jircd (print) issn 2040–5111

jircd (online) issn 2040–512x

Article

Phonological Development in a Child with

Autism Spectrum Condition: Case Study of an
Intervention

Pia M. Nordgren

University of Gothenburg, Gothenburg, Sweden

Abstract

In this single case study with a multiple baseline approach, a five-year-old Swedish
boy, diagnosed with an autism spectrum condition, and having a bilingual back-
ground, was followed weekly for one year. In order to facilitate the acquisition of pho-
nological contrasts and knowledge of symbolic representations, a training package
with minimal pairs was gradually introduced. A corpus of the utterances in interac-
tion with teachers and a speech pathologist was collected, transcribed, and analyzed.
The results of the study showed a quantitative and qualitative development of sound
production involving the use of new phonological feature types, new syllable con-
structions, and new words during the year. The delayed and deviant speech develop-
ment, both concerning segments, syllables, and word boundaries in this child, is in
line with studies that describe a deviant phonological development in individuals
with autism.

Keywords: autism spectrum condition; Language development; phonological

contrasts; speech production

Introduction
There is a lack of studies that focus on increasing speech output in nonspeak-
ing children with autism spectrum condition (ASC). More specifically there are

* Correspondence address: Email: pia.nordgren@ling.gu.se

jircd vol 6.1 2015  25–51 doi : 10.1558/jircd.v6i1.25

©2015, equinox publishing
26 Phonological Development

practically no studies on school-aged children with ASC and their vocalizations,

production of syllables and acquisition of words. Most studies have been carried
out with speaking pre-school children with ASC below five years of age (Cirrin
and Gilliam, 2008). One explanation for this absence of studies is that children
with ASC and no or limited speech output may previously have been excluded
from studies because of factors such as difficulties estimating the limited oral
speech production these children show (Rapin et al., (2009). Non-speaking chil-
dren with ASC may exhibit great difficulties with all aspects of language: per-
ception, phonological contrasts, production, syntax, vocabulary, semantics and
pragmatics as well as prosodic aspects. Segmental levels, for example, phonemes
and syllables, are not often investigated in ASC. The acquisition of phonologi-
cal contrasts is an important step in early language acquisition and in ASC this
development may be delayed. The present case study seeks to investigate the
nature of early sound production (babbling) and phonological skills in a school-
aged child with autism and limited speech output. Ultimately, the purpose of
the study is to add to the general debate on the extent to which the acquisition
of phonological contrasts is of importance for language learning in autism. For
children with autism, who are at risk of language delay and speech sound disor-
ders this question could have pedagogical implications.

Background
Early Interventions in Autism
Recent research on ASC points to the importance of early interventions. Strat-
egies in for example enhancing communication, influencing joint attention
and modifying behaviors, are described in the literature. These strategies are
grounded in theories that a lack of theory of mind in autism (Baron-Cohen
et al., 1985), leads to social and communicative disabilities. Environmental
support programs, parental support programs, behavioral programs, sign-
language treatment, alternative and augmentative communication programs
are frequently used as interventions in autism. Visual communication strate-
gies, such as The Picture Exchange Communication System, PECS (Bondy and
Frost, 2002) are also frequently used, in order to help the child with ASC to
communicate. In the literature, two types of training strategies are described:
Adult-Directed Instructions and Natural Language Treatment (Rogers et al.,
2006; Grela and McLaughlin, 2006). In Adult-Directed Instructions, the adult
controls the learning environment by selecting activities and materials to
promote learning (Grela and McLaughlin, 2006). Natural Language Treat-
ment uses techniques to follow the child´s initiative, which shifts intervention
control from the adult to the child so that the child is free to select the learning
situations.
 Pia M. Nordgren 27

The present study deviates from recent studies in autism, in that it focuses
on speech production and the acquisition of phonological contrasts and
it combines the two treatment strategies described above. Few studies have
investigated phonological acquisition in children with autism and limited
speech output. In a review article, Goldstein (2002) found a limited number of
articles concerning autism and phonological interventions. An example is one
article (Koegel et al., 1988) in Goldstein’s review, which compared two differ-
ent methods of phonological interventions. In the first condition (The Verbal
Attempts Condition), all the speech attempts were reinforced whether they
were correct or not. The second condition (The Successive Approximation
Condition) included reinforcement of more and more correct speech produc-
tions. The results showed that the most effective method included reinforce-
ment for all verbal attempts, not only if the attempts were correct. Ten years
later, Koegel et al. (1998) presented a study of motivational aspects in speech
training. Correct speech production according to target sounds was mea-
sured in two conditions (A Naturalistic and a Non-Naturalistic Approach).
The results showed that the approach, where the child had an interest in the
objects (The Naturalistic Approach) was the most effective approach. Results
from these two studies form a backdrop for the present study.

Phonological development in autism

Limitations in auditory perception, the acquisition of phonological contrasts
and speech production may restrict language development in children with
ASC and limited speech output. A small group of children with autism never
develop any speech. A third of the children develop some words, but do not
use them for conversational purposes (Hadwin et al., 1997). Rapin et al. (2009)
found two subtypes in children with autism, one subtype with problems con-
cerning perception and production of phonology and another subtype with
semantic-pragmatic difficulties. Delayed phonological development in autism
may be similar to phonological development in intellectual disability (Barto-
lucci et al., 1976; Tager-Flusberg, 1990). A delayed or deviant phonologi-
cal development and the existence of unusual sounds in children with autism
have been suggested (Bartolucci and Pierce, 1977; Wolk and Edwards, 1993;
Wolk and Giesen, 2000). A recent study by Cleland et al. (2010) showed that
children with ASC exhibit normal phonological processes, but that deviant
processes (such as phoneme specific nasal omission and initial consonant
deletion) also occur. Bishop (2004) found no disproportionate phonological
difficulties in persons with ASC and normal verbal ability. In recent years, a
specific mechanism impairing the development of speech in autism has been
suggested, which is oral motor dysfunction (Rogers et al., 2006). Dziuk et al.
(2007) describe the condition as dyspraxia in autism. The basis for under-
28 Phonological Development

standing apraxia, difficulties with motor planning, in relation to social and

communicative deficits is not yet fully understood. The studies on phonology
and dyspraxia, now mentioned, form an additional backdrop for the present
study.
There have been few empirical studies on vocalizations and babbling in
ASC. A recent study, Schoen et al. (2011), found that speech-like vocalizations
similar to those of language-matched peers were produced in the group with
autism, but significantly more atypical non-speech vocalizations occurred
in this group when it was compared to two control groups (Schoen et al.,
2011). Another previous study found no difficulties with canonical babbling
in autism (Scheinkopf et al., 2000). Babbling is relevant in relation to non-
speaking children, as a ground for phonological development. The child in the
present study was non-speaking to a great extent at the beginning of the study.
Therefore it is important to see how his vocalizations compare with babbling
in typical development.

Babbling and Word Acquisition in Typical Child Language Development

Five developmental babbling stages occur during typical children´s first 18
months of life, (Roug et al., 1989). The first stage involves glottal consonants
and nasalized vowels. The second stage, the velar and uvular stage, is char-
acterized by vibration of the uvula. After that follows a period in which the
infant produces a large number of non-consonantal utterances. Roug et al.
(1989) describe these utterances as ‘relatively long vocalizations with non-
speech-like intonation patterns resembling singing patterns rather than
speech’. During the next stage, the reduplicated babbling stage, the child
develops the ability to produce repeated syllables, which starts to occur
around the seventh month (Steinberg and Sciarini, 2006). Most reduplicated
syllables are of the basic consonant-vowel type, while some of them consist of
CVC combinations. In Swedish, children say, for example, mamma [mama]
for ‘mummy’, pappa [papa] for ‘daddy’. The CVCV tier is acquired first in
Swedish. Consonant clusters are acquired later. In the final babbling stage,
the variegated babbling stage, the child produces babbling with varied con-
sonants and sentence-like intonation patterns (Roug et al., 1989). Children
can be said to have learnt their first word when they make a recognizable
speech form in relation to an object or event in the environment (Steinberg et
al., 2006). A crucial step in cognitive and linguistic development is children´s
understanding of symbols with signs for referents. Language can be seen as
a system of signs, a semiotic system, where feature types may distinguish
meaning. Since there are few (if any) studies on babbling and autism in
Swedish children, it will be interesting to follow how the child’s language
development in the present study evolves.
 Pia M. Nordgren 29

Swedish phonology and phonotactics

Swedish has a rich vowel system with nine vowel phonemes (Elert, 1989; Plun-
kett and Strömqvist, 1990), with varying length, long-short, which is distinc-
tive in prosody and can form minimal pairs, for example, [vi:t]-[vɪt]. Swedish
consonants are shown in Table 1. Engstrand (1990) describes Central Stan-
dard Swedish as spoken in the Stockholm area.

Table 1. Swedish consonants are shown

Bilabials Labiodentals Dentals Palatals Velars Glottals

Plosives p,b t,d k,g
Nasals m n ŋ
Fricatives f,v s j h
Approximants ɹ
Laterals l

Swedish has a complex phonotactic structure with a maximum of three

consonants present initially and finally in words (Sigurd, 1965). In inflected
word endings, up to five consonants are allowed. Three phonemic prosodic
features are word accents (accent 1 or 2), duration of vowels (long-short), and
word stress (varied). Words may consist of only one stressed syllable (mono-
syllables), for example, häst, ‘horse’, strå, ‘straw’, or words may consist of a
stressed syllable followed by one or several unstressed syllables, for example,
spindlar, ‘cobs’,´överste, ‘colonel’, baderskorna, ‘the bathing women’, hederli-
gaste, ‘the most honest’, or preceded by one or several unstressed syllables. The
basic phonotactic pattern in Swedish at the early steps of language acquisition
is the alternation between vocalic and consonantal segments (Sigurd, 1965).
Since no studies of the acquisition of the Swedish consonant system in autism
have been carried out, it will be particularly interesting to see how the child in
this study develops consonants.

Bilingualism in autism
Bilingualism is common worldwide, but little is published about how to deal
with this phenomenon in autism. The literature on this matter concludes that
children with autism would rather benefit from a multilingual environment,
when several languages occur at home, rather than the opposite. A recent
study showed no differences in language development between one mono-
lingual and one bilingual group of children with autism concerning use of
spoken language (Valicenti-McDermott et al., 2013). The bilingual children
in their study were more likely to vocalize and utilize gestures. Other studies
have suggested similar results (Ohash et al., 2012; Kremer-Sadlik, 2005).
30 Phonological Development

Despite research results it happens that parents are advised to speak only one
language to their child (Kremer-Sadlik, 2005), in order to create a simplified
linguistic input. The child in the present study has a bilingual background,
which will be discussed below.

Autism and intellectual disability

Deficits in various developmental diagnoses, including intellectual function-
ing, often co-occur with ASC (Levy et al., 2010). Intellectual functioning
includes various mental abilities, such as reasoning, problem solving, plan-
ning, abstract thinking, judgment and academic and learning through experi-
ences. Mental abilities are measured by IQ tests and a score of approximately
two standard deviations below average represents a significant cognitive
deficit, typically IQ score 70 or below.

The current study

The current study is a descriptive case study of an intervention with a multiple
baseline approach. Phonological and syllabic development is investigated for
one year in a boy diagnosed with ASC.

Research questions
1. Can the boy in the present study acquire phonological contrasts, more
advanced syllable constructions and use of new words, following the
intervention?
2. Can a deviant phonological development with presence of atypical
production of speech in this child with ASC, be detected?

Method
Participant
The boy, who we call Ted, exhibits ASC and severe intellectual disability and
was 5;9 years old, at the beginning of the study and 7;2 years at the last
follow-up.
Ted attended a special class for children with autism, which was integrated
in a regular school in Sweden. Since the age of 3, Ted had all his educational
training at this school, which offered him a proficient language model from an
early age. The TEACCH methodology (Schopler et al., 1995), using structured
settings was employed. Between age 3;0 and 5;9, the parents had received
some guidance and courses concerning Augmentative and Alternative Com-
munication (AAC) from a speech pathologist and Ted had received a picture
board. Ted’s vision and hearing were normal. L1 (for both parents) was a
non-Germanic language, while L2 (Ted’s language and the language spoken
 Pia M. Nordgren 31

at home) was Swedish. Both L1 and L2 were spoken at home, together with
some English, until Ted reached the age of 1;6 years. The parents noticed Ted’s
delay in language development (at 1;6 years) and were then given the advice
to change languages at home to only L1, which they did. An investigation
of psychomotor development started (at the age of 2;5 years) and then the
parents switched to using L2, as the only language. Ted was diagnosed with
ASC at 2;9 years of age. The Griffiths Test (Griffith, 1954) and the Vineland
Adaptive Behaviour Scales (Sparrow and Cicchetti, 1985) showed that Ted
scored for autism in the following areas: language, social interaction, play, and
stereotyped behaviors (moderate to severe autism spectrum condition). The
observation also showed a limited attention span and concentration prob-
lems. A speech and language pathologist estimated that Ted at age 2;9 almost
totally lacked language and word comprehension, with an expressive language
of four to five one-syllable utterances. Reynell Language Developmental Scales
(Reynell, 1977) testing at age 5;6 revealed that Ted was able to point out eight
out of 15 objects (nouns) and to follow a simple instruction. Bilabials, dentals,
and velars were judged as established in initial and medial positions, in that
five objects were named with these sounds. The majority of his utterances con-
sisted of one syllable. Reduplicated babbling was rare, but occurred, such in
[pɛpɛpɛ] for ‘apple’, äpple [ɛplə] in Swedish.
At the onset of the present study, Ted exhibited extremely limited oral
speech and communicative skills. Ted used picture boards, pointing, and
facial expressions in communication, to some extent. His expressive language
was estimated to be about five to ten one-syllable utterances (the first or the
second syllable), according to the child’s parents and teachers.
In Table 2 there is a description of Ted’s sounds preceding the study, which
were very few [pɛ], [tɔ:], [ɡʊ], [bi:].

Table 2. Ted´s speech production (in Swedish) before the study is shown

Speech Standard Produced Description Description Syllable

Sounds (in Phonetic Syllables (by of Word of Word Structure in
Swedish) Representation the boy) (in Examples (in Examples (in Produced
(in Swedish) Swedish) Swedish) English) Syllables
[p] [ɛplə] [pɛ] äpple apple CV
[t] [tɔ:ɡ] [tɔ:] tåg train CV
[k] [kɑ:ka] [ɡʊ] kaka cookie CV
[b] [bi:l] [bi:] bil car CV

Procedure and considerations

This study was originally designed as a single-subject design, with repeated
measurements of the dependent variable, but due to the participant’s very
limited speech output and involvement, the study is better described as a case
32 Phonological Development

study. Some changes and special adjustments to the research design were also
necessary. However, the single-subject strategy was the preferred starting
point for the study.
The study involved a playful and somewhat unstructured puppet game with
minimal pairs, where we tried to induce the child to interact with an adult.
The puppets were used as tools in the interaction of both adult and child. The
training was in contrast to most of his other school assignments dialogic and
presupposed interaction. Several aspects were taken into consideration as
the study was performed. In the training, Ted had the possibility of match-
ing specific objects (which were stimulating for the senses) with sound pat-
terns in the environment. By the presence of objects, which Ted and the adult
could refer to, understanding was facilitated. The training can be described as
a combination of natural language treatment and adult-directed instructions,
where all speech attempts were reinforced. The teachers and the speech and
language pathologist had an educational and therapeutic quite calm approach
where they had an on-going positive attitude towards the boy. They gave the
boy general praise throughout the sessions, but used some additional positive
signals (change in voice, words) now and then when Ted tried to produce
sounds or words. During the second period, when the SLP was involved, Ted
took more initiatives, which the SLP could follow. An example is when Ted
showed the dolls (one at a time) to the video camera.
We thought it would be important for the intervention to be performed
with Ted’s ordinary teachers so Ted did not need to connect with a new person
(e.g. the researcher). A speech pathologist was later involved, but this person
was part of the staff at school. The study was performed over one year and
took place in two periods, each five months long. After that follow-up mea-
sures were performed. We called the weekly video recordings made during
the weeks without treatment baseline sessions (BL), and those made during
periods of treatment probe sessions. Ted’s summer holiday from June to August
interrupted the intervention study, which led to a withdrawal of treatment.
A second baseline, with video recordings without weekly training was estab-
lished at the beginning of August. Establishing a baseline (with data of the
child’s speech before treatment) was, however, difficult, because of the boy’s
very limited speech output. Pictures of the puppets in the study were used
as stimuli during the baseline and probe sessions. As there was virtually no
sound production during the first baseline, we made the decision to use the
puppets in addition to pictures during the remaining baseline sessions and the
weekly probe sessions.

Stimuli – materials
The materials used in the study (Figure 1) were puppets, made of differ-
ent colors and materials, such as wood or soft stuffed fabrics, that would
 Pia M. Nordgren 33

be stimulating for different modalities like vision and audition (IAKM, The
International Association for the Karlstad Model, Johansson, 1988). Differ-
ent names (CVCV syllables), for the puppets were decided on beforehand.
All the puppets’ names exhibited the word accent of Swedish (accent 2). In
period 1, eight puppets (+ pictures) were used during baseline and probe and
in period 2, 13 puppets (+ pictures). Every fourth week, if Ted participated, we
also showed the puppets from period 1.

Figure 1. Two examples of puppets, [kʊkʊ] (left) and [bʊbʊ] (right) are shown.

Period 1 (Age: 5;9 – 6;1 years, February to June) – stimuli

Period 1, (age: 5;9 to 6;1 years), had three baseline sessions (in February)
and ended with only two baseline sessions (due to the summer holiday,
June to August). This period contained 13 probe sessions (Table 3). During
period 1, daily puppet play, (for 5–20 minutes or as long as Ted partici-
pated, see stimuli), was performed once a day, i.e. in mean eight minutes’
training per day, by Ted’s teachers. The researcher was only present during
the weekly sessions, the probe sessions, where video recordings were made.
During period 1, the weekly probe sessions were video recorded by the
researcher.
Table 3 shows the trained features during period 1 in chronological order.
Stimuli were gradually introduced using two stimuli per week (two puppets per
week), forming minimal pairs. The order of the sounds followed the general
order of language acquisition, according to Jakobson (1968). Four distinctive
feature contrasts were systematically introduced, place features (− voice) [p],
[t], [k], place features (+ voice) [b], [d], [ɡ], laryngeal features [p]–[b], [t]–[d],
[k]–[ɡ], and manner features, [m], [p], [b], [n], [t], [d]. The vowel [ʊ] was used
consistently, as well as the CVCV syllable structure. The names of the puppets
were: [pʊpʊ], [tʊtʊ], [kʊkʊ], [bʊbʊ], [dʊdʊ], [gʊgʊ], [mʊmʊ] and [nʊnʊ].
Eight puppets (+pictures) in total were included period 1.
34 Phonological Development

Table 3. Description of features for the 13 probe sessions and the representation of
baseline sessions, Period 1 are shown

Order Sound (in Phonetic Feature Overall Description:

Swedish) Representation (in Contrast Feature Examples of words
Swedish) Contrast (in Swedish)
1 Baseline
2 Baseline
3 Baseline
4 p–t [pʊpʊ] [tʊtʊ] labial - dental Place contrast ”Poppo” ”Totto”
5 p–k [pʊpʊ] [kʊkʊ] labial - velar Place contrast ”Poppo” Kocko”
6 t–k [tʊtʊ] [kʊkʊ] dental -velar Place contrast ”Totto” ”Kocko”
7 b–d [bʊbʊ] [dʊdʊ] labial - dental Place contrast ”Bobbo” ”Doddo”
8 b–g [bʊbʊ] [ɡʊɡʊ] labial - velar Place contrast ”Bobbo” ”Goggo”
9 d–g [dʊdʊ] [ɡʊɡʊ] dental - velar Place contrast ”Doddo” Goggo”
10 p–b [pʊpʊ] [bʊbʊ] +/- voice Laryngeal ”Poppo” Bobbo”
(labial) contrast
11 t–d [tʊtʊ] [dʊdʊ] +/- voice Laryngeal ”Totto” ”Doddo”
(dental) contrast
12 k–g [kʊkʊ] [ɡʊɡʊ] +/- voice Laryngeal ”Kocko” ”Goggo”
(velar) contrast
13 m–p [mʊmʊ] [pʊpʊ] nasal - stop Manner ”Mommo” Poppo”
contrast
14 m–b [mʊmʊ] [bʊbʊ] nasal - stop Manner ”Mommo” Bobbo”
contrast
15 n–t [nʊnʊ] [tʊtʊ] nasal - stop Manner ”Nonno” ”Totto”
contrast
16 n–d [nʊnʊ] [dʊdʊ] nasal - stop Manner ”Nonno” ”Doddo”
contrast
17 Baseline
18 Baseline

Period 2 (Age: 6;3 to 6;7 years, August to December) – stimuli

Period 2 (age: 6;3 to 6;7 years) began with a baseline period of four weeks.
After the period of training there was another baseline period of three weeks.
Period 2 contained nine probe sessions and the sound features were changed
three times (Table 4). During period 2, the training stimuli occurred over
three weeks and included the use of four to five puppets per week. At first, the
stimuli had an initial and medial bilabial consonant, [mɪmɪ], [pɪpɪ], [bɪbɪ],
[baba], and then an initial and medial dental consonant, [nɪnɪ], [tɪtɪ], [dɪdɪ],
[tata], [dada]. Finally, an initial and medial velar consonant were used, [kɪkɪ],
[ɡɪɡɪ], [kaka], [ɡaɡa]. In period 2 the vowels were varied during training
(between [i] and [a]). In total, 13 puppets (+ pictures) were included.
The school decided that Ted should receive speech and language training in
period 2, and therefore a licensed speech and language pathologist, who was
 Pia M. Nordgren 35

part of the staff at school, was responsible for the weekly sessions in period 2,
while teachers continued the daily training. There were no other interventions
by the speech pathologist during this period of time. Table 4 shows the trained
features during period 2 in chronological order.

Table 4. The sound features for the nine probe sessions and the representation of base-
line sessions period 2 are shown.

Order Sounds Phonetic Feature Overall Description:

Representation (in Contrasts Feature Example of Words
Swedish) Contrasts (in Swedish)
1-4 Baseline
5-7 m-p-b [mɪmɪ] [pɪpɪ] voice – laryngeal ”Mimmi” ”Pippi”
[bɪbɪ] [baba] voiceless manner ”Bibbi” ”Babba”
stop - nasal
8-10 n-t-d [nɪnɪ] [tɪtɪ] [dɪdɪ] voice - laryngeal ”Ninni” ”Titti” ”Diddi”
[tata] [dada] voiceless manner ”Tatta” ”Dadda”
stop - nasal
11-13 k-ɡ [kɪkɪ] [ɡɪɡɪ] voice - laryngeal ”Kicki” ”Giggi”
[kaka] [ɡaɡa] voiceless ”Kacka” ”Gagga”
14-16 Baseline

The researcher video recorded the two first baseline sessions, period 2, until
the speech and language pathologist began weekly treatment sessions with
Ted and took over the video filming. The researcher met with the speech and
language pathologist on a regular basis. In consideration of Ted’s development
and to keep up Ted’s interest, we decided to use several puppets during period
2, and train the features for longer periods.

Follow-up measures (Age: 6;8 to 7;2 years, January to June)

Follow-up was conducted one month until four months after completion of
the study. The training continued once a week with the SLP and follow-up
measures were done. The results during the treatment period were compared
to the baselines, historical data, the speech pathologist’s medical records, and
interviews (see below).

Recordings and transcriptions

All the sessions, 34 occasions in total, were video recorded with a Panasonic
HDC SD-700, once a week during the year. The length of the recordings varied
between five to 20 minutes with consideration taken for the child´s participa-
tion and attention span. In mean, the sessions were 8,7 minutes long with a
standard deviation of 5,5 minutes. The sessions in period 2 had a somewhat
longer duration than the sessions during period 1, because Ted participated
for longer periods when the SLP was involved. It was possible for the SLP to
36 Phonological Development

use more interactive games than were used during period 1. The mean dura-
tion per session period 1 was 5,5 minutes with a standard deviation of 2,65
minutes, while period 2 had a mean of 12,3 minutes and a standard deviation
of 5,7 minutes. Ted’s total talk time (which included vocalizations) had a mean
of 35 seconds per session, with a mean standard deviation of 29 seconds. Ted’s
talk time (including vocalizations) increased from a mean of 23 seconds per
session during period 1 (standard deviation = 17 seconds) to a mean of 48
seconds (standard deviation = 35 seconds) per session period 2. The corpus
was transcribed and annotated. Orthographic transcriptions were made for the
teacher’s or the speech and language pathologist’s speech and phonetic tran-
scription (according to IPA, the International Phonetic Alphabet) was made
for the child´s speech. An utterance here refers to at least one syllable with
at least one consonant and one vowel, since non-speech vocalizations were
removed from the corpus data. The setting was a special room at Ted’s school,
constructed for teaching. Only Ted and his teacher or SLP were present in the
room (apart from the researcher in period 1). Four to five children with ASC
had their daily education in the classroom (together with six to seven teach-
ers) outside this room. Recordings began when Ted and the teacher or speech
and language pathologist entered the room.

Method of analysis – measurements

The method of analysis was both quantitative, concerning, for example, the
number of speech productions or feature types and qualitative, concerning,
for example, syllable constructions. The transcribed material was analyzed
and measured for general sound production, phonological features, and syl-
lables. Not all of the vocalizations were transcribed since some of them were
considered to be non-speech vocalizations (see previous section). T-test was
used in comparing period 1 and 2, where df was represented by number of
sessions during period 2, 16−1= 15.
Feature accuracy analysis was performed, using a method similar to that
used by Koegel et al. (1988) for calculating whether phonological features
were correct. Feature accuracy was calculated by giving scores from 0 to 5
on every answer to a specific prompt, 5 = maximum score and 0 = minimum
score. Scores were estimated for the following: place features, manner features,
laryngeal features, syllable features, and correct vowel. If all features were
correct the answer was scored 5. Feature scores were converted to a percent-
age. T-test was used for comparing period 1 and 2.
In addition, auditory perceptual skills were annotated as responses to spe-
cific prompts. Pointing or gestures such as grabbing the object were judged as
correct responses. Eye pointing was rare and not counted in the study. In Table
5, the prompts are described.
 Pia M. Nordgren 37

Table 5. Examples of the prompts are shown

Number Example
1 Two to five different puppets are shown and named by teachers or speech and
language pathologist so the boy can listen to the sound-combination repeatedly
and is given the possibility to imitate.
2 Phrases like: ‘Now comes …’
3 Phrases like: ‘Where is [tata]?’ ‘Point at [tata]!’

Inter/intra-observer agreement
Two individuals (T1 and T2) transcribed the corpus and an inter-observer
agreement between the T1 (Transcriber 1) and T2 (Transcriber 2) during all
the sessions was then calculated. Only sounds annotated by the two transcrib-
ers were used in the calculations. This means that if one transcriber had not
annotated an utterance, it was removed from the calculations. Of the utter-
ances transcribed, 70% were considered utterances by both transcribers.
For speech sounds, the inter-observer agreement was 81%. Specific agree-
ment for the number of syllables was 59%. The lower agreement for syllables
may be related to difficulties with estimating word boundaries. Some words
were produced which included long pauses between the syllables and it was
difficult to estimate whether they belonged to the same word or not.
It was not possible to reach 100% agreement, which is why one of the tran-
scribers tried to reach consensus with a third transcriber so as to have as much
valid data as possible. The third transcriber (phonetician) had consensus dis-
cussions with transcriber 1 (T1) and phonetic transcriptions according to
IPA were made after thorough discussions. After the estimation of consensus,
the inter-observer agreement was almost 100%. There was high agreement
as regards syllables, vowels, and the distinction stop-nasal, but the feature
voiced-voiceless was challenging to estimate.
Cohen’s kappa was also calculated for T1 and T2. Cohen’s kappa (Table 6)
for sounds was Κ = 0.77, which was quite good (accuracy 81%, 0.81 – 0.22/1 –
0.22 = 0.77). Cohen’s kappa (Table 6) for syllables was Κ = 0.44 (accuracy 59%,
0.59 – 0.26/ 1 – 0.26 = 0.44) which was perhaps a bit low, but acceptable.

Table 6. Cohen’s kappa and accuracy for sounds and syllables is shown

Calculated Feature Accuracy Cohen’s kappa

Sounds 81% K = 0.77
Syllables 59% K = 0.44

Interviews
The researcher interviewed one of Ted’s parents twice during the study and
took notes from the meetings. During the second interview, the speech and
38 Phonological Development

language pathologist was present. One and a half years after the beginning of
the study, the parents participated in a written interview. Notes from discus-
sions with teachers were collected on a regular basis during the researcher’s
visits to the school. Data concerning the speech pathologist’s experience rely
on the speech and language pathologist’s medical records of Ted.

Results
General sound production
The development of sound production in general is presented in Figure 2.
Increased sound production shows as a positive trend during the year, com-
pared to the baseline sessions. There were 112 speech productions (31%)
during period 1 and 254 speech productions (69%) during period 2, for a total
corpus of 366 utterances. Using a t-test, the difference in number of utterances
between period 1 and 2 was significant, t (15) = −4,9, p < 0.0001. A complicat-
ing factor was that a change in ADHD medication co-occurred with introduc-
tion of treatment, week 6–10, which possibly affected the work in general, in
that Ted was tired and less eager to participate during this period. This may
have led to no or very few utterances during this period. The graphs in Figures
3 to 4 show that the initial sessions contain dental sounds, since it took a while
until labials and velars were produced. Session 16 shows a decreased number
of productions, which did not occur in conjunction with change in ADHD
medication.

Figure 2. The general development of speech sounds during the study is shown. The
vertical dashed line shows the end of period 1. P1 = Period 1. P2 = Period 2. The vertical
grey lines show the divisions between baseline and treatment phases. T1 = Transcriber
1. T2 = Transcriber 2. Consensus = both Transcribers.
 Pia M. Nordgren 39

Phonologically distinctive features

The specific phonological features increased during the year, as was the case
for general sound production, seen in Figures 3 to 5. Figure 3 shows that labial
sounds were produced more frequently as the contrasts [m]-[p], [m]-[b]
were introduced (session 13–14). Words for [mama] and [papa], mummy
and daddy occurred in session 13, when labials (such as [mʊmʊ], [pʊpʊ],
and [bʊbʊ]) were introduced. Learning labials at this stage (Ted’s age) was
quite late, according to the literature on child language, but nevertheless this
is a possible effect of training. These new words temporally coincided with
treatment including labials. A noteworthy fact is that when m–p was trained,
place features were held constant and it was manner features that were trained.
The labial effect remained during the second baseline (period 1). The positive
increase of labials continued during period 2. The difference between the two
periods, using the t-test, was, however, not significant, t (15) = 0.540, p < 0.597
in production of labials.

Figure 3. The production of labial utterances is shown. The vertical dashed line shows
the end of period 1. P1 = Period 1. P2 = Period 2. The vertical grey lines show the division
between baseline and treatment phases. T1 = Transcriber 1. T2 = Transcriber 2. Consen-
sus = both Transcribers.

Dentals were the sounds that dominated overall in Ted’s production and
increased more generally throughout the study, which is seen in Figure 4.
It may have been the case that dentals followed the general productions of
sounds, which is seen in the beginning of dentals directly after baseline when
treatment has begun. The difference in production of dentals (between period
1 and period 2) was significant, t (15) = −3.507, p = < 0.003.
40 Phonological Development

Figure 4. The production of dental utterances is shown. The vertical dashed line shows
the end of period 2. P1 = Period 1. P2 = Period 2. The vertical grey lines show the division
between baseline and treatment phases. T1 = Transcriber 1. T2 = Transcriber 2. Consen-
sus = both Transcribers.

Velar sounds were almost completely absent in the production until the
introduction of velar puppets (Figure 5), when the number of velar sounds
produced increased. The results for velars (the difference between period 1
and period 2) were significant on the t-test, t (15), −2.389, p = < 0.030.

Figure 5. The production of velars is shown. The vertical dashed line shows the end of
period 1. P1 = Period 1. P2 = Period 2. The vertical grey lines show the division between
baseline and treatment phases. T1 = Transcriber 1. T2 = Transcriber 2. Consensus = both
Transcribers.

In summarizing the results shown on Figures 3 to 5, we may conclude that

labials increased by training (however, not significantly), dentals increased
continually (significantly) and velars increased (significantly) through training,
but only when a certain level of development was achieved (in this case period
 Pia M. Nordgren 41

2). General sound production followed about the same path as dentals, with a
continuous development.

Syllables
The target syllables in the study consisted of CVCV structure. As seen in
Figure 6, the production of CVCV syllables increased between sessions 15 to
18. The difference was statistically significant, t (15) = −3.8, p = <0.002.

Figure 6. The number of CVCV syllables in production during the different sessions is
shown. The vertical dashed line shows the end of period 1. P1 = Period 1. P2 = Period 2.
The vertical grey lines show the division between baseline and treatment phases. T1 =
Transcriber 1. T2 = Transcriber 2. Consensus = both Transcribers.

Figure 7 shows that one-syllable utterances also increased during the study.
The difference between period 1 and 2 was significant on a t-test, t (15), =
–3.5, p = < 0.003.

Figure 7. The number of CV/VC syllables in production during the different sessions is
shown. The vertical dashed line shows the end of period 1. P1 = Period 1. P2 = Period 2.
The vertical grey lines show the division between baseline and treatment phases. T1 =
Transcriber 1. T2 = Transcriber 2. Consensus = both Transcribers.
42 Phonological Development

In summary, Figures 6–7 show an increase in CVCV constructions as well

as in CV and VC constructions. These differences were significant and are
remarkable because of very limited production period 1.

Atypical production of speech

Atypical syllabic constructions for language development that were present in
the corpus were, for example, VV combinations, such as [ʏɯ:] and CVV, such
as [nʏʊ] [nʏœ:]. Other aspects in the corpus that differed from that of typi-
cally developing children were: palatal stops, increased length of aspiration on
syllables, lateral fricatives, interlabial stops, nasalized vowels and diphthongs.
These sounds are found neither in Ted’s L1 (a non-Germanic language) or
L2 (Swedish). In addition, some words were produced with more than the
expected number of syllables, due to Ted not yet having acquired knowledge
of syllable structure and the property of a word. Examples were [ma ma ma],
[pa pa pa], [kʊ pa tʊ].

Feature accuracy
Scores were estimated for place features, manner features, laryngeal features,
syllable features, and correct vowels, and were compared to the number of
possible scores for each prompt. For example, if the five features were correct,
the score was 5, but if just one feature was correct, the score was 1. Prompts
were counted if the adult said the prompt word or said: ‘give me’ + prompt
word, ‘look’ + prompt word, ‘where is’ + prompt word, ‘there was’ + prompt
word, or ‘here comes’ + prompt word.
Period 1 showed 121 feature scores out of 535 possible scores (23%
correct) and period 2 showed 328 feature scores out of 756 possible scores
(43% correct), which means that production during period 2 showed better
feature accuracy than production during period 1. The total number of pos-
sible scores in relation to the number of prompts for the whole period was
1,291 (535+756) possible scores. The 449 (121+328) feature scores out of 1,291
possible scores are about 35% correctly produced utterances. In total, 27%
of the features scores were found in period 1 and 73% were found period 2,
which is in large part, related to a general increase in speech production. In
analysing feature scores for every session in relation to the number of possible
scores, a positive trend is shown in Figure 8, which shows feature scores as a
percentage for every session. The graph suggests that Ted’s awareness of the
different features increased. On a t-test, the difference between period 1 and 2
in a percentage was significant, t (15) = −5.3, [p < 0.000.
 Pia M. Nordgren 43

Figure 8. Feature points in percentage for every session during the year. P1 = Period
1. P2 = Period 2. The vertical dashed line shows the end of period 1. The vertical grey
lines show the division between baseline and treatment phases. T1 = Transcriber 1. T2 =
Transcriber 2. Consensus = both Transcribers.

In summary, as regards feature accuracy, there was a positive trend as

shown in Figure 8 in the correctness of production.

Auditory perceptual skills/language comprehension

The number of responses for auditory perception (answer to a prompt (such
as where is + name?) during period 1 was near zero. During period 2, 17 out
of 47 responses were correct, which means that the difference (an increase
from 0% to 36% correct responses) between period 1 and 2 was statistically
significant, t (15) − 4, p = < 0.001.

Results from tests and interviews

The SLP, 1 month and 4 months after the study, performed the tests below.
During this period the training continued weekly with the speech pathologist.

The Peabody Picture Vocabulary Test IV (1 month after data collection)

The Peabody Picture Vocabulary Test IV (PPVT, Dunn and Dunn, 1959) was
tested one month after completion of the study (after data collection). Ted
exhibited a raw score of 20 on the PPVT, which is the equivalent for 2;1 years
of age, with a standard score of 61.

Hellquist Phoneme Test (1 month after data collection)

The Swedish Hellquist’s Phoneme Test (Hellquist, 1995) investigates the pro-
nunciation of speech sounds and children’s phonological system. The child is
44 Phonological Development

induced to produce the words, which correspond with black and white pic-
tures in a picture book.
Ted was tested with Hellquist’s Phoneme Test one month after comple-
tion of the study. On the Hellquist’s Phoneme Test, Ted named 21 out of 42
pictures. From the videotapes it was noted that most pictures were named
spontaneously, but some were repeated. During testing, bilabials, dentals, and
nasals and one fricative [f] were produced.

Story reading (1 month after data collection)

Ted was tested with story reading one month after data collection. Ted par-
ticipated in naming 11 out of 13 prompted words during story reading. Two
CVCV constructions were used in a story reading test one month after comple-
tion of the study. Ted’s productive vocabulary was estimated to be more than
25 words, compared to five words, as estimated by the speech and language
pathologist at the neuropsychiatric center before the study and 5–10 words in
notes from parents’ and teachers’ interviews before the study.

Hellquist Phoneme Test (4 months after data collection)

After 4 months, post intervention, Ted was tested again with Hellquist’s
Phoneme Test. Ted was able to give a response for 35 out of 45 pictures on the
test, in comparison with test 1, where he names 21 out of 42 pictures of words
with a syllable or sound. All stops (apart from [ɡ]) were produced. Nasals
[n] and [ŋ] were produced, as were fricatives [f], [s] and the approximate [j].
Words with one, two and three syllables were produced. Examples of words
produced were: [fɔfa] for soffa, ‘sofa’, [fi:afa] for giraffe, ‘giraffe’, [tɛtɛ] for plåster,
‘sticking plaster’, [ɔta] for åtta, ‘eight’ and [tʏfɛfɛ] for tjugofem , ‘twenty-five’.
Most of the ‘error types’ seen in Ted´s speech resemble speech in verbal
dyspraxia, with inconsistent pronunciation of words, simplification of word
structure through omission of syllables, and certain syllable reduplications,
in addition to a visible ‘seeking behavior’ in articulation. Table 7 shows an
overview of the test results.

Table 7. An overview of test results after the training period

PPVT PPVT Hellquist Phoneme test Story Reading

Raw scores Standard scores
1m 20 61 21/42 11/13 words
4m 35/45

Interviews with parents and teachers

Ted’s parents confirmed his development in language and imitative skills
during period 1 (Interview 1). New skills such as matching pictures with words
 Pia M. Nordgren 45

had also occurred. He used prosody to express emotions. During period 2

(Interview 2), his parents reported that Ted became more communicative,
attentive, and social, and began to approach his parents, saying [ma:] or [pa:],
and expecting a response. Pointing increased.
In the third (written) interview, the parents revealed that communica-
tion (and Ted´s understanding at home) had become more effective, includ-
ing more communicative initiatives. Eye contact had increased and Ted had
become ‘more present’ and social, and had begun to show an understanding
of intentions (for example, parents wanting him to attend to them). Sound-
ing had increased and signs of an interest in others’ lip movements occurred.
Overextension (‘mummy’ for all women and ‘daddy’ for all men) had begun
developing. Ted had become interested in books and realized the connection
between letters, words, and pictures, according to the parents. Ted’s vocabu-
lary now consisted of more than 60 words. Below follows an example from the
written interview:
Researcher ‘What progress have you noted’?

Mother ‘He understands more, he produces more, he looks more firmly

at our mouths, he imitates.’ ‘He started imitating at 6 years of
age.’ ‘Our son now understands that fun things may happen if
he stays and listens to what we say.’

‘He understands that there is a connection between us calling

for him and him being able to get some information.’

Researcher ‘Have the number of communicative initiatives increased’?

Mother ‘Yes, without doubts. He is more present, more communicative

and more social. He looks directly at my eyes if I talk or if he
talks to me.’

The teachers confirmed that Ted´s expressive vocabulary, pointing, and

interest in reading books had developed during the year.

Discussion
The results of the study showed that listening to sounds in CVCV construc-
tions with minimal pairs in interaction seemed to have had an impact on oral
production in this child with ASC. A comparison of the two periods showed
both a quantitative and a qualitative development of general speech produc-
tion during the year.
The quantitative increase consisted of a significant increase in production
of speech sounds but also in use of new features. It is noteworthy that velars
increased during Period 2, when a certain level of development had taken
46 Phonological Development

place. In period 1, 31% of the total numbers of produced sounds were found,
compared to 69% during Period 2. The difference in speech production during
the two periods was, for the most part, statistically significant. Furthermore,
the number of responses concerning auditory perceptual skills and correct
phonological features increased.
The qualitative difference during the two periods of training consisted of
use of new phonological distinctive features, such as labials, velars, and new
syllable types (CVCV syllables). Feature accuracy increased. Historical data in
medical records and historical data from interviews with parents and teachers
revealed that Ted rarely exhibited oral production with more than one syllable
before the study. In addition, historical data revealed that speech production
increased during the study. Thus, historical data was in line with the results of
the study.
This study detected a deviant development of syllable constructions, that
is the presence of CVV constructions and VV constructions. These non-
speech vocalizations are atypical, compared to typical speech development
in the literature. The presence of phonological processes and deviant sounds
occurring in this child´s speech is in accordance with the literature regarding
phonology and vocalizations in autism (Wolk and Edwards, 1993; Wolk and
Giesen, 2000; Schoen et al., 2011). Examples of atypical sounds in the present
study were: palatal stops, an increased length of aspiration in syllables, lateral
fricatives, interlabial stops, nasalized vowels, cluster reduction, and diphthongs.
None of these sounds are found in L1 or L2. They may instead be related to
different developmental babbling stages. Some of Ted´s vocalizations may, for
example, be related to the ‘vocalic stage’. The language difficulties in this case
may be related to autism per se and delayed language development, but they
may also be due to intellectual disability. There were examples in the corpus,
which suggested that some language difficulties were related to problems with
the identification of word boundaries. For example, too many syllables per
word were present, such as [ma ma ma]. Pauses also existed in inappropriate
places.
All speech attempts were reinforced with a positive response from teachers
and SLP during the study, which was in accordance with the method of earlier
studies (Koegel et al., 1988, 1998). Koegel et al. (1998) described the impor-
tance of motivation in speech training in autism. An additional conclusion
from this study may be that use of puppets with varying colors and textures
may be motivating for children with autism. In the study, place (of articula-
tion) features were kept the same and involved less effort than changing these
features.
This case had many complicating factors, with the use of several languages
being one of them. Use of several languages may influence language develop-
 Pia M. Nordgren 47

ment and in this case parents were advised to just speak one language with
their child. There is however reason to believe that the use of several languages
would not lead to this severe language delay. A recent study by Valicenti-
McDermott et al. (2012) showed no differences between a monolingual and
a bilingual group of children with ASC regarding language development. Ted
in the present study also exhibited additional developmental delay, i.e. intel-
lectual disabilities, in addition to autism, which is quite common. There is
also the matter of general development, which may interfere with the results
of a study. One may argue that we were ‘lucky’ and that this boy was on his
way to developing language and speech when the study took place, also due
to the school’s educational input. Even if general development or maturation
may not be completely ruled out as making a contribution to the results of the
present study, the rapid progression seen in this boy and his history of being
non-speaking until 5:9 years, leads us to conclude that the results have been
at least partly due to the intervention. Labials and dentals increased by train-
ing and velars increased when a certain level of development is achieved (in
period 2). In addition, we know that this boy exhibited only a few syllables,
before the study and no CVCV constructions. An additional complicating
factor in the study is the change in ADHD medication, which co-occurred with
the introduction of treatment (weeks 6–10); it affected Ted’s performance in
general negatively. It is possible that correct medication could affect his atten-
tion span, but it is not likely that medication per se would lead to specific
improvement in specific phonological features or syllable constructions. The
qualitative differences with an increase of certain syllables suggest that Ted
improved his speech and language competence because of the intervention,
rather than the medication alone. The last complicating factor concerns the
speech and language pathologist being responsible for training during Period
2; the therapeutic approach per se may have been responsible for some of the
improvement.
It could be argued that the increase of speech in this study is very limited
and therefore not important. On the other hand, it may be of great impor-
tance for this child’s quality of life to be able to hear his own voice and play
with speech sounds. Diehl et al. (2012) proposed that a child needs using
non-communicative utterances or ‘read aloud’, in order to train motor
planning and perceptual skills. To be aware of the difference between the
sound features leads to an increased vocabulary and a possible effect on
quality of life. In the present study, Ted managed to begin using speech
sounds for communicative purposes (he said [ma] or [pa] when he needed
his parents’ attention). An important aspect is the minimum of training
given in this study, which suggests that a minimum of training can have
important implications.
48 Phonological Development

The results here are in accordance with Wolk and Edwards (1993), suggest-
ing that auditory perception may influence production. Beginning to be able
to identify word boundaries may have been contributing to the results both
as regards improved perceptual skills and also increased speech production.
Other factors contributing to the result may also have been increased working
memory or the inhibition of non-important environmental stimuli. Working
memory is in this study was trained by using a specific number of CVCV
constructions.
The most important contributions made by the present study are related
to the following: 1. This is a study of a child with (from the start) no speech
output. 2. It is a study of a school child. 3. An intervention which was limited in
time and easy to carry out (for both child and adult) seems to lead to acquisi-
tion of phonological contrasts, speech production, syllable constructions and
use of phonological features.

Future work
The findings in this study provide implications for further research about
school children with ASC and limited speech output and their acquisition of
phonological contrasts.

Acknowledgements
Sincere thanks to the family, the teachers and the speech and language pathol-
ogist, involved in the study, for your participation. I would like to extend my
sincere gratitude towards associate professor Åsa Abelin and professor Sally
Boyd at the Department of Philosophy, Linguistics and Theory of Science
for on-going support and commenting on the manuscript. Thanks also to Dr
Simon Dobnik at the Department of Philosophy, Linguistics and Theory of
Science for comments, which contributed to this work.
The Region Västra Götaland, Sweden, University of Gothenburg, Sweden,
Kempe-Carlgrenska Foundation, Sweden, and the Royal Hvitfeldtska Founda-
tion supported this work.

About the author

Pia Nordgren is active in Linguistics at the Department of Philosophy, Linguis-
tics and Theory of Science, University of Gothenburg, Sweden. Correspondence
concerning this article should be addressed to: Pia Nordgren, Department of Phi-
losophy, Linguistics and Theory of Science, University of Gothenburg, Box 200,
405 30 GOTHENBURG or Phone: +46 (31) 786 43 92, E - mail: pia.nordgren@
ling.gu.se.
 Pia M. Nordgren 49

References
Baron-Cohen, S., Leslie, A. M. and Frith, U. (1985), Does the autistic child have a ‘Theory of
Mind’? Cognition, 21 (1): 37−46. http://dx.doi.org/10.1016/0010-0277(85)90022-8
Bartolucci, G., Pierce, S., Streiner, D. and Eppel, P. T. (1976), Phonological investigation of
verbal autistic and mentally retarded subjects. Journal of Autism and Childhood Schizo-
phrenia, 6 (4): 303–316. http://dx.doi.org/10.1007/BF01537908
Bartolucci, G. and Pierce, S. (1977), A preliminary comparison of phonological develop-
ment in autistic, normal and mentally retarded subjects. International Journal of Lan-
guage & Communication Disorders, 12 (2): 137–147. http://dx.doi.org/10.3109/1368282​
7709011317
Bishop, D., Wong, M. M. D., Maley, A., Hill, W. and Hallmayer, J. (2004), Are phonologi-
cal processing deficits part of the broad autism phenotype? American Journal of Medical
Genetics Part B (Neuropsychiatric Genetics), 128 B: 54–60.
Bondy, A. and Frost, L. (2002), A Picture’s Worth: PECS and Other Visual Strategies in
Autism. Topics in Autism. ERIC, Institute of Education Sciences.
Cirrin, F. M. and Gilliam, R. B. (2008), Language intervention practices for school-age chil-
dren with spoken language disorders – a systematic review. Language, Speech and Hear-
ing Services in Schools, 39: 110–137. http://dx.doi.org/10.1044/0161-1461(2008/012)
Cleland, J., Gibbon, F. E., Peppe´, Sue J. E., O´Hare, A. and Rutherford, M. (2010), Phonetic
and phonological errors in children with high functioning autism and Asperger syn-
drome. International Journal of Speech-Language Pathology, 12 (1): 69–76. http://dx.doi.
org/10.3109/17549500903469980
Diehl, J. J. and Paul, R. (2012), Acoustic differences in the imitation of prosodic patterns in
children with prosodic patterns. Research in Autism Spectrum Disorders, 6 (1): 123–134.
http://dx.doi.org/10.1016/j.rasd.2011.03.012
Dziuk, M. A., Gidley Larson, J. C., Apostu, A., Mahone, E. M., Denckla, M. B. and Mostof-
sky, S. H. (2007), Dyspraxia in autism: Association with motor, social, and communica-
tive deficits. Developmental Medicine & Child Neurology, 49 (10): 734–739. http://dx.doi.
org/10.1111/j.1469-8749.2007.00734.x
Dunn, D. M. and Dunn, L. M. (1959), Peabody Picture Vocabulary Test – III, San Antonio,
TX: Pearson.
Engstrand, O. (1990), Swedish. Journal of the International Phonetic Association, 20-01:
42–44.
Elert, C. (1989), [General and Swedish Phonetics] Allmän och Svensk Fonetik, Norstedts
förlag AB, Stockholm.
Goldstein, H. (2002). Communication intervention for children with autism: A review
of treatment efficacy. Journal of Autism and Developmental Disorders, 32 (5): 373–396.
http://dx.doi.org/10.1023/A:1020589821992
Grela, B. G. and McLaughlin, K. S. (2006). Focused stimulation for a child with autism
spectrum disorder: A treatment study. Journal of Autism and Developmental Disorders
36 (6): 753–756. http://dx.doi.org/10.1007/s10803-006-0122-1
50 Phonological Development

Griffith, R. (1954), The Abilities of Babies, a Study in Mental Measurement, Worcester, MA:
American Psychological Association.
Hadwin, J., Baron-Cohen, S., Howlin, P. and Hill. K. (1997), Does teaching theory of mind
have an effect on the ability to develop conversation in children with autism? Journal of
Autism and Developmental Disorders, 27 (5): 519–537. http://dx.doi.org/10.1023/A:1025​
826009731
Hellquist, B. (1995), Phoneme Test (Fonemtest): Short version (Kortversion) (third edition),
Löddeköpinge, Pedagogical design, (Pedagogisk design), Sweden.
Jakobson, R. (1968). Child Language Aphasia and Phonological Universal. The Hague, The
Netherlands, Mouton & Co. http://dx.doi.org/10.1515/9783111353562
Johansson, I. (1988), [Language Development in Disabled Children 1] Språkutveckling hos
Handikappade Barn 1, Studentlitteratur, Lund, Sweden.
The Karlstad Model, IAKM, International Association for the Karlstadmodel.
Koegel, R. L., O´Dell, M. and Dunlap, G. (1988), Producing speech use in nonverbal autis-
tic children by reinforcing attempts. Journal of Autism and Developmental Disorders, 18
(4): 525–538. http://dx.doi.org/10.1007/BF02211871
Koegel, R. L., Camarata, S., Koegel, L. K., Ben-Tall, A. and Smith, A. E. (1998), Increasing
speech intelligibility in children with autism. Journal of Autism and Developmental Dis-
orders, 28 (3): 241–251. http://dx.doi.org/10.1023/A:1026073522897
Levy, S. E., Giarelli, E., Lee, L-C, Schieve, L. A., Kirby, R. S., Cunniff, C., Nicholas, J., Reaven,
J. and Rice, C. E. (2010), Autism spectrum disorder and co-occurring developmental, psy-
chiatric and medical conditions among children in multiple populations of the United
States. Journal of Developmental Behavioral Pediatrics, 31 (4): 267–275. http://dx.doi.
org/10.1097/DBP.0b013e3181d5d03b
Ohashi, J. K., Mirenda, P., Marinova-Todd, S., Hambly, C., Fombonne, E., Szatmari, P.,
Bryson, S., Roberts, W., Smith, I., Vaillancourt, T., Volden, J., Waddell, C., Zwaigen-
baum, L., Georgiades, S., Duku, E. and Thompson, A. (2012), Comparing early language
development in monolingual- and bilingual- exposed young children with autism spec-
trum disorders. Research in Autism Spectrum Disorders, 6 (2): 890–897. http://dx.doi.
org/10.1016/j.rasd.2011.12.002
Plunkett, K. and Strömqvist, S. (1990), The acquisition of Scandinavian languages, Gothen-
burg Papers in Theoretical Linguistics, 59.
Rapin, I., Dunn, M. A., Allen, D. A., Stevens, M. C. and Fein, D. (2009), Subtypes of lan-
guage disorders in school-age children with autism. Developmental Neuropsychology, 34
(1): 66–84. http://dx.doi.org/10.1080/87565640802564648
Reynell, C. (1977), Reynell Developmental Language Scales, Windsor, England.
Rogers, S. D., Hayden, D., Hepburn, S., Charlifue-Smith, R., Hall, T. and Hayes, A. (2006),
Teaching young nonverbal children with autism useful speech: a pilot study of the Denver
Model and PROMPT interventions. Journal of Autism and Developmental Disorders 36
(8): 1007–1024. http://dx.doi.org/10.1007/s10803-006-0142-x
 Pia M. Nordgren 51

Roug, L., Landberg, I and Lundberg, L. J. (1989), Phonetic development in early infancy: A
study of four Swedish children during the first eighteen months of life. Journal of Child
Language, 16 (1): 19–40. http://dx.doi.org/10.1017/S0305000900013416
Scheinkopf, S. J., Mundy, P., Oller, D. K. and Steffens, M. (2000), Vocal atypicalities in pre-
verbal autistic children. Journal of Autism and Developmental Disorders, 30 (4): 345–354.
http://dx.doi.org/10.1023/A:1005531501155
Schoen, E., Paul, R. and Chawarska, K. (2011), Phonology and vocal behavior in toddlers with
autism spectrum disorders, Autism Research, 4: 1–12. http://dx.doi.org/10.1002/aur.183
Schopler, E., Mesibov, G. B. and Hearsey, K. (1995), Structured teaching in the TEACCH
system, learning and cognition in autism. Current Issues in Autism: 243–268.
Sigurd, B. (1965), Phonotactic Structures in Swedish. Lund: Berlingska boktryckeriet.
Sparrow, S. S. and Cicchetti , D. V. (1985), Diagnostic uses of the vineland adaptive behav-
iour scales. Journal of Pediatric Psychology, 10 (2): 215–225. http://dx.doi.org/10.1093/
jpepsy/10.2.215
Steinberg, D. D. and Sciarini, N. V. (2006) An Introduction to Psycholinguistics (second edi-
tion). Harlow: Pearson Education.
Tager-Flusberg, H., Calkins, S., Nolin, T., Baumberger, T., Anderson, M. and Chadwick
Dias, A. (1990), A longitudinal study of language acquisition in autistic and Down Syn-
drome children. Journal of Autism and Developmental Disorders, 20 (1): 1–21. http://dx.​
doi.org/10.1007/BF02206853
Valicenti-McDermott, M., Tarshis, N., Schouls, M., Galdston, M, Hottinger, K., Seijo, R.,
Shulman, L. and Shinnar, S. (2013), Language differences between monolingual English
and bilingual English-Spanish young children with autism spectrum disorders. Journal
of Child Neurology, 28 (7): 945–948. http://dx.doi.org/10.1177/0883073812453204
Wolk, L. and Edwards, M. L. (1993), The emerging phonological system of an autis-
tic child. Journal of Communication Disorders, 26 (3): 161–177. http://dx.doi.org/10.​
1016/0021-9924(93)90006-V
Wolk, L. and Giesen, J., (2000), A phonological investigation of four siblings with childhood
autism. Journal of Communication Disorders, 33 (5): 371–389. http://dx.doi.org/10.1016/
S0021-9924(00)00021-6