 2018) and has tenseness. This scoring system is sensitive to production
been shown to be a predictor of expressive vocabulary gains by showing improvement in production of phonemes
growth in children with ASDs (Chenausky et al., 2018; and syllables before gains are seen at the word level.
Smith et al., 2007; Thurm et al., 2007). A consonant
inventory has been described as a set of consonants pro- Clinical Implications
duced spontaneously by participants. Wetherby et al.
(2007) found a substantial correlation between the inven- Knowing that children with ASDs produce new spo-
tory of consonants on the Communication and Symbolic ken words based on phonological selection can have an
Behavior Scales Developmental Profile Infant–Toddler enormous impact on their treatment programs. The
Checklist (Wetherby & Prizant, 2002) and the inventory lexical–phonological association for a child can be deter-
of words produced by children with ASDs and children mined using results from a CDI and the child’s consonant
with developmental delays. Additionally, they found that inventory. Since word comprehension precedes production
the best predictors of verbal skills at age 3 years were acts by at least 5 months (Benedict, 1979), the words that chil-
for behavior regulation and inventory of consonants. dren understand on the CDI can be used as a list of
In a study by Yoder et al. (2015), extensive model- potential target words to be trained in therapy. A clinician
ing was used to determine four value-added predictors of could select words from the list that begin with the pho-
expressive language growth for preschoolers with ASDs. nemes from the child’s speech sound inventory. Other fac-
Value-added predictors are determined by considering the tors for word selection such as semantic class, syllable
previous correlation between predictors used during the shape, and the child’s own preferred items would be con-
modeling process. The four predictive factors were inten- sidered as well. This method of selecting target words to
tional communication, consonant inventory, responding to expand a child’s spoken vocabulary can be enhance the
joint attention, and parent/caregiver linguistic responses. use of augmentative alternative communication. One
A more recent study has investigated phonetic repertoire example would be using a speech-generating application,
and consonant inventory in children with ASDs (Saul & installed on a tablet computer (Apple iPad, Microsoft Sur-
Norbury, 2020). Twenty-seven preschool children with face Pro, Samsung Galaxy Tab, etc.), and programming a
ASDs were seen by the researchers in their homes every set of words that match the criteria for target words
4 months over a 12-month period. The researchers found described in this study. All the words on the tablet screen
that both the phonetic repertoire and consonant inventory could be “masked” (or hidden) except for the target words
were significant in predicting the children’s number of spo- meeting the lexical–phonological criteria, and only those
ken words for the vocabulary measures completed at every words would be trained. Since the words begin with con-
4-month interval. Saul and Norbury (2020) confirmed the sonants children are already producing, they would have a
importance of consonant production to the development of better chance of imitating the words while using the
spoken language in children with ASDs. speech-generating device. As the children learn one set of
We concur with Yoder et al. that there needs to be words, more words that meet the lexical–phonological cri-
further research in consonant inventory growth in children teria could be unmasked, slowly expanding their target

Biller & Yeager: Lexical and Phonological Development in ASDs 1083


words and growing their vocabulary/lexicon with their small doses of discrete trial training (DTT) using traditional
phonological skills. articulation therapy involving auditory, visual, and tactile
The lexical–phonological association has been consid- cues. For example, treatment sessions could include a speech
ered in very few treatment studies. A clinical case study practice time in which the child receives focused trials of the
reported by Biller and Johnson (in press) used a combined target words similar to “shaping procedures characteristic of
intervention program, incorporating both communication the DTT approach, which allows the child to produce at
and speech production strategies with a minimally verbal first relatively gross approximations of target words that are
child with ASDs. The study examined the spoken produc- gradually required to move closer to adult forms” (Tsiouri
tion of preselected words by the participant, using a com- et al., 2012, p. 1290). In communication intervention studies
bined intervention approach, consisting of three communi- using DTT, children have been taught receptive and expres-
cation strategies and three speech production strategies dur- sive language skills, joint attention, symbolic play, verbal
ing structured play. The target words were chosen based on and motor imitation, production of individual sounds, sin-
the in-phonology and out-of-phonology word learning pre- gle words, or phrases, (Biller & Johnson, 2022; Goldstein,
sented in this study. The results indicated that the child 2002; Kaiser et al., 2017, Kaiser et al., 2006; Paul et al.,
demonstrated improved spontaneous and imitative produc- 2013; Prizant et al., 2000; Tsiouri et al., 2012).
tion of the target words during the treatment and main-
tained production of the target words 1 month after the
treatment ended. Additionally, the participant demon- Limitations and Future Research
strated spontaneous production of seven out of the eight
target words on the posttreatment MacArthur–Bates CDI. A primary limitation of this study was the small sam-
One treatment study reported that consonant inven- ple size, which was partially due to the inclusionary criteria.
tories were considered when selecting target words for There were several children with autism whose families were
intervention (Brady et al., 2015). A study by Brady et al. willing to participate in the study; however, the children
(2015) used a multimodal approach (e.g., the use of speech were either nonverbal or had a spoken vocabulary over 50
and a speech generating device during intervention) to words. Since the study focused on the first 50-word stage of
teach preselected spoken vocabulary to minimally verbal language development, we strictly adhered to that criterion.
children with ASDs, based on a child’s initial speech Even though there was strong evidence that the children in
sound repertoire, phonotactic probability, and neighbor- this study used in-phonology and out-of-phonology word
hood density. Phonotactic probability refers to the fre- learning, a larger number of participants could give a
quency of occurrence of sound combinations in the lan- clearer understanding of the relationship between number of
guage, which can have either a high or a low probability words produced by the child and the number of phonemes
of occurrence. Neighborhood density refers to a cluster of present in a child’s speech sound inventories.
words that sound similar, differing by only one sound sub- A second consideration was the method used to
stitution, addition, or deletion (e.g., cat, cad, cap, and gather the children’s speech sound inventories. Unlike the
cab). The researchers chose high-probability and high- methodology used by Leonard et al. (1982), this study
density target words that were easier for the children to analyzed imitative, elicited, and spontaneous productions
produce. Results of Brady et al.’s study indicated that in all positions of words to acquire the children’s conso-
their intervention package promoted improved spoken nant inventories. Combining the child’s productions dur-
words for some minimally verbal children with ASDs. ing the assessment sessions allowed for a broader range of
Furthermore, the authors reported that the children more phonemes to be counted that may not have been heard if
easily acquired new words that incorporated consonants the child was required to perform an imitative task. To
from their existing consonant repertoire. make the speaking situations as uniform across the partici-
A second clinical implication from this study per- pants as possible, the participants were recorded during
tains to the children’s individual phonetic inventories. the administration of the MSEL while carrying out the
Whereas many of the children in this study produced most communication temptation activities from the CSBS.
of the Early 8 consonants and some of the Middle 8 con- However, those speaking situations are not as uniform as
sonants, there were some idiosyncrasies among the chil- using a list of preselected targets used in the study of
dren’s consonant inventories, such as the absence of /f/. Schwartz and Leonard. Additionally, the children’s pro-
After conducting a probe, it would be beneficial to teach ductions could have been divided by position in the word
each child the age-appropriate consonants missing from (e.g., initial, medial, or final position) to yield a more
their inventory. By teaching the missing consonants (and refined analysis. A future study might compare the use of
vowels), it would give the child an expanded phonological more uniform stimuli to elicit speech sound production and
base with which to increase lexical acquisition. Teaching the use of spontaneous language samples to determine the
the age-appropriate phonemes could be accomplished with best sampling method for obtaining consonant inventories

1084 Language, Speech, and Hearing Services in Schools • Vol. 53 • 1074–1087 • October 2022
from children with ASDs. Brady et al. (2015) used another American Psychiatric Association. (2013). Diagnostic and statisti-
type of sampling method that provided an even more cal manual of mental disorders (5th ed.). https://doi.org/10.
1176/appi.books.9780890425596
extensive inventory of the children’s speech sounds—a Lan- Bates, E., Bretherton, I., & Snyder, L. (1988). From first words to
guage Environmental Analysis (LENA) device. It is a small grammar: Individual differences and dissociable mechanisms.
recorder often worn in a vest that records continuous ver- Cambridge University Press.
bal productions of the children. Benedict, H. (1979). Early lexical development: Comprehension
It would be noteworthy to conduct a treatment and production. Journal of Child Language, 6(2), 183–200.
https://doi.org/10.1017/S0305000900002245
study with children with ASDs that compares use of in- Bhat, A. N., Landa, R. J., & Galloway, J. C. (2011). Current per-
phonology and out-of-phonology words as target stimuli. spectives on motor functioning in infants, children, and adults
One would predict that the in-phonology words would be with autism spectrum disorders. Physical Therapy, 91(7),
learned at a faster rate than the out-of-phonology words. 1116–1129. https://doi.org/10.2522/ptj.20100294
In conclusion, there appears to be mounting evi- Bleile, K. M. (2004). Manual of articulation and phonological dis-
orders: Infancy through adulthood, second edition. Thomson/
dence that phonological skills have an impact on expres- Delmar Learning.
sive language growth of children with ASDs (Chenausky Brady, N. C., Kosirog, C., Fleming, K., & Williams, L. (2021).
et al., 2018; Saul & Norbury, 2020; Yoder et al., 2015). Predicting progress in word learning for children with autism and
Children with autism who have a limited spoken vocabu- minimal verbal skills. Journal of Neurodevelopmental Disorders,
lary should be given the same opportunity to improve 13(1), Article 36. https://doi.org/10.1186/s11689-021-09386-x
Brady, N. C., Storkel, H. L., Bushnell, P., Barker, R. M.,
their speech sound production as other clinical popula- Saunders, K., Daniels, D., & Fleming, K. (2015). Investigating
tions. Leonard et al. (1982) suggested that target words in a multimodal intervention for children with limited expressive
treatment programs should adapt to the child’s phonolo- vocabularies associated with autism. American Journal of
gical characteristics because it is an effective method of Speech-Language Pathology, 24(3), 438–459. https://doi.org/
expanding of a child’s scant lexicon. They made this state- 10.1044/2015_AJSLP-14-0093
Biller, M. F., & Johnson, C. J. (2022). A combined spoken lan-
ment about phonological characteristics and children with guage intervention for a minimally verbal child with autism: A
specific language impairment; however, could it hold true clinical case study. Perspectives of the ASHA Special Interest
for children with autism as well. Groups, 7(3), 696–711. https://doi.org/10.1044/2022_PERSP-21-
00253
Broome, K., McCabe, P., Docking, K., & Doble, M. (2017). A
systematic review of speech assessments for children with au-
Data Availability Statement tism spectrum disorder: Recommendations for best practice.
American Journal of Speech-Language Pathology, 26(3),
The data sets generated during and/or analyzed dur- 1011–1029. https://doi.org/10.1044/2017_ajslp-16-0014
ing this study are available from the corresponding Charman, T., Baron-Cohen, S., Swettenham, J., Baird, G., Drew,
author. They are stored on the campus of the University A., & Cox, A. (2003). Predicting language outcome in infants
with autism and pervasive developmental disorder. Interna-
of Central Arkansas and are available upon reasonable
tional Journal of Language & Communication Disorders, 38(3),
request. The data sets contain analyses of language sam- 265–285. https://doi.org/10.1080/136820310000104830
ples and speech sound inventories of minimally verbal Chenausky, K., Brignell, A., Morgan, A., & Tager-Flusberg, H.
children with autism, ages 2;11–6;0 (years;months). (2019). Motor speech impairment predicts expressive language
in minimally verbal, but not low verbal, individuals with
autism spectrum disorder. Autism & Developmental Language
Impairments, 4. https://doi.org/10.1177/2396941519856333
Acknowledgments Chenausky, K., Norton, A., Tager-Flusberg, H., & Schlaug, G.
(2018). Behavioral predictors of improved speech output in
This work was supported by a University Research minimally verbal children with autism. Autism Research,
Grant from the University of Central Arkansas (Grant 11(10), 1356–1365. https://doi.org/10.1002/aur.2006
348R09 awarded to Maysoon F. Biller). The authors would De Giacomo, A., Portoghese, C., Martinelli, D., Fanizza, I.,
L’Abate, L., & Margari, L. (2009). Imitation and communica-
like to thank the many families and children who partici- tion skills development in children with pervasive develop-
pated in this study, one spoken word at a time. They would mental disorders. Neuropsychiatric Disease and Treatment, 5,
also like to thank the agencies in their region who promoted 355–362. https://doi.org/10.2147/NDT.S5679
the study and aided them in securing participants. Dyson, A. T. (1988). Phonetic inventories of 2- and 3-year-old
children. Journal of Speech and Hearing Disorders, 53(1), 89–93.
https://doi.org/10.1044/jshd.5301.89
Eigsti, I.-M., de Marchena, A. B., Schuh, J. M., & Kelley, E.
References (2011). Language acquisition in autism spectrum disorders: A
developmental review. Research in Autism Spectrum Disorders,
Adams, L. (1998). Oral-motor and motor-speech characteristics of 5(2), 681–691. https://doi.org/10.1016/j.rasd.2010.09.001
children with autism. Focus on Autism and Other Developmental Dis- Fenson, L., Dale, P. S., Reznick, J. S., Bates, E., Thal, D. J., &
abilities, 13(2), 108–112. https://doi.org/10.1177/108835769801300207 Pethick, S. J. (1994). Variability in early communicative

Biller & Yeager: Lexical and Phonological Development in ASDs 1085


development. Monographs of the Society for Research in Child of autism. Journal of Autism and Developmental Disorders,
Development, 59(5), i+iii-v+1–185. https://doi.org/10.2307/ 36(6), 807–828. https://doi.org/10.1007/s10803-006-0111-4
1166093 Koegel, L. K., Bryan, K. M., Su, P. L., Vaidya, M., & Camarata,
Fenson, L., Marchman, V. A., Thal, D. J., Dale, P. S., Reznick, S. (2020). Definitions of nonverbal and minimally verbal in
J. S., & Bates, E. (2007). MacArthur–Bates Communicative research for autism: A systematic review of the literature. Jour-
Development Inventories: User’s guide and technical manual nal of Autism and Developmental Disorders, 50(8), 2957–2972.
(2nd ed.). Brookes. https://doi.org/10.1007/s10803-020-04402-w
Gernsbacher, M. A., Sauer, E. A., Geye, H. M., Schweigert, Leonard, L. B., Schwartz, R. G., Chapman, K., Rowan, L. E.,
E. K., & Goldsmith, H. H. (2008). Infant and toddler oral- Prelock, P. A., Terrell, B., Weiss, A. L., & Messick, C.
and manual-motor skills predict later speech fluency in (1982). Early lexical acquisition in children with specific lan-
autism. The Journal of Child Psychology and Psychiatry, guage impairment. Journal of Speech and Hearing Research,
49(1), 43–50. https://doi.org/10.1111/j.1469-7610.2007.01820.x 25(4), 554–564. https://doi.org/10.1044/jshr.2504.554
Goldstein, H. (2002). Communication intervention for children Lord, C., & Paul, R. (1997). Language and communication in
with autism: A review of treatment efficacy. Journal of Autism autism. In D. J. Cohen & F. R. Volkmar (Eds.), Handbook
and Developmental Disorders, 32(5), 373–396. https://doi.org/ of autism and pervasive developmental disorders (2nd ed.,
10.1023/A:1020589821992 pp. 195–225). Wiley.
Hart, B., & Risley, T. R. (1975). Incidental teaching of language Lord, C., Rutter, M., Dilavore, P. C., Risi, S., Gotham, K., &
in the preschool. Journal of Applied Behavior Analysis, 8(4), Bishop, S. L. (2012). Autism Diagnostic Observation Schedule–
411–420. https://doi.org/10.1901/jaba.1975.8-411 Second Edition (ADOS-2). WPS.
Hoff, E. (2003). The specificity of environmental influence: Socio- Luyster, R., Lopez, K., & Lord, C. (2007). Characterizing com-
economic status affects early vocabulary development via municative development in children referred for autism spec-
maternal speech. Child Development, 74(5), 1368–1378. https:// trum disorders using the MacArthur–Bates Communicative
doi.org/10.1111/1467-8624.00612 Development Inventory (CDI). Journal of Child Language,
Horvath, S., McDermott, E., Reilly, K., & Arunachalam, S. 34(3), 623–654. https://doi.org/10.1017/S0305000907008094
(2018). Acquisition of verb meaning from syntactic distribu- MacRoy-Higgins, M., Schwartz, R. G., Shafer, V. L., & Marton,
tion in preschoolers with autism spectrum disorder. Language, K. (2013). Influence of phonotactic probability/neighbourhood
Speech, and Hearing Services in Schools, 49(3S), 668–680. density on lexical learning in late talkers. International Journal of
https://doi.org/10.1044/2018_lshss-stlt1-17-0126 Language & Communication Disorders, 48(2), 188–199. https://
Howlin, P. (2003). Outcome in high-functioning adults with doi.org/10.1111/j.14606984.2012.00198.x
autism with and without early language delays: Implications Mullen, E. (1995). Mullen Scales of Early Learning. Pearson Clin-
for the differentiation between autism and Asperger syn- ical Assessment.
drome. Journal of Autism and Developmental Disorders, 33(1), Nicolosi, L., Harryman, E., & Kresheck, J. (2006). Terminology
3–13. https://doi.org/10.1023/a:1022270118899 of communication disorders: Speech-language-hearing (4th ed.).
Ingram, D. (1989). Phonological disability in children (2nd ed.). Lippincott Williams & Wilkins.
Whurr. Page, J., & Boucher, J. (1998). Motor impairments in children
Jimenez, E., Haebig, E., & Hills, T. T. (2021). Identifying areas with autistic disorder. Child Language Teaching and Therapy,
of overlap and distinction in early lexical profiles of children 14(3), 233–259. https://doi.org/10.1177/026565909801400301
with autism spectrum disorder, late talkers, and typical Patten, E., Belardi, K., Baranek, G. T., Watson, L. R., Labban,
talkers. Journal of Autism and Developmental Disorders, 51(9), J. D., & Oller, D. K. (2014). Vocal patterns in infants with
3109–3125. https://doi.org/10.1007/s10803-020-04772-1 autism spectrum disorder: Canonical babbling status and vocaliza-
Kaiser, A. P., Scherer, N. J., Frey, J. R., & Roberts, M. Y. tion frequency. Journal of Autism and Developmental Disorders,
(2017). The effects of enhanced milieu teaching with phonolo- 44(10), 2413–2428. https://doi.org/10.1007/s10803-014-2047-4
gical emphasis on the speech and language skills of young Paul, R., Campbell, D., Gilbert, K., & Tsiouri, I. (2013). Compar-
children with cleft palate: A pilot study. American Journal of ing spoken language treatments for minimally verbal pre-
Speech-Language Pathology, 26(3), 806–818. https://doi.org/ schoolers with ASD spectrum disorders. Journal of Autism
10.1044/2016_AJSLP-16-0008 and Developmental Disorders, 43(2), 418–431. https://doi.org/
Kasari, C., Freeman, S., & Paparella, T. (2006). Joint attention 10.1007/s10803-012-1583-z
and symbolic play in young children with autism: A random- Plumb, A. M., & Wetherby, A. M. (2013). Vocalization develop-
ized controlled intervention study. The Journal of Child Psy- ment in toddlers with autism spectrum disorder. Journal of
chology and Psychiatry, 47(6), 611–620. https://doi.org/10. Speech, Language, and Hearing Research, 56(2), 721–734.
1111/j.1469-7610.2005.01567.x https://doi.org/10.1044/1092-4388(2012/11-0104)
Kasari, C., Kaiser, A., Goods, K., Nietfeld, J., Mathy, P., Landa, Prizant, B. M., Wetherby, A. M., & Rydell, P. (2000). Communi-
R., Murphy, S., & Almirall, D. (2014). Communication inter- cation intervention issues for children with autism spectrum
ventions for minimally verbal children with autism: A sequen- disorders. In A. Wetherby & B. Prizant (Eds.), Autism spec-
tial multiple assignment randomized trial. Journal of the trum disorders: A transactional developmental perspective (pp.
American Academy of Child & Adolescent Psychiatry, 53(6), 193–224). Brookes.
635–646. https://doi.org/10.1016/j.jaac.2014.01.019 Psychology Research and Reference. (2022). Developmental quotient.
Kazak, S., Collis, G. M., & Lewis, V. (1997). Can young people http://psychology.iresearchnet.com/developmental-psychology/per-
with autism refer to knowledge states? Evidence from their sonality-and-intelligence/developmental-quotient/
understanding of “know” and “guess.” Journal of Child Psychol- Rapin, I., & Dunn, M. A. (1997). Language disorders in children
ogy and Psychiatry and Allied Disciplines, 38(8), 1001–1009. with autism. Seminars in Pediatric Neurology, 4(2), 86–92.
https://doi.org/10.1111/j.14697610.1997.tb01617.x https://doi.org/10.1016/s1071-9091(97)80024-1
Kelley, E., Paul, J. J., Fein, D., & Naigles, L. R. (2006). Residual Rapin, I., Dunn, M. A., Allen, D. A., Stevens, M. C., & Fein, D.
language deficits in optimal outcome children with a history (2009). Subtypes of language disorders in school-age children

1086 Language, Speech, and Hearing Services in Schools • Vol. 53 • 1074–1087 • October 2022
with autism. Developmental Neuropsychology, 34(1), 66–84. Stone, W. L., & Yoder, P. J. (2001). Predicting spoken language
https://doi.org/10.1080/87565640802564648 level in children with autism spectrum disorders. Autism, 5(4),
Rescorla, L., & Safyer, P. (2013). Lexical composition in children 341–361. https://doi.org/10.1177/1362361301005004002
with autism spectrum disorder (ASD). Journal of Child Lan- Swensen, L. D., Kelley, E., Fein, D., & Naigles, L. R. (2007). Pro-
guage, 40(1), 47–68. https://doi.org/10.1017/s0305000912000232 cesses of language acquisition in children with autism: Evi-
Rogers, S. J., Hayden, D., Hepburn, S., Charlifue-Smith, R., dence from preferential looking. Child Development, 78(2),
Hall, T., & Hayes, A. (2006). Teaching young nonverbal chil- 542–557. https://doi.org/10.1111/j.1467-8624.2007.01022.x
dren with autism useful speech: A pilot study of the Denver Tager-Flusberg, H., Lord, C., & Paul, R. (1997). Language and
Model and PROMPT interventions. Journal of Autism and communication in autism. In D. J. Cohen & F. R. Volkmar
Developmental Disorders, 36(8), 1007–1024. https://doi.org/10. (Eds.), Handbook of autism and pervasive developmental dis-
1007/s10803-006-0142-x orders (2nd ed., pp. 195–225). Wiley.
Rowe, M. L. (2012). A longitudinal investigation of the role of Tager-Flusberg, H., Paul, R., & Lord, C. (2005). Language and
quantity and quality of child-directed speech in vocabulary communication in autism. In F. R. Volkmar, R. Paul, A.
development. Child Development, 83(5), 1762–1774. https:// Klin & D. Cohen (Eds.), Handbook of autism and pervasive
doi.org/10.1111/j.1467-8624.2012.01805.x developmental disorders: Diagnosis, development, neurobiology
Saul, J., & Norbury, C. (2020). Does phonetic repertoire in mini- and behavior (Vol. 1, pp. 335–364). Wiley. https://doi.org/10.
mally verbal autistic preschoolers predict the severity of later 1002/9780470939345.ch12
expressive language impairment? Autism, 24(5), 1217–1231. Tek, S., Jaffery, G., Fein, D., & Naigles, L. R. (2008). Do chil-
https://doi.org/10.1177/1362361319898560 dren with autism spectrum disorders show a shape bias in
Schopler, E., Van Bourgondien, M. E., Wellman, G. J., & Love, word learning? Autism Research, 1(4), 208–222. https://doi.
S. R. (2010). Childhood Autism Rating Scale–Second Edition org/10.1002/aur.38
(CARS2). Western Psychological Services. Thurm, A., Lord, C., Lee, L. C., & Newschaffer, C. (2007). Pre-
Schwartz, R. G., & Leonard, L. B. (1982). Do children pick and dictors of language acquisition in preschool children with autism
choose? An examination of phonological selection and avoid- spectrum disorders. Journal of Autism and Developmental Disor-
ance in early lexical acquisition. Journal of Child Language, ders, 37(9), 1721–1734. https://doi.org/10.1007/s10803-006-0300-1
9(2), 319–336. https://doi.org/10.1017/s0305000900004748 Tsiouri, I., Schoen Simmons, E., & Paul, R. (2012). Enhancing
Shriberg, L. D. (1993). Four new speech and prosody-voice the application and evaluation of a discrete trial intervention
measures for genetics research and other studies in develop- package for eliciting first words in preverbal preschoolers with
mental phonological disorders. Journal of Speech and Hear- ASD. Journal of Autism and Developmental Disorders, 42(7),
ing Research, 36(1), 105–140. https://doi.org/10.1044/jshr. 1281–1293. https://doi.org/10.1007/s10803-011-1358-y
3601.105 Warlaumont, A. S., Richards, J. A., Gilkerson, J., & Oller, D. K.
Shriberg, L. D., Paul, R., Black, L. M., & van Santen, J. P. (2014). A social feedback loop for speech development and its
(2011). The hypothesis of apraxia of speech in children with reduction in autism. Psychological Science, 25(7), 1314–1324.
autism spectrum disorder. Journal of Autism Spectrum Disor- https://doi.org/10.1177/0956797614531023
der and Developmental Disorders, 41(4), 405–426. https://doi. Wetherby, A. M., & Prizant, B. M. (2002). Communication and
org/10.1007/s10803-010-1117-5 Symbolic Behavior Scales Developmental Profile (CSBS DP),
Smith, V., Mirenda, P., & Zaidman-Zait, A. (2007). Predictors of First Normed Edition. Brookes.
expressive vocabulary growth in children with autism. Journal Wetherby, A. M., Prizant, B. M., Hutchinson, T., Wetherby, A.,
of Speech, Language, and Hearing Research, 50(1), 149–160. & Prizant, B. (2003). Communication and Symbolic Behavior
https://doi.org/10.1044/1092-4388(2007/013) Scales, Normed Edition. Brookes.
Sosa, A. V., & Stoel-Gammon, C. (2006). Patterns of intra-word Wetherby, A. M., Watt, N., Morgan, L., & Shumway, S. (2007).
phonological variability during the second year of life. Journal Social communication profiles of children with autism spec-
of Child Language, 33(1), 31–50. https://doi.org/10.1017/ trum disorders late in the second year of life. Journal of
S0305000905007166 Autism and Developmental Disorders, 37(5), 960–975. https://
Stoel-Gammon, C. (1985). Phonetic inventories, 15-24 months: A doi.org/10.1007/s10803-006-0237-4
longitudinal study. Journal of Speech and Hearing Research, Yoder, P., Watson, L. R., & Lambert, W. (2015). Value-added
28(4), 505–512. https://doi.org/10.1044/jshr.2804.505 predictors of expressive and receptive language growth in initially
Stoel-Gammon, C. (2011). Relationships between lexical and pho- nonverbal preschoolers with autism spectrum disorders. Journal of
nological development in young children. Journal of Child Lan- Autism and Developmental Disorders, 45(5), 1254–1270. https://doi.
guage, 38(1), 1–34. https://doi.org/10.1017/s0305000910000425 org/10.1007/s10803-014-2286-4

Biller & Yeager: Lexical and Phonological Development in ASDs 1087


Copyright of Language, Speech & Hearing Services in Schools is the property of American
Speech-Language-Hearing Association and its content may not be copied or emailed to
multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.

You might also like