Prosody, Phonology and Phonetics

Mari Wiklund

Speech and
Interaction
of Preadolescents
with Autism
Spectrum Disorder
Focus on Prosody, Disfluencies and
Comprehension Problems
Prosody, Phonology and Phonetics

Series Editors
Daniel J. Hirst, CNRS Laboratoire Parole et Langage, Aix-en-Provence, France
Hongwei Ding, School of Foreign Languages, Shanghai Jiao Tong University,
Shanghai, China
Qiuwu Ma, School of Foreign Languages, Tongji University, Shanghai, China
The series will publish studies in the general area of Speech Prosody with a particular
(but non-exclusive) focus on the importance of phonetics and phonology in this field.
The topic of speech prosody is today a far larger area of research than is often realised.
The number of papers on the topic presented at large international conferences such as
Interspeech and ICPhS is considerable and regularly increasing. The proposed book
series would be the natural place to publish extended versions of papers presented
at the Speech Prosody Conferences, in particular the papers presented in Special
Sessions at the conference. This could potentially involve the publication of 3 or
4 volumes every two years ensuring a stable future for the book series. If such
publications are produced fairly rapidly, they will in turn provide a strong incentive
for the organisation of other special sessions at future Speech Prosody conferences.
Mari Wiklund

Speech and Interaction

of Preadolescents
with Autism Spectrum
Disorder
Focus on Prosody, Disfluencies
and Comprehension Problems
Mari Wiklund
Department of Languages
University of Helsinki
Helsinki, Finland

ISSN 2197-8700 ISSN 2197-8719 (electronic)

Prosody, Phonology and Phonetics
ISBN 978-981-19-8116-6 ISBN 978-981-19-8117-3 (eBook)
https://doi.org/10.1007/978-981-19-8117-3

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Singapore Pte Ltd. 2023
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface

The idea of writing this book came from the observation that there was a need for a
scientific book that approaches autism spectrum disorder (ASD) from the perspective
of conversation analysis. Instead of experimental settings, studies carried out within
the framework of conversation analysis are based on naturally occurring data, i.e., on
spontaneous situations of interaction that occur independently of the study. I also felt
that there was a need for a book that addressed such aspects of interaction as speech
prosody, disfluencies of speech, comprehension problems, and gaze behaviour, as
well as therapists’ response strategies and teaching orientations, that have so far
been studied much less than, for example, the pragmatic problems of persons with
ASD.
The book is based on a number of studies carried out within the framework of
a project entitled “Interaction of Preadolescents with Autism—Focus on Speech
Prosody, Gaze Behavior and Misunderstanding Situations”, which was launched in
2009 as a postdoctoral research project funded by the Kone Foundation (Finland).
In 2014–2015, the project was funded by the Emil Aaltonen Foundation (Finland),
in 2015–2019 by the Helsinki Collegium for Advanced Studies of the University of
Helsinki (Finland), and in 2022 by the Future Fund of the University of Helsinki. I
am extremely grateful to the funders of the project as well as to all the colleagues
with whom I have had the privilege to work within the framework of this project.
I would like to thank the following colleagues in particular for their cooperation:
Minna Laakso, Simo Määttä, Satu Saalasti, Melisa Stevanovic, Lari Vainio, and
Martti Vainio.
I would also like to thank Springer for agreeing to publish this book, as well as
my family and friends for their support during the process.

Helsinki, Finland Mari Wiklund, Ph.D.

September 2022

v
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2.1 Finnish Data on Participants with ASD . . . . . . . . . . . . . . . . . . 5
1.2.2 French Data with Participants with ASD . . . . . . . . . . . . . . . . . 8
1.2.3 Control Group Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.2.4 Experimental Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.3 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2 Prosody . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1 Overview of Previous Research on the Prosody of Persons
with ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.2 Prosodic Characteristics of Preadolescent Boys with ASD
and Perception of Atypicality by Neurotypical Listeners . . . . . . . . . 22
2.2.1 Sing-Song-Like Pitch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
2.2.2 Bouncing Pitch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.2.3 Disconnected Speech Rhythm . . . . . . . . . . . . . . . . . . . . . . . . . 27
2.2.4 Large Pitch Excursions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
2.2.5 Flat Pitch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.3 The Role of Prosodic Features in the Creation
of Comprehension Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
2.4 The Use of Prosody in Questions and Statements . . . . . . . . . . . . . . . . 35
2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3 Disfluencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.1 Overview of Previous Research on Disfluencies in the Speech
of Persons with ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
3.2 Disfluencies in the Speech of Preadolescent Boys
with and Without ASD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

vii
viii Contents

3.3 The Role of Disfluencies in the Creation of Comprehension

Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
3.4 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
4 Comprehension Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4.1 Overview of Previous Research on Conversational Repairs
in the Interaction of Persons with ASD . . . . . . . . . . . . . . . . . . . . . . . . 75
4.2 Causes of Comprehension Problems in Group Therapy
Sessions Involving Preadolescent Boys with ASD . . . . . . . . . . . . . . . 77
4.2.1 Overly Literal Interpretation of Speech . . . . . . . . . . . . . . . . . . 78
4.2.2 Topical Discontinuities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
4.2.3 Non-verbal Features Associated with Trouble-Source
Turns: The Role of Eye Contact . . . . . . . . . . . . . . . . . . . . . . . . 86
4.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5 Gaze Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
5.2 Three Main Patterns Used for Avoiding Eye Contact in Group
Therapy Sessions Involving Preadolescent Boys with ASD . . . . . . . 97
5.2.1 Fixing One’s Gaze Straight Ahead . . . . . . . . . . . . . . . . . . . . . . 97
5.2.2 Letting One’s Gaze Wander Around . . . . . . . . . . . . . . . . . . . . 101
5.2.3 Looking at One’s Own Hands When Speaking . . . . . . . . . . . 107
5.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
6 Therapists’ Response Strategies and Teaching Orientations . . . . . . . . 117
6.1 Therapists’ Response Strategies in a Group Therapy Session
Involving Preadolescent Boys with ASD . . . . . . . . . . . . . . . . . . . . . . . 117
6.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
6.1.2 Five Most Common Response Strategies . . . . . . . . . . . . . . . . 121
6.2 Therapists’ Teaching Orientations in Group Therapy Sessions
Involving Preadolescent Boys with ASD . . . . . . . . . . . . . . . . . . . . . . . 132
6.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
6.2.2 Implicit Teaching Orientation . . . . . . . . . . . . . . . . . . . . . . . . . . 134
6.2.3 Explicit Teaching Orientation . . . . . . . . . . . . . . . . . . . . . . . . . . 139
6.2.4 Prosodic Problems: Beyond Teaching Orientations? . . . . . . . 142
6.3 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146

List of Transcription Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Chapter 1
Introduction

Abstract Autism Spectrum Disorder (ASD) is a neurobiological developmental

disorder characterised by deficits in social–emotional reciprocity and non-verbal
communication; restricted, repetitive patterns of behaviour, interests, or activities; as
well as deficits in developing, maintaining, and understanding relationships. In addi-
tion to these core features, atypical prosody was mentioned even in early descriptions
of autism. For example, people with ASD may have a limited pitch range, sing-song
pitch, bouncy pitch, overly fast speech rate, jerky speech rhythm, remarkably quiet
or loud voice, inconsistent pause structure, deviant word stress, and/or creaky or
nasal voice. It is also well known that persons with ASD often fail to use language
appropriately or effectively in interaction. Producing and understanding speech acts,
learning conversational skills such as turn-taking, as well as drawing upon contextual
information when interpreting the interlocutor’s speech can be difficult for persons
with ASD. The tendency of persons with ASD to understand things literally and to
miss implicit messages in interactions has been widely documented. In addition, ASD
is characterised by the avoidance of eye contact and other deviant features related
to gaze, facial expressions, postures, and gestures. Several studies have also demon-
strated that many speakers with ASD produce disfluent speech. On the other hand,
the discourse practices of children with ASD are also like those of their neurotypical
peers in many respects.

Keywords Autism · Autism spectrum disorder · ASD · Prosody · Intonation ·

Speech · Interaction

This chapter will provide a general introduction to the book. The chapter will start
with a brief description of the background (Sect. 1.1). Then a general description of
the data (Sect. 1.2) and methods (Sect. 1.3) will be given.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 1
M. Wiklund, Speech and Interaction of Preadolescents with Autism Spectrum Disorder,
Prosody, Phonology and Phonetics, https://doi.org/10.1007/978-981-19-8117-3_1
2 1 Introduction

1.1 Background

Autism Spectrum Disorder (ASD) is a neurobiological developmental disorder char-

acterised by deficits in social–emotional reciprocity and non-verbal communication;
restricted, repetitive patterns of behaviour, interests, or activities; as well as deficits
in developing, maintaining, and understanding relationships (APA, 2013: 50). The
prevalence of ASD is estimated to be 1.9% among children aged 8 years (Maenner
et al., 2020). It is approximately 4.3 times more frequent among boys than among
girls (Maenner et al., 2020).
In addition to the above-mentioned core features, atypical prosody was already
mentioned in early descriptions of autism (Asperger, 1944; Kanner, 1943). ‘Prosody’
encompasses phenomena related to pitch (voice fundamental frequency, f 0 ), loudness
(intensity), duration (timing), speech rhythm, speech rate, stress, phrasing, and voice
quality (Couper-Kuhlen, 2000; Crystal, 1980). Prosodic phenomena are ‘supraseg-
mental’: they typically involve units of speech that are at least one syllable in length.
It is known that people with ASD1 often have deviant prosodic features in their
speech. For example, people with ASD may have a limited pitch range, sing-song
pitch, bouncy pitch, overly fast speech rate, jerky speech rhythm, remarkably quiet or
loud voice, inconsistent pause structure, deviant word stress, and/or creaky or nasal
voice (e.g., Asghari et al., 2021; Asperger, 1944; Baltaxe, 1984; Baltaxe & Simmons,
1985, 1992; Diehl et al., 2009; Diehl & Paul, 2013; Fay & Schuler, 1980; Hubbard,
2017; Nadig & Shaw, 2012; Nakai et al., 2014; Paul, 1987; Paul et al., 2005a, 2005b;
Shriberg et al., 2001; Tager-Flusberg, 2000; Thorson et al., 2016; Wiklund & Vainio,
2019; Wiklund et al., forthcoming). When these features occur, they may constitute a
significant obstacle to the social acceptance of the individual (Paul et al., 2005a: 205).
Indeed, deviant prosodic features may create an immediate impression of ‘oddness’
(Van Bourgondien & Woods, 1992): atypical prosody has been identified as one of
the most recognisable ASD-specific behaviours (De Marchena & Miller, 2017). It
affects how ASD speakers are rated in terms of social and communicative competence
(Paul et al., 2005b) as well as in terms of likeability (Redford et al., 2018). Atypical
prosodic features may also impair the intelligibility of speech (Redford et al., 2018).
However, Holbrook and Israelsen (2020) study shows that interventions of more than
one treatment day that target specifically speech prosody result in moderate to large
improvements in speech prosody in persons with ASD.
Concerning the perception of prosody, it is known that the processing of emotional
prosody is impaired in autism (Globerson et al., 2014; Korpilahti et al., 2007;

1 The terminology that is used to refer to persons with ASD is currently an area of major debate
in the field of autism research (see, for example, Kenny et al., 2016; Vivanti 2020). According
to Kenny et al. (2016), academics and service providers favour person-first language (e.g., ‘person
with ASD’), whereas autistic persons themselves often prefer condition-first language (e.g., ‘autistic
person’). Vivanti (2020) recommends that academics should consult participants in each scientific
study about their actual preferences, and that the judicious use of person-first and condition-first
language should be supported as appropriate for the context. In this book, the author has decided to
use person-first language (e.g., ‘person/child/boy/preadolescent with ASD’), because she has not
had an opportunity to ask the participants about their own preferences.
1.1 Background 3

Rosenblau et al., 2017). Lindström’s (2019) results based on Finnish data show
that the processing of natural speech prosody is impaired in children with ASD
on various levels of information processing, including aberrant discrimination of
prosodic changes, as well as problems in orienting to such changes and sluggishness
in responding to them. Findings reported by Zhou et al. (2019) concerning Mandarin-
speaking preschool children with ASD indicate that the ability to use prosodic cues
to understand the speaker’s communicative intentions is impaired in autism. Cheval-
lier et al. (2009), in turn, studied the extent to which persons with ASD understood
grammatical prosody, and their results support the view that grammatical prosody is
not affected in ASD.
Pragmatic problems of people with ASD have been studied extensively (e.g.,
Cummings, 2009, 2014a, 2014b). It is, for example, well known that persons with
ASD often fail to use language appropriately or effectively in interaction (Cummings,
2009: 56). Producing and understanding speech acts and learning conversational
skills such as turn-taking, as well as drawing upon contextual information when inter-
preting the interlocutor’s speech, can be difficult for persons with ASD (Cummings,
2009: 56, 2014b: 49). The tendency of persons with ASD to understand things liter-
ally and to miss implicit messages in interactions has also been widely documented
(Cummings, 2009; Lehtinen, 2012; Lewis et al., 2008; Martin & McDonald, 2004;
Wiklund & Stevanovic, 2018; Wiklund, 2016). According to Cummings (2009: 57),
these difficulties are probably related to the fact that, in order to understand an utter-
ance which is used to imply something beyond what is explicitly stated in it, one
must be able to establish the ‘communicative intention’ of the interlocutor. This, in
turn, requires that the listener be able to make inferences about the mental states of
others (that is, to have a ‘theory’ of other ‘minds’), and this ability is known to be
impaired in autism.
ASD is also characterised by the avoidance of eye contact and other deviant
features related to gaze, facial expressions, postures, and gestures (McPartland &
Klin, 2006; Szatmari et al., 1989; Tantam et al., 1993; Wiklund, 2012). Indeed, when
ASD is suspected in a child, the lack of eye contact is typically one of the first
symptoms to draw adults’ attention (Hutt & Ounsted, 1966; Kylliäinen, 2007).
Several studies have also demonstrated that many speakers with ASD produce
disfluent speech (e.g., Lake et al., 2011; Scaler Scott et al., 2014; Scott et al., 2006;
Shriberg et al., 2001; Sisskin, 2006; Wiklund & Laakso, 2019, 2020). The definitions
of disfluencies differ, but the studies report that speakers with ASD have been found
to produce not only stuttering-like disfluencies such as sound, syllable, or word
repetitions, as well as blocks, but also atypical disfluencies such as breaks, insertions,
and word-final sound prolongations. In addition, syntactic impairments have been
found in the speech of persons with ASD (Cummings, 2014a). For example, children
with ASD may align their use of syntactic structures to that of the interlocutor (Allen
et al., 2011). Some studies also indicate that persons with ASD revise and self-repair
their speech (Shriberg et al., 2001; Sisskin, 2006; Wiklund & Laakso, 2019), but
contradictory findings have also been reported (Lake et al., 2011).
However, the discourse practices of children with ASD are also in many respects
similar to those of their neurotypical (i.e., typically-developing) peers (Ochs &
4 1 Introduction

Solomon, 2004: 139; Wiklund, 2012, 2016; Wiklund et al., 2021). For example,
findings from several studies demonstrate that children with ASD are able to partic-
ipate in question–answer adjacency pair sequences in relatively competent ways
(Kremer-Sadlik, 2004; Wiklund & Määttä, 2021; Wiklund, 2016; Wiklund et al.,
2021). Children with ASD are also able to interpret the implicit conversational
meanings of combinations of dialogue particles, prosodic features, and gestures
correctly (Wiklund, 2012, 2016; Wiklund et al., 2021), and at least in our data, they
can use prosodic features—such as utterance-final pitch rises, creaky voice, stress,
quiet voice, and falling pitch—as interactional resources (Wiklund et al., 2021).
Children with ASD are also able to launch narratives in conversation (Solomon,
2004; Wiklund, 2012, 2016; Wiklund et al., 2021) and notice social rule violations
(Sterponi, 2004).
In addition, Volden’s (2004) results show that the repair abilities of children with
ASD are in many respects similar to those of children without ASD. For example, the
children with ASD used a variety of repair strategies. They were also able to respond
to requests for clarification and add more information when a breakdown persisted.
However, compared with the neurotypical controls, the children with ASD were also
significantly more likely to respond inappropriately when they were faced with a
request for clarification. According to Geller’s (1998) results, children with ASD
attempt to repair most of the communication breakdowns occurring in a free play
context. Similarly, in a study by Ohtake et al. (2011), 12 verbal students with ASD
repaired 80% of communication breakdowns. In their repair attempts, the informants
used forms of communication that reflected the social meaning of the breakdown as
well as more unconventional forms.
This book is based on a research project entitled “Interaction of Preadolescents
with Autism—Focus on Speech Prosody, Gaze Behavior and Misunderstanding Situ-
ations”, which was launched in 2009 as a postdoctoral research project funded
by the Kone Foundation (Finland). In 2014–2015, the project was funded by the
Emil Aaltonen Foundation (Finland), in 2015–2019 by the Helsinki Collegium for
Advanced Studies of the University of Helsinki (Finland), and in 2022 by the Future
Fund of the University of Helsinki. This book will focus on the following aspects
of speech and interaction of preadolescents with ASD, which were studied within
the framework of the project: speech prosody (characteristic features, perception of
atypicality by neurotypical listeners, prosody, and comprehension problems), disflu-
encies of speech (comparison with neurotypical controls, disfluencies, and compre-
hension problems), comprehension problems (overly literal interpretation of speech,
topical discontinuities, the role of eye contact, and other causes), gaze behaviour
(eye contact avoidance strategies, using gaze as a source of feedback) as well as
therapists’ response strategies and teaching orientations in group sessions involving
preadolescent boys with ASD.
1.2 Data 5

1.2 Data

Where most studies on ASD are based on data from experimental settings, this book
is based on naturally occurring audio-visual data from group therapy sessions where
11–13 year-old Finnish (Sect. 1.2.1) and French-speaking (Sect. 1.2.2) boys with
ASD talk with each other and with their neurotypical, adult therapists. Control group
data with age- and gender-matched controls have also been collected for the Finnish
data (Sect. 1.2.3). In addition, experimental data in which neurotypical university
students assess the atypicality of the speech of the Finnish-speaking informants with
and without ASD have been collected (Sect. 1.2.4). The data will be presented in
greater detail in the following sub-sections.

1.2.1 Finnish Data on Participants with ASD

The Finnish data on participants with ASD consist of audio-visual recordings of

neuropsychiatric group therapy sessions in which two groups of 11–13 year-old
boys (Group A and Group B) engage in a discussion with their therapists and with
each other. The aim of these sessions is to teach the children interactional skills and
group activities. Group A consists of three participants and two therapists, and Group
B consists of four participants and two therapists. One of the therapists is a man and
the other one is a woman. The male therapist is the same in both sessions; the female
therapist is not the same. The Group A session was recorded in 2009, and the Group
B session was recorded a few months later in 2010 in a hospital located in southern
Finland.2 The author of this book was present when the data were collected, but she
sat behind a one-way mirror so that the participants could not see her. Figure 1.1
illustrates the setting during the Group A session. The names of the participants
have been changed, and the image has been manipulated so that the persons are
not recognisable. ‘MT’ refers to the male therapist, and ‘FT’ refers to the female
therapist.
Figure 1.2 illustrates the setting during the session of Group B. The names of
the participants have been changed, and the image has been manipulated so that the
persons are not recognisable.
Finnish is the only native language of all the participants. The boys all lived in
southern Finland at the time the data were collected. The places of birth of the boys
are not known by the author of this book, and it is not known how long each of
them had lived in southern Finland when the data were collected. The participants
were recruited through the staff of the hospital. The participants and their parents
were provided with an informed consent form as well as a detailed information sheet
in Finnish. Separate forms were given to the children and the parents. Consent was
given in writing by signing the informed consent form and the information sheet. The
study was evaluated and approved by the hospital ethics committee (decision number

2 The names of the city and the hospital are not mentioned due to ethical restrictions.
6 1 Introduction

Fig. 1.1 Setting during the Group A session

Fig. 1.2 Setting during the Group B session

284/13/03/03/2009). A research permit was granted, which has been extended once.
The participants received no monetary compensation, and their participation was
entirely voluntary. All participants were treated equally. The participants had the right
to refuse to participate and to withdraw at any time without any adverse consequences.
There was no risk of coercion or harm to the participants. The data have been treated
anonymously.
Most participants of the Finnish data had been diagnosed with Asperger Syndrome
at the time the data were collected. Both groups also included one member who
had not been officially diagnosed at the time. However, according to the therapists,
these two boys had the same symptoms as the other members of the group and
participated in the group sessions due to these symptoms. The two participants (AFi6
and AFi7) without an official ASD diagnosis have been excluded from the quantitative
studies presented in this book. As members of the group, these informants have,
however, been taken into account in certain qualitative analyses included here. All the
1.2 Data 7

Table 1.1 Diagnoses of the Finnish-speaking participants with ASD

Participant Diagnoses
AFi1 Asperger Syndrome
AFi2 Asperger Syndrome, ADHD, and Tic disorder
AFi3 Asperger dyndrome, with difficulties in executive functions
AFi4 Asperger dyndrome, with difficulties in social interaction, attention, and executive
functions
AFi5 Asperger dyndrome
AFi6 Non-verbal learning disorder, tourette Syndrome
AFi7 –

participants had normal IQ. Some of the participants also had co-morbid diagnoses:
it is noteworthy that three boys had attentional deficits and/or difficulties in executive
functioning as co-morbid diagnoses, in addition to the Asperger Syndrome diagnosis.
The diagnoses of the Finnish-speaking participants of the ASD group are given in
Table 1.1.
Asperger Syndrome used to be its own specific diagnosis (APA, 2000), but nowa-
days it is included in the larger diagnostic category of Autism Spectrum Disorder in
the widely used DSM-5 classification (APA, 2013). The separate Asperger Syndrome
diagnosis will also disappear from the WHO’s ICD-11 classification, which will be
implemented in 2022. Accordingly, this author will mainly use the term ‘Autism
Spectrum Disorder’ (ASD) to refer to the condition of the informants as well as to
refer to the condition in general (for example, when referring to previous research
literature). Separate diagnoses (such as Asperger Syndrome and High-Functioning
Autism) will be used only when considered particularly relevant.
The total duration of each group therapy session was two hours. Two cameras
were used to film the sessions. One camera was stable and filmed the situation as
a whole. The other camera was directed towards the face of the participant who
was speaking. Each informant also wore a microphone that was attached to his ear
so that the distance between the mouth and the microphone remained the same the
whole time, even when the participants were moving. The sessions started with
sharing news: each boy told the group what he had been doing lately, how school
was going, and other related matters. After hearing one participant’s news, the others
asked questions about what they had just heard.3 After this, the group discussed a
predetermined theme with the help of a series of drawn pictures. In both sessions
included in the data, the theme was ‘bullying at school’. About one hour after the
beginning of the session, the boys and the therapists took a 20 min break, during
which they had a snack in another room. The break was not included in the data.
The last part of the session consisted of playing a traditional Finnish board game

3 One of the characteristics of ASD is a lack of eagerness to share interests, joy, and achievements
with others or to show interest in other people’s preoccupations (APA, 2013). Reciprocal social
interaction in general is difficult for people with ASD. For these reasons, sharing news and asking
questions are practised in group therapy sessions.
8 1 Introduction

called Star of Africa. The last part of the session has not been taken into account in
the studies presented in this book, because participation in game interaction differs
considerably from participation in a group discussion. The recording was conducted
by a group of audio-visual technology professionals. The data have been annotated
with the speech analysis program Praat (Boersma & Weenink, 2021) and transcribed
in detail using the Jeffersonian conventions of Conversation Analysis (Seppänen,
1997).

1.2.2 French Data with Participants with ASD

The French data consist of an audio-visual recording of a naturally occurring group

therapy session in which four 11–13 year-old boys with ASD talk with each other
and with their two female therapists. The duration of the session is 55 min, and it
was recorded at a private clinic in western Switzerland in 2016. As in the case of
the Finnish data, the aim of the sessions was to teach the children interactional skills
and group activities. The theme of this particular session was ‘taunting’. The boys
were all native speakers of French. All the participants had normal IQ, and had been
diagnosed with ASD (Asperger Syndrome, ASD, or Infantile autism) when the data
were collected. Some of the boys also had co-morbid diagnoses (e.g., ADHD). The
diagnoses of the French-speaking participants are given in Table 1.2.
One camera was used to film the session. Each informant also wore a microphone
that was attached to his ear so that the distance between the mouth and the micro-
phone remained the same the whole time. The filming and the sound recording were
conducted by a local research assistant. As in the Finnish sessions, the French session
started with sharing news: each boy told the group about what he had been doing
lately and how he was feeling at that moment, how school and hobbies were going,
etc. After this, the group watched a video together in which a child was taunted. After
having watched the video and discussed it, the group started a more general discus-
sion on the theme of taunting. During the discussion, the boys were encouraged to
talk about their own experiences and conceptions about taunting and bullying.
Figure 1.3 illustrates the setting during the session of the French-speaking group.
The names of the participants have been changed, and the image has been manipulated
so that the persons are not recognisable.

Table 1.2 Diagnoses of the French-speaking participants with ASD

Participant Diagnoses
AFr1 ASD, Dysgraphia, and ADHD
AFr2 Infantile autism, Buccofacial dyspraxia, Sensory dysorality syndrome
AFr3 Asperger Syndrome
AFr4 Asperger Syndrome
1.2 Data 9

Fig. 1.3 Setting during the session of the French-speaking group

The boys all lived in western Switzerland at the time the data were collected. The
boys’ places of birth are not known to the author of this book. The informants were
recruited through the staff of the private clinic in which the data were collected. All
the participants and their parents were provided with an informed consent form as
well as a detailed information sheet in French. Separate forms were given to the chil-
dren and the parents. The participants gave their consent in writing, having signed
the informed consent form and the information sheet. The participants received no
monetary compensation. The participation was entirely voluntary, and all participants
were treated equally. The participants had the right to refuse to participate and to
withdraw at any time without any adverse consequences. There was no risk of coer-
cion or harm to the participants. The data have been treated anonymously. The data
have been transcribed in detail using the Jeffersonian conventions of Conversation
Analysis (Seppänen, 1997).

1.2.3 Control Group Data

The control group data come from a 30 min-long conversation in which six 11–
13 year-old Finnish-speaking boys talk with their female teacher on the premises
of a school in southern Finland. The control group data were recorded in 2016.
Two cameras were used to film the session, and each participant had a microphone
behind one ear. The filming and the sound recording were conducted by audio-
visual technology professionals. The session began with sharing news, and then the
group discussed bullying at school. The situation was not naturally occurring; it was
constructed for the purposes of the current research project to match the therapy
discussions with participants with ASD. All the informants of the control group
were neurotypical and native speakers of Finnish with normal IQ. None of them was
bilingual.
The members of the control group all lived in southern Finland at the time the
data were collected. The participants’ places of birth are not known to the author of
10 1 Introduction

this book. The participants were recruited through the staff of the school in which
the data were collected. All the participants and their parents were provided with an
informed consent form as well as a detailed information sheet in Finnish. Separate
forms were given to the children and the parents. The participants gave their consent
in writing. The participants received no monetary compensation, and their partic-
ipation was entirely voluntary. All participants were treated equally, and they had
the right to refuse to participate and to withdraw at any time without any adverse
consequences. There was no risk of coercion or harm to the participants. The data
have been treated anonymously. The control group data have been annotated with the
speech analysis program Praat (Boersma & Weenink, 2021) and transcribed in detail
using the Jeffersonian conventions of Conversation Analysis (Seppänen, 1997).

1.2.4 Experimental Data

The experimental data consist of the test results of 50 neurotypical adults who partic-
ipated in a perception test exploring the perception of atypicality of speech prosody
of Finnish-speaking preadolescents with ASD (Wiklund et al., forthcoming). The
stimuli consisted of brief (5.0 s) speech samples drawn from the Finnish ASD groups’
data and the control group data presented above. The stimuli were rated on a five-point
Likert scale ranging from one (‘very typical’) to five (‘very atypical’). In addition, the
test subjects had the opportunity to explain their choices in writing and to comment
on the ratings. The perception test was carried out in a quiet room on the premises of a
Finnish university. The room was furnished with a laptop computer and headphones.
The same computer and headphones were used by all the test subjects. The duration
of the perception test was approximately 60 min. The perception test was conducted
by a research assistant, supervised by the author of this book.
The test subjects were 19–45 year-old neurotypical university students living in
southern Finland (mean age 27.2 years). They were all native, monolingual speakers
of Finnish. All test subjects had normal hearing and normal or corrected-to-normal
eyesight. A large majority (44) of them were female. Five test subjects were male,
and one was self-declared as an agender. The test subjects were recruited through
university students’ mailing lists by a research assistant. All the participants were
provided with an informed consent form in Finnish. They gave their consent by
signing this form. The test subjects did not receive any monetary compensation for
their participation, but they received a cinema ticket as a reward. Participation in the
study was entirely voluntary: all the participants had the right to refuse to participate
and to withdraw their participation at any time without any consequences. They were
at no risk of coercion or harm. All participants were treated equally. The data have
been anonymised.
1.3 Methods 11

1.3 Methods

Different methods have been used in the project. Most of the studies presented in
the book are based on Conversation Analysis (CA), which is a qualitative, inter-
disciplinary framework used for example in the fields of linguistics, sociology, and
logopedics. In general, CA is the study of recorded, naturally occurring talk-in-
interaction. The aim of studies carried out within this framework “is to discover how
participants understand and respond to one another in their turns at talk, with a central
focus on how sequences of actions are generated” (Hutchby & Wooffitt, 2008: 12).
However, CA is not just about how people understand and respond to each other. It
is also about how participants in social activities—conducted through interaction—
build their actions and social activity together through observable and reportable
practices. In contrast with research that taps into impairments by testing individuals,
studying atypical and asymmetric interactions qualitatively using the methods of
CA has the potential to reveal the collaboration and resources of the participants in
the interaction (Wiklund & Laakso, 2019). Therefore, studies using the methods of
CA may lead to an improved comprehension of the causes and interactional conse-
quences of the impairment phenomena that are being studied (cf. Damico & Nelson,
2005). CA has been used in research on ASD (see, e.g., Reilly et al., 2016), but most
studies focused on ASD come from experimental settings.
Detailed transcriptions are an essential part of CA research. As already mentioned,
the entire data set has been transcribed following Jeffersonian CA transcription
conventions (Seppänen, 1997). The CA transcriptions aim at capturing not only
what is being said, but also how something is being said (Hepburn & Bolden, 2013:
57). Indeed, CA transcriptions are based on the assumption that “no order of detail in
interaction can be dismissed a priori as disorderly, accidental, or irrelevant” (Heritage,
1984: 241). The durations of pauses are indicated in brackets in tenths of a second (for
example, (0.5) corresponds to 0.5 s), interrupted words are indicated with a hyphen
(for example, koir-), lengthened sounds are indicated with a colon (for example,
koiraa:), rising pitch is indicated with a question mark (?), flat pitch with a comma
(,), and falling pitch with a full stop (.). A complete list of signs and abbreviations
used in the transcription of the examples is given at the end of the book. The names
of the participants have been changed in the transcriptions to anonymise them.
In the studies presented in this book, comprehension problems are materialised
by other-initiated repairs in which the repair process is started by the recipient of
the problematic utterance (Schegloff et al., 1977; SIL Glossary of Linguistic Terms,
2021). Repair, in the sense used in CA, was first defined by Schegloff et al. (1977) as
“the set of practices whereby a participant interrupts the ongoing course of action to
attend to possible trouble in speaking, hearing or understanding the talk” (Kitzinger,
2014: 229). Trouble refers to such things as “misarticulations, malapropisms, use
of a ‘wrong’ word, unavailability of a word when needed, failure to hear or to be
heard, trouble on the part of the recipient in understanding, incorrect understandings
by recipients” (Schegloff, 1987: 210). Repair ensures “that the interaction does not
freeze in its place when trouble arises, that intersubjectivity is maintained or restored,
12 1 Introduction

and that the turn and sequence and activity can progress to possible completion”
(Schegloff, 2007: xiv). Repairs can be initiated by the speakers themselves or by the
recipients of talk. In other words, the speaker can self-repair, or the recipients can
other-initiate repair by requesting clarification from the speaker or by offering an
interpretation—a candidate understanding—to be confirmed by the speaker (Sche-
gloff, 2007; Schegloff et al., 1977). In ordinary conversation, repair operations are
usually short side events after which the interlocutors return to the ongoing topic of
talk. Quick self-repair within the speaker’s own turn is preferred, and if the recipi-
ents other-initiate repair, they typically leave it to the speakers themselves to do the
actual repair. When studying comprehension problems from the point of view of CA,
it is the subsequent action that shows what the trouble is or whether something in
the previous speaker’s turn is oriented to as troublesome, problematic, repairable, or
ungrammatical, and thus correctable (Schegloff et al., 1977).
In addition to CA, methods of acoustic phonetics and experimental psycholin-
guistics (perception tests based on stimuli drawn from naturally occurring data) have
been used in the project presented in this book. The prosodic analyses have been
carried out with the speech analysis program Praat (Boersma & Weenink, 2021).
Praat makes it possible to objectively measure prosodic parameters, such as funda-
mental frequency (f 0 ), intensity, speech rate, and duration, and to draw acoustic
curves illustrating phenomena related to these parameters. Presentation (Neurobe-
havioral Systems, 2021), a stimulus delivery and experiment control program, was
used to construct the perception test, in which neurotypical test subjects assessed the
atypicality of speech prosody of the Finnish-speaking informants with and without
ASD (Wiklund et al., forthcoming). As the methods used within the framework of
this project have differed significantly depending on the objectives of each specific
study, the methods used in each part of the project will be described in greater detail
in the relevant chapter.

References

Allen, M. L., Haywood, S., Rajendran, G., & Branigan, H. (2011). Evidence for syntactic alignment
in children with autism. Developmental Science, 14, 540–548.
American Psychiatric Association (APA). (2000). Diagnostic and statistical manual of mental
disorders (DSM-IV) (4th ed.). American Psychiatric Association.
American Psychiatric Association (APA). (2013). Diagnostic and statistical manual of mental
disorders (DSM-V) (5th ed.). American Psychiatric Association.
Asghari, S. Z., Farashi, S., Bashirian, S., & Jenabi, E. (2021). Distinctive prosodic features of people
with autism spectrum disorder: A systematic review and meta-analysis study. Scientific Reports,
11(1), 23093.
Asperger, H. (1944). Die “Autistischen Psychopathen” im Kindesalter. Archiv Für Psychiatrie Und
Nervenkrankheiten, 117, 76–136.
Baltaxe, C. (1984). The use of contrastive stress in normal, aphasic and autistic children. Journal
of Speech and Hearing Research, 27, 97–105.
Baltaxe, C., & Simmons, J. (1985). Prosodic development in normal and autistic children. In E.
Schopler & G. Mesibov (Eds.), Communication problems in autism (pp. 95–125). Plenum Press.
References 13

Baltaxe, C., & Simmons, J. (1992). A comparison of language issues in high-functioning autism
and related disorders with onset in children and adolescence. In E. Schopler & G. Mesibov (Eds.),
High-functioning individuals with autism (pp. 210–225). Plenum Press.
Boersma, P., & Weenink, D. (2021). Praat: Doing phonetics by computer. [Computer program].
http://www.praat.org/.
Chevallier, C., Noveck, I., Happé, F., & Wilson, D. (2009). From acoustics to grammar: Perceiving
and interpreting grammatical prosody in adolescents with Asperger Syndrome. Research in
Autism Spectrum Disorders, 3(2), 502–516.
Couper-Kuhlen, E. (2000). Prosody. In J. Verschueren, J.-O. Östman, & C. Bulcaen (Eds.),
Handbook of pragmatics (pp. 1–19). John Benjamins Publishing Company.
Crystal, D. (1980). A first dictionary of linguistics and phonetics. Deutsch.
Cummings, L. (2009). Clinical pragmatics. Cambridge University Press.
Cummings, L. (2014a). Pragmatic disorders. Springer.
Cummings, L. (2014b). Communication disorders. Palgrave Macmillan.
Damico, J. S., & Nelson, R. L. (2005). Interpreting problematic behavior: Systematic compensatory
adaptations as emergent phenomena in autism. Clinical Linguistics & Phonetics, 19(5), 405–417.
De Marchena, A., & Miller, J. (2017). “Frank” presentations as a novel research construct and
element of diagnostic decision-making in autism spectrum disorder. Autism Research, 10(4),
653–662.
Diehl, J., & Paul, R. (2013). Acoustic and perceptual measurements of prosody production on the
profiling elements of prosodic systems in children by children with autism spectrum disorders.
Applied Psycholinguistics, 34(1), 135–161.
Diehl, J., Watson, D., Bennetto, L., Mcdonough, J., & Gunlogson, C. (2009). An acoustic analysis
of prosody in high-functioning autism. Applied Psycholinguistics, 30(3), 385–404.
Fay, W., & Schuler, A. L. (1980). Emerging language in autistic children. University Park Press.
Geller, E. (1998). An investigation of communication breakdowns and repairs in verbal autistic
children. The British Journal of Development Disabilities, 44(87), 71–85.
Globerson, E., Amir, N., Kishon-Rabin, L., & Golan, O. (2014). Prosody recognition in adults
with high-functioning autism spectrum disorders: From psychoacoustics to cognition. Autism
Research, 8(2), 153–163.
Hepburn, A., & Bolden, G. B. (2013). The conversation analytic approach to transcription. In J.
Sidnell & T. Stivers (Eds.), The handbook of conversation analysis (pp. 57–76). Wiley Blackwell.
Heritage, J. (1984). Garfinkel and ethnomethodology. Polity Press.
Holbrook, S., & Israelsen, M. (2020). Speech prosody interventions for persons with autism spec-
trum disorders: A systematic review. American Journal of Speech-Language Pathology, 29(4),
2189–2205.
Hubbard, D. J., Faso, D. J., Assmann, P. F., & Sasson, N. J. (2017). Production and perception of
emotional prosody by adults with autism spectrum disorder. Autism Research, 10(12), 1991-2001.
Hutchby, I., & Wooffitt, R. (2008). Conversation analysis (2nd ed.). Cambridge: Polity Press.
Hutt, C., & Ounsted, C. (1966). The biological significance of gaze aversion with particular reference
to the syndrome of infantile autism. Behavioral Science, 11(5), 346–356.
Kanner, L. (1943). Autistic disturbances of affective contact. The Nervous Child, 2, 217–250.
Kenny, L., Hattersley, C., Molins, B., Buckley, C., Povey, C., & Pellicano, E. (2016). Which terms
should be used to describe autism? Perspectives from the UK autism community. Autism, 20(4),
442–62.
Kitzinger, C. (2014). Repair. In J. Sidnell & T. Stivers (Eds.), The handbook of conversation analysis
(pp. 229–256). Wiley Blackwell.
Korpilahti, P., Jansson-Verkasalo, E., Mattila, M.-L., Kuusikko, S., Suominen, K., Rytky, S., Pauls,
D. L., & Moilanen, I. (2007). Processing of affective speech prosody is impaired in Asperger
Syndrome. Journal of Autism and Developmental Disorders, 37(8), 1539–1549.
Kremer-Sadlik, T. (2004). How children with Autism and Asperger Syndrome respond to questions:
a ‘naturalistic’ theory of mind task. Discourse Studies, 6(2), 185–205.
14 1 Introduction

Kylliäinen, A. (2007). Face and Gaze processing in children with autism. [Doctoral dissertation].
University of Tampere.
Lake, J. K., Humphreys, J. R., & Cardy, S. (2011). Listener versus speaker oriented aspects of
speech: Studying the disfluencies of individuals with autism spectrum disorders. Psychonomic
Bulletin and Review, 18, 135–140.
Lehtinen, M. (2012). “Aiheesta toiseen”—Ymmärrysongelmatilanteita Asperger-nuorten kuntou-
tuskeskusteluissa. In: L. Leskelä & C. Lindholm (Eds.), Haavoittuva keskustelu. Keskustelu-
nanalyyttisia tutkimuksia kielellisesti epäsymmetrisestä vuorovaikutuksesta (pp. 235–277).
Kehitysvammaliitto.
Lewis, F. M., Woodyatt, G. C., & Murdoch, B. E. (2008). Linguistic and pragmatic language skills
in adults with autism spectrum disorder: A pilot study. Research in Autism Spectrum Disorders,
2(1), 176–187.
Lindström, R. (2019). Neural and perceptual processing of emotional speech prosody in typi-
cally developed children and in children with autism spectrum disorder. [Doctoral dissertation].
University of Helsinki.
Maenner, M.J., Shaw, K.A., Baio, J., Washington, A., Patrick, M., Di Rienzo, M., Christensen, D. L.,
Wiggins, L. D., Pettygrove, S., Andrews, J. G., Lopez, M., Hudson, A, Baroud, T., Schwenk, Y.,
White, T., Rosenberg, C. R., Lee, L. C., Harrington, R. A., Huston, M., Hewitt, A., Esler, A., Hall-
Lande, J., Poynter, J. N., Hallas-Muchow, L., Constantino, J. N., Fitzgerald, R. T., Zahorodny,
W., Shenouda, J., Daniels, J. L., Warren, Z., Vehorn, A., Salinas, A., Durkin, M. S., & Dietz,
P. M. (2020). Prevalence of autism spectrum disorder among children aged 8 years—Autism
and developmental disabilities monitoring network, 11 sites, United States, 2016. Morbidity and
Mortality Weekly Report—Surveillance Summaries, 69(4), 1–12.
Martin, I., & McDonald, S. (2004). An exploration of causes of non-literal language problems in
individuals with Asperger Syndrome. Journal of Autism and Developmental Disorders, 34(3),
311–328.
McPartland, J., & Klin, A. (2006). Asperger’s syndrome. Adolescent Medicine Clinics, 17(3),
771–788.
Nadig, A., & Shaw, H. (2012). Acoustic and perceptual measurement of expressive prosody in
high-functioning autism: Increased pitch range and what it means to listeners. Journal of Autism
and Developmental Disorders, 42(4), 499–511.
Nakai, Y., Takashima, R., Takiguchi, T., & Takada, S. (2014). Speech intonation in children with
autism spectrum disorder. Brain and Development, 36(6), 516–522.
Ochs, E., & Solomon, O. (2004). Introduction: Discourse and autism. Discourse Studies, 6(2),
139–146.
Ohtake, Y., Wehmeyer, M., Nakaya, A., Takahashi, S., & Yanagihara, M. (2011). Repair strategies
used by verbal students with autism during free play. Focus on Autism and Other Developmental
Disabilities, 26(1), 3–14.
Paul, R. (1987). Communication. In D. Cohen & Anne Donnellan (Eds.), Handbook of autism and
pervasive developmental disorders (pp. 61–84). Wiley.
Paul, R., Augustyn, A., Klin, A., & Volkmar, F. R. (2005a). Perception and Production of Prosody
by Speakers with Autism Spectrum Disorders. Journal of Autism and Developmental Disorders,
35(2), 205-220.
Paul, R., Shriberg, L. D., McSweeney, J., Cicchetti, D., Klin, A., & Volkmar, F. R. (2005b). Brief
report: relations between prosodic performance and communication and socialization ratings in
high functioning speakers with autism spectrum disorders. Journal of Autism and Developmental
Disorders, 35(6), 861-869.
Redford, M. A., Kapatsinski, V., & Cornell-Fabiano, J. (2018). Lay listener classification and
evaluation of typical and atypical children’s speech. Language and Speech, 61(2), 277–302.
Reilly, M. J. O., Lester, J. N., & Muskett, T. (Eds.). (2016). Journal of Autism and Developmental
Disorders, Special Issue: ‘Discourse and Conversation Analytic Approaches to the Study of
Autism Spectrum Disorder’, 46, 2.
References 15

Rosenblau, G., Kliemann, D., Dziobek, I., & Heekeren, H. R. (2017). Emotional prosody processing
in autism spectrum disorder. Social Cognitive and Affective Neuroscience, 12(2), 224–239.
Scaler Scott, K., Tetnowski, J. A., Flaitz, J. R., & Yaruss, J. S. (2014). Preliminary study of disflu-
ency in school-aged children with autism. International Journal of Language & Communication
Disorders, 49(1), 75–89.
Schegloff, E. (1987). Analyzing single episodes of interaction: An exercise in conversation analysis.
Social Psychology Quarterly, 50(2), 101–114.
Schegloff, E. A. (2007). Sequence organization in interaction. Cambridge University Press.
Schegloff, E., Jefferson, G., & Sacks, H. (1977). The preference for self-correction in the
organization of repair in conversation. Language, 53(2), 361–382.
Scott, K. S., Grossman, H. L., Abendroth, K. J., Tetnowski, J. A., & Anddamico, J. S. (2006).
Asperger Syndrome and attention deficit disorder: clinical disfluency analysis. In: J. Au-Yeung &
M. M. Leahy (Eds.), Research, treatment, and self-help in fluency disorders: New horizons.
Proceedings of the fifth world congress on fluency disorders (pp. 273–278). International Fluency
Association.
Seppänen, E.-L. (1997). Vuorovaikutus paperilla. In: L. Tainio (Ed.), Keskustelunanalyysin perusteet
(pp. 18–31). Vastapaino.
Shriberg, L., Paul, R., McSweeney, J., Klin, A., Cohen, D., & Volkmar, F. R. (2001). Speech and
prosody characteristics of adolescents and adults with high functioning autism and Asperger
Syndrome. Journal of Speech, Language and Hearing Research, 44, 1097–1115.
SIL Glossary of Linguistic Terms. (2021). https://glossary.sil.org/.
Sisskin, V. (2006). Speech disfluency in Asperger’s Syndrome: Two cases of interest. Perspectives
on Fluency and Fluency Disorders, 16(2), 12–14.
Solomon, O. (2004). Narrative introductions: Discourse competence of children with autistic
spectrum disorders. Discourse Studies, 6(2), 253–276.
Sterponi, L. (2004). Construction of rules, accountability and moral identity by high-functioning
children with autism. Discourse Studies, 6(2), 207–228.
Neurobehavioral Systems. (2021). Presentation. [Computer program]. https://www.neurobs.com/.
Szatmari, P., Brenner, R., & Nagy, J. (1989). Asperger Syndrome: A review of clinical features. The
Canadian Journal of Psychiatry, 34, 554–560.
Tager-Flusberg, H. (2000). Understanding the language and communicative impairments in autism.
International Review of Research on Mental Retardation, 23, 185–205.
Tantam, D., Holmes, D., & Cordess, C. (1993). Non-verbal expression in autism of Asperger type.
Journal of Autism and Developmental Disorders, 23(1), 111–133.
Thorson, J. C., Usher, N., Patel, R., & Tager-Flusberg, H. (2016). Acoustic analysis of prosody in
spontaneous productions of minimally verbal children and adolescents with autism. In: J. Barnes,
A. Brugos, S. Shattuck-Hufnagel & N. Veilleux (Eds.), Proceedings of Speech Prosody (Vol. 8).
USA.
Van Bourgondien, M. E., & Woods, A. V. (1992). Vocational possibilities for high functioning
adults with autism. In E. Schopler & G. Mesibov (Eds.), High functioning individuals with autism
(pp. 227–242). Plenum Press.
Vivanti, G. (2020). Ask the editor: What is the most appropriate way to talk about individuals with
a diagnosis of autism?Journal of Autism and Developmental Disorders 50, 691–693. https://doi.
org/10.1007/s10803-019-04280-x.
Volden, J. (2004). Conversational repair in speakers with autism spectrum disorder. International
Journal of Language & Communication Disorders, 39(2), 171–189.
Wiklund, M. (2012). Gaze behavior of pre-adolescent children afflicted with Asperger Syndrome.
Communication & Medicine, 9(2), 173–186.
Wiklund, M. (2016). Interactional challenges in conversations with autistic preadolescents: the role
of prosody and nonverbal communication in other-initiated repairs. Journal of Pragmatics, 94,
76–97.
16 1 Introduction

Wiklund, M., & Laakso, M. (2019). Ungrammatical utterances and disfluent speech as causes
of comprehension problems in interactions of preadolescents with highly functioning autism.
Clinical Linguistics & Phonetics, 33(7), 654–676.
Wiklund, M., & Laakso, M. (2020). Comparison of disfluent and ungrammatical speech of pread-
olescents with and without ASD. Journal of Autism and Developmental Disorders, 51(8),
2773–2789.
Wiklund, M., & Määttä, S. (2021). Therapists’ response strategies in a group session involving
French-speaking children with autism spectrum disorder. Journal of Interdisciplinary Voice
Studies, 6(1), 109–130.
Wiklund, M., & Stevanovic, M. (2018). Ymmärrysongelmat ja vuorovaikutustaitojen opet-
taminen lievästi autististen varhaisnuorten kuntoutuskeskusteluissa [Understanding problems
and the teaching of social interaction skills in group therapy sessions of preadolescents with
high-functioning autism]. Psykologia, 53(05–06), 421–451.
Wiklund, M., & Vainio, M. (2019). Pitch-related features in the speech of Finnish- and French-
speaking boys with autism in data coming from group therapy sessions. In H. E. H. Lenk, J.
Härmä, B. Sanromán Vilas, & E. Suomela-Härmä (Eds.), Studies in comparative pragmatics
(pp. 45–63). Cambridge Scholars Publishing.
Wiklund, M., Ihaksinen, K., & Vainio, M. (2021). Autismikirjon poikien intonaation käyttö
kysymys-ja kerrontavuoroissa spontaanissa vuorovaikutuksessa [The Use of Intonation in Ques-
tions and Narrative Turns in the Spontaneous Interaction of Boys Afflicted with Autism Spectrum
Disorder]. Puhe Ja Kieli, 41(1), 43–70.
Wiklund, M., Vainio, L., Saalasti, S., & Vainio, M. (forthcoming). Puheen prosodian havaittu
epätyypillisyys suomenkielisillä autismikirjon varhaisnuorilla [Perceived Atypicality of Speech
Prosody of Finnish-Speaking Preadolescents with Autism Spectrum Disorder]. Puhe ja kieli.
Zhou, P., Ma, W., & Zhan, L. (2019). A deficit in using prosodic cues to understand communicative
intentions by children with autism spectrum disorders: An eye-tracking study. First Language,
40(1), 41–63.
Chapter 2
Prosody

Abstract ‘Prosody’ encompasses phenomena related to aspects of speech such as

melody, rhythm, stress, phrasing, pauses, and voice quality. These are related to
pitch (voice fundamental frequency, f 0 ), loudness (intensity), duration (timing), and
speech rate. Prosodic phenomena are ‘suprasegmental’. In other words, they typically
involve units of at least one syllable in length. The results of a study by Wiklund
et al. (forthcoming) proved both experimentally and statistically that neurotypical
Finnish-speaking adults find the prosody of preadolescent, Finnish-speaking boys
with ASD more atypical than the prosody of age- and gender-matched controls. In
general, the speech samples of the ASD individuals were rated as significantly more
atypical than the samples of the control group. Potential causes of the perception of
atypicality include the following prosodic features: sing-song pitch, bouncing pitch,
disconnected speech rhythm, large pitch excursions, and flat pitch. It is noteworthy
that these features surprisingly often gave the impression of a non-native accent.
Wiklund (2016) demonstrates that the most common prosodic feature related to
trouble-source turns causing comprehension problems in interaction between pread-
olescents with ASD and neurotypical adults is a creaky voice. A quiet voice and large
pitch excursions are also relatively frequent. It is noteworthy that even if the trouble-
source turns often carry certain prosodic features in these data and these features
may contribute to the creation of the repair sequence, in very few cases do prosodic
features seem to be the main cause of the comprehension problem (Lehtinen, 2012;
Wiklund, 2016). The results of a study by Wiklund et al. (2021) show that speakers
with ASD can use utterance-final pitch rises as an interactional resource. Indeed, they
can call for other participants’ reactions and indicate that they take other participants
into account with the help of utterance-final pitch rises (Routarinne, 2003). They
can also ‘recycle’ prosody, that is, repeat the prosody of a previous speaker. In addi-
tion, the informants can correctly produce and interpret prosodic features indicating
finality in a spontaneous interaction. They can also emphasise words with the help
of stress and changes in intonation.

Keywords Prosody · Intonation · Interaction · Perception of atypicality ·

Comprehension problems · Utterance-final pitch rises · ASD · Autism spectrum
disorder

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 17
M. Wiklund, Speech and Interaction of Preadolescents with Autism Spectrum Disorder,
Prosody, Phonology and Phonetics, https://doi.org/10.1007/978-981-19-8117-3_2
18 2 Prosody

2.1 Overview of Previous Research on the Prosody

of Persons with ASD

Within the field of phonetics, prosody encompasses phenomena related to aspects

of speech, such as melody, rhythm, stress, phrasing, pauses, and voice quality.
These are related to pitch (voice fundamental frequency, f 0 ), loudness (intensity),
duration (timing), and speech rate (Couper-Kuhlen, 2000; Crystal, 1980). Prosodic
phenomena are suprasegmental. In other words, these phenomena typically involve
units of at least one syllable in length. It is well known that people with ASD tend
to have deviant prosodic features in their speech. Indeed, disordered prosody was
already included in early descriptions of autism (Asperger, 1944; Kanner, 1943).
The speech of persons with ASD is often described as monotonous (see, e.g., Nakai
et al., 2014). However, the pitch range of people with ASD may also be excep-
tionally wide (see, e.g., Baltaxe, 1984; Fosnot & Jun, 1999; Diehl & Paul, 2013;
Nadig & Shaw, 2012). Their speech may also be overly fast, ‘jerky’, or loud, or it
may be characterised by large pitch excursions, a quiet voice, an inconsistent pause
structure, deviant word stress, and/or a creaky or nasal voice (Asghari et al., 2021;
Asperger, 1944; Baltaxe, 1984; Baltaxe & Simmons, 1985, 1992; Fay & Schuler,
1980; Lehtinen, 2010a; Paul, 1987; Paul et al., 2005a, 2005b; Shriberg et al., 2001;
Tager-Flusberg, 2000; Thorson et al., 2016). Generally speaking, prosodic extremes
are typical of persons with ASD (Wiklund & Vainio, 2019). As mentioned in the intro-
duction chapter, according to a study by Paul et al. (2005a: 205), deviant prosodic
features may constitute a significant obstacle to the social acceptance of the indi-
vidual. Indeed, atypical prosody has been identified as one of the most recognisable
features specific to ASD (De Marchena & Miller, 2017). Deviant prosodic features
may create an immediate impression of ‘oddness’ (Van Bourgondien & Woods,
1992). These features also affect how speakers with ASD are rated in terms of social
and communicative competence (Paul et al., 2005b) as well as in terms of likeability
(Redford et al., 2018). In addition, prosodic features may impair the intelligibility of
speech (Redford et al., 2018).
According to a study by Baltaxe (1984), children with ASD often have very
narrow or very wide pitch (fundamental frequency, f 0 ) ranges. The results of Fosnot
and Jun (1999) are similar: children with ASD exhibit greater pitch variation and
pitch range than age-matched typically developing controls. Increased pitch range
among children with ASD has also been reported by Diehl and Paul (2013) as well
as by Nadig and Shaw (2012). In the study by Nadig and Shaw (2012), the speakers
with ASD were rated by listeners as having atypical prosody, but they were not rated
as having increased pitch variation, even if the ASD group in fact demonstrated
increased pitch range compared to typically developing speakers. The authors suggest
that the listeners not rating the speakers with ASD as having increased pitch variation
was related to their non-conventional use of pitch variation. The informants with ASD
did not differ from typically developing controls concerning mean pitch or speech
rate (Nadig & Shaw, 2012).
2.1 Overview of Previous Research on the Prosody of Persons with ASD 19

Earlier, Fine et al. (1991) reported that persons with high-functioning autism
(HFA) were less likely to adopt useful patterns of intonation for communication than
persons with Asperger’s syndrome (AS) or psychiatric out-patient controls. Subjects
with AS, in turn, differed little from the controls.
Paul et al. (2005a, 2005b) have shown that persons with ASD may have difficulties
in producing appropriate stress patterns. On the other hand, informants with ASD
did not have difficulties in the use or perception of pragmatic-affective prosody (Paul
et al., 2005b). In other words, they could recognise speaking styles directed towards
children and adults as well as the control group members. Fine et al. (1991), in turn,
reported that persons with Asperger’s Syndrome (AS)1 were able to adopt useful
patterns of intonation for communication, whereas persons with High-Functioning
Autism (HFA) were less likely to do so. According to the results of Shriberg et al.
(2001), people with ASD have notable difficulties with the pragmatic and affective
use of prosody, but not with its grammatical functions. Similarly, when Grossman
et al. (2010) studied lexical stress production among children with ASD, their results
showed that these children were able to produce appropriately differentiated lexical
stress patterns. However, the stress patterns were atypically long. According to the
authors, this result indicates a reduced ability to produce natural prosody among
children with ASD.
Olivati et al. (2017) have studied the prosodic features of speech samples produced
by students with ASD.2 The authors found significant differences between the ASD
group and the control group concerning the following variables: tessitura (pitch
range), melodic amplitude of tonic vowel, melodic amplitude of pretonic vowel,
maximum intensity, minimum intensity, tonic vowel duration, pretonic vowel dura-
tion, and phrase duration. Results of recent studies based on the Finnish and French
data of this book (see Sect. 1.3) show that prosodic extremes are typical of preadoles-
cents with ASD (Wiklund & Vainio, 2019; Wiklund et al., forthcoming; Lohi, 2020).
The pitch of the informants may range from a remarkably flat and monotonous pitch
to a sing-song pitch characterised by large excursions of the f 0 . The same studies
show that the speech rate of the informants with ASD may also be remarkably fast or,
on the other hand, very slow. On the other hand, prosodic extremes are not exhibited
by all persons with ASD, and for example sing-song pitch does not necessarily occur
in the speech of an individual speaker all the time (Wiklund & Vainio, 2019).
In a study by Green and Tobin (2008), Hebrew-speaking children with ASD
produced more pitch accents than typically developed controls, and the use of the
high tone (H*) was more common in the ASD group than in the control group.
Another difference reported by the authors was that the repertoire of prosodic edge
tones was more limited among the participants of the ASD group than among those of
the control group. These findings correspond to the wider pitch range, ‘bounciness’,
and unusual pause structures reported in other studies. The same authors (Green &

1 The names of the specific diagnoses have been maintained here, because two different groups are
being compared.
2 Olivati et al.’s (2017) data are in Brazilian Portuguese. The other studies presented in this sub-

chapter are based on English data if the language is not specified.

20 2 Prosody

Tobin, 2009) have also found that, as a group, children with ASD exhibit wider
pitch ranges and larger pitch variability compared to typically developed children,
but when considered individually, children with ASD could be divided into three
distinct groups: those with narrow, typical, or wide pitch ranges.
The speech of adults with ASD has been studied much less than the speech of
children with ASD. A large pitch range has, however, also been reported among adults
with ASD (DePape et al., 2012). More precisely, the findings of DePape et al. (2012)
indicate that adults with ASD and high language functioning generally use a wider
pitch range than neurotypical controls but do not mark the information structure,
whereas adults with ASD and with moderate language functioning generally use a
narrower pitch range but mark the information structure appropriately.
A recent study by Lau et al. (2022) concerning the prosody of American English
and Hong Kong Cantonese speakers with ASD revealed rhythm-relative features that
are consistent across the two languages. Intonational features, in turn, appeared to
be language-specific. The participants of the study were 6–35 year-old males and
females. The data were collected from a narrative task.
Concerning the perception of prosody, it is known that the processing of emotional
prosody is impaired among those with ASD. Rosenblau et al. (2017), for example,
found evidence of aberrant processing of emotional prosody among adults with
ASD. Similarly, results reported by Korpilahti et al. (2007) concerning 9–12 year-
old Finnish-speaking boys indicate atypical neural responses to affective prosody
among children with ASD, as well as among their fathers. According to the authors,
these results provide evidence of the existence of familial patterns of abnormal audi-
tory brain reactions to prosodic features. Similarly, Globerson et al. (2014) found
that adults with ASD performed more poorly than neurotypical controls in affective
prosody recognition. In the same study, no differences were found between the ASD
group and the control group concerning pragmatic prosody recognition or psychoa-
coustic tasks that tested the recognition of pitch direction and pitch discrimination.
However, the findings of Zhou et al. (2019), who have studied the perception of
prosodic cues among Mandarin-speaking preschool children with ASD, suggest that
the ability to use prosodic cues to understand a speaker’s communicative intention
is impaired in ASD.
The ability of persons with ASD to understand grammatical prosody has been
studied by Chevallier et al. (2009). The authors conducted experiments that tested
the interpretation of word stress, grammatical pauses, and discrimination of the ques-
tion contour. The experiments also included acoustic tasks to assess perceptions of
pitch, duration, intensity, and prosodic contours. The test subjects were teenaged boys
diagnosed with ASD. According to the results, the informants with ASD performed
as well as the typically developing control group members in all the tasks. This result
suggests that grammatical prosody is not affected by ASD. On the other hand, Lind-
ström’s (2019) results based on Finnish data suggest that the processing of natural
speech prosody is impaired in children with ASD on various levels of information
processing. These levels include aberrant discrimination of prosodic changes, as well
as problems in orienting to as well as sluggishness in responding to such changes.
2.1 Overview of Previous Research on the Prosody of Persons with ASD 21

A study by Le Normand et al. (2008) focused on prominence and prosodic contours

related to speech acts. The data came from recorded free play sessions during which
French-speaking children with ASD played with a speech pathologist. The findings
showed that the children with ASD had difficulties in producing prosodic promi-
nence: low prominence was highly frequent in their speech compared with typically
developing controls. Although prominence is a perceptual term, this result seems to
point to more monotonous prosodic patterns—further strengthened by the fact that
the children with ASD also produced many words with flat contours. According to
the same study, on the level of communication, the children with ASD had diffi-
culties in matching the form and function of the prosodic contours. For example,
declaratives were frequently produced with a flat contour. Exclamations were less
atypical, consisting mainly of screams out of pitch range. Concerning questions, no
clear pattern of prosodic contours emerged in this study, which, according to the
authors, could have been related to the low number of questions produced by the
children with ASD in the data analysed for the study.
Prosodic atypicality among children with ASD has been studied experimentally
by Peppé et al. (2006). The authors carried out tests assessing the functionality
of prosody in four aspects of speech: phrasing, affect, turn end, and focus. The
objective was to find out if deviations in these aspects also affected the communicative
functions of prosody. In addition, five adults with phonetic awareness and a group of
22 phonetically naïve adult judges assessed prosodic atypicality among the children
with ASD (AS and HFA) as well as among typically developing children. The native
language of the participants was English (with a Scottish accent). The HFA3 group
scored significantly lower than the typically developing group on affect, focus, turn
end, and chunking. The HFA group also scored significantly lower than the AS group
on imitating words, and much lower on affect, focus, and the imitation of phrases.
Interestingly, the difference between the test results of the AS group and the typically
developing group was significant on only one task: the imitation of phrases. In the
assessment of perceived atypicality, the phonetically aware judges rated the HFA
group as more atypical than the AS group. The AS group, in turn, was rated as
more atypical than the typically developing group. According to the ratings of the
phonetically naïve judges, the perceived difference of atypicality between the HFA
and the typically developing groups was highly significant. There was no perceived
difference between the HFA and the AS groups. There was a tendency to perceive
more prosodic atypicality in the AS group than in the typically developing group, but
the difference did not quite reach statistical significance. These results imply that—
at least in a test environment—the prosody of children with HFA is perceived as
atypical, and that their use of prosody may also be misleading or at least ambiguous.
Children with AS sound less atypical than children with HFA, and they also have
better functional prosodic skills. The causes of the perceptions of atypicality are not
analysed in the study.

3The names of the specific diagnoses (HFA and AS) have been maintained here, because the study
compares the results of two diagnostic groups.
22 2 Prosody

Another study concerning prosodic atypicality among English-speaking children

with ASD has been conducted by Redford et al. (2018). In this study, college-aged
neurotypical adults assessed speech samples obtained from children with ASD from
a storytelling task. The results were compared with those of typically developing
children. The test subjects were asked to make judgments of ‘typical’ and ‘disordered’
based on the speech samples. The results of the study indicate that test subjects are
more likely to assess speech produced by children with ASD as ‘disordered’ than
speech produced by typically developing children. According to the same study,
college-aged test subjects rate children with ASD more negatively than their typically
developing peers concerning ‘likeability’. In addition, the ASD group was rated as
less intelligible on average than the typically developing control group.
Thus, it is known that persons with ASD tend to have salient prosodic features
in their speech, and that their perceptions of prosody are in many respects impaired.
Studies have also shown that neurotypical persons perceive the prosody of persons
with ASD as ‘atypical’ or ‘disordered’, at least in a test environment in the case of
English-speaking children (Peppé et al., 2006; Redford et al., 2018).
The following sub-chapters will explore the prosodic characteristics of preado-
lescent boys with ASD (Sect. 2.2), the perception of ‘atypicality’ by neurotypical
test subjects (Sect. 2.3) as well as the role of prosodic features in the creation of
comprehension problems (Sect. 2.4) and the use of prosody by preadolescent boys
with ASD (Sect. 2.5) in the Finnish and French data presented in the Sect. 1.2.

2.2 Prosodic Characteristics of Preadolescent Boys

with ASD and Perception of Atypicality
by Neurotypical Listeners

This sub-chapter is based on a recent study by Wiklund et al. (forthcoming)4

describing prosodic characteristics of preadolescent boys with ASD, and the percep-
tion of atypicality of the speech prosody of these persons by neurotypical listeners.
The data consist of the Finnish data on informants with ASD (see Sect. 1.2.1), the
control group data (see Sect. 1.2.3), and the experimental data (see Sect. 1.2.4). In
total, 150 speech samples were collected from the production data (Sects. 1.2.1 and
1.2.3), and a perception test was constructed. The test included 75 speech samples
(stimuli) from the ASD groups and 75 samples from the control group. The duration
of each stimulus was five seconds. The speech analysis program Praat (Boersma &
Weenink, 2021) was used to collect the stimuli, in chronological order from the begin-
ning of the individual sound recordings of each participant. Sequences in which there
was overlapping speech or poor sound quality were not included. The intensity levels
of the stimuli were normalised. ‘Presentation’ (Neurobehavioral Systems, 2021), a

4 The article is published in Finnish.

2.2 Prosodic Characteristics of Preadolescent Boys with ASD … 23

stimulus delivery and experiment control program, was used to construct the percep-
tion test, in which the stimuli were presented in a randomised order. The number
of speech samples per participant in the ASD group was 15. The number of control
group members’ samples varied between 4 and 20 per participant (mean: 13), because
the data come from spontaneous interaction, and therefore some of the participants
spoke much less than the others; there was also variation in the quality of sound and
the amount of overlapping speech.
In total, 50 neurotypical subjects participated in the perception test. The stimuli
were rated on a five-point Likert scale ranging from one (‘very typical’) to five
(‘very atypical’). In addition, the test subjects had the opportunity to explain their
choices in writing and to comment on the ratings. The written answers gave us
important information concerning the prosodic features and other features of speech
that attracted the test subjects’ attention. The perception test was carried out in a
quiet room furnished with a laptop computer and headphones: the same computer
and headphones were used by all the test subjects. The duration of the perception
test was approximately 60 min.
The test subjects were 19–45 year-old neurotypical university students living in
southern Finland (mean age 27.2 years). They were all monolingual native speakers
of Finnish with normal hearing and normal or corrected-to-normal eyesight. A large
majority (44) were female. Five were male, and one was self-declared as an agender.
The test subjects were naïve about the aim of the study. A research assistant recruited
the subjects through university students’ mailing lists. All the participants were
provided with an informed consent form in Finnish: they gave consent by signing
the form. They were not paid for their participation, but received a cinema ticket as
a reward.
A total of 7,500 responses from the participants (n = 50) were divided unevenly
among the response levels, one being ‘very typical’ and five ‘very atypical’ (1 =
34.6%, 2 = 28.7%, 3 = 19.5%, 4 = 12.8%, and 5 = 4.3%). Only a few participants
rated the samples as ‘very atypical’ (5). Notably, as Fig. 2.1 shows, the speech
samples of persons with ASD yielded strikingly fewer ‘very typical’ (1) responses
in comparison to those of the control group. In addition, the speech samples of
the individuals with ASD, in comparison with members of the control group, were
more commonly associated with the third (moderately atypical) and fourth (atypical)
response levels.5
The following sub-chapters will present the five stimuli that obtained the highest
atypicality ratings in the perception test (Wiklund et al., forthcoming). The exam-
ples are analysed acoustically using the Praat speech analysis program (Boersma &
Weenink, 2021), supported by the test subjects’ comments on their ratings.

5 For more precise statistical analyses, see Wiklund et al. (forthcoming).

24 2 Prosody

Fig. 2.1 The mean

typicality ratings (1 = very
typical, 2 = typical, 3 =
moderately atypical, 4 =
atypical, and 5 = very
atypical) of the speech
samples associated with the
ASD individuals and the
control group: error bars
depict the standard error of
the mean and asterisks
indicate statistically
significant differences
(***p < 0.001)

2.2.1 Sing-Song-Like Pitch

The first example is of a salient pitch pattern in the speech of a boy with ASD,
consisting of large, sing-song pitch movements. Indeed, as Fig. 2.2 shows, the pitch
goes up and down almost as if the speaker were singing. During the production of
the word luistelulla (‘skating’), for example, the pitch level increases approximately
one octave or 12.1 semitones (st) between the first and the last syllable, and during
the production of the word kaaduin (‘I fell’) it rises to 8.2 st. Moreover, the pitch rise
occurs in an unstressed syllable, which is atypical in Finnish with its fixed lexical
stress on the first syllable of the word. The fundamental frequency (f 0 ) mean in this
speech sample is 207.8 Hz, which is a normal value for a preadolescent boy. The
standard deviation of the f 0 is 90.7 Hz (i.e., 8.3 st), which corresponds to a large
range. The second half of the speech sample (beginning from the interrupted word
pol-, ‘kne-’) is produced with a creaky voice, which breaks the pitch curve and makes
it impossible to analyse acoustically. Figure 2.2 illustrates the pitch curve during the
production of the sample.
As mentioned above, the subjects (n = 50) in the perception test also rated the
stimuli on a five-point Likert scale, 1 being ‘very typical’ and 5 being ‘very atypical’.
This sample obtained the highest rating of all 150 stimuli, with a mean rating of 4.0.
Such a rating is not surprising given that, in general, Finnish intonation is charac-
terised by small pitch intervals and a narrow variation range in pitch movements
(Iivonen, 1998). Thus, the large, sing-song pitch excursions occurring in this sample
create a strong impression of atypicality.
The test subjects (n = 50) also had the opportunity to describe their ratings in
writing. Six of these written responses could, however, not be taken into account
in this study due to a technical mistake made by the test subjects that resulted in a
mismatch between the stimuli and the written responses. Thus, the written responses
of 44 test subjects were included in the study (n = 44). Within this sample (Example
2.1), 30 subjects provided a written description. As many as 14 of these subjects
2.2 Prosodic Characteristics of Preadolescent Boys with ASD … 25

30
24
Pitch (semitones re 100 Hz)

12

100 Hz
0

-12
mä olin luistelulla mä kaaduin ja mun pol- (0.4) ja polvi (0.3) vähän

I was skating I fell and my kne- and knee a little

0 5
Time (s)

Fig. 2.2 Example of a sing-song-like pitch produced by a boy with ASD

(46.7%) wrote that the speech sounded as if the speaker were not a native speaker of
Finnish—even if Finnish was in reality the speaker’s only native language. However,
only five test subjects (16.7%) explicitly mentioned ‘intonation’ in their responses.
The ungrammaticality of the language was mentioned more frequently (10 responses,
33.3%) as a salient feature: indeed, the word luistelu (‘skating’) is in the wrong case
here, which sounds atypical in the speech of a native Finnish speaker. Thus, the
results indicate that the perception of atypicality is attributable here not only to the
sing-song pitch excursions but also to the grammatical problem at the beginning of
the sample. Four test subjects (13.3%) mention ‘word searches’ (i.e., problems in
lexical retrieval) in their responses, which probably relates to the occurrence of the
interrupted word pol- (‘kne-’) in the middle of the sample.
Interestingly, none of the test subjects mentioned that the second half of the sample
was produced in a creaky voice. This probably reflects the fact that a creaky voice is
very common in spoken Finnish. It is often used as an interactional cue indicating
the end of a speaking turn (Ogden, 2001). It is also common for a whole utterance
to be produced mostly, or even entirely, with a creaky voice (Aho, 2010).

2.2.2 Bouncing Pitch

The second example concerns another salient pitch pattern in the speech of a boy with
ASD, a ‘bouncing pitch’. First, the pitch rises rapidly (9.9 st) during the production
of the words oon nyt (‘I have’ + ‘now’), and then falls to 5.1 st between the word
nyt (‘now’) and kiirees-, which is the beginning of the word ‘quickly’ (see Fig. 2.3).
The last syllable ([ti]) of the word kiireesti (‘quickly’) is produced on a very low
pitch level with somewhat irregular phonation, making pitch detection difficult. The
syllable [ti] is also produced separately from the beginning of the word, after a brief
interruption. Then the pitch rises again suddenly between katto- and nu (katto is the
26 2 Prosody

30
24
Pitch (semitones re 100 Hz)

12

100 Hz
0

-12
oon nyt kiirees -ti kattonu sen m kuinka se on vedetty läpi

I have now quick -ly checked it m how people have played through it

0 4.995
Time (s)

Fig. 2.3 Example of a bouncing pitch produced by a boy with ASD

beginning of the word ‘checked’, and nu is the last syllable of the same word): the
size of the pitch rise is as much as 13.5 st. After this strong rise, the pitch falls to 8.7 st
during the word sen (‘it’). The boy then produces an isolated sound [m], which again
briefly interrupts the speech flow. The rest of the utterance (kuinka se on vedetty läpi,
“how people have played through it”) is produced on a low pitch level with a creaky
voice: this breaks the pitch curve and makes it impossible to analyse acoustically,
although the last disyllabic word (läpi) has pitch rises in both syllables. The mean
f 0 in the sample is 134.0 Hz, which is rather low for a preadolescent boy, although
the standard deviation is 35.2 Hz (or 5.3 st), which falls within the moderate range.
According to our analyses, these consecutive strong pitch rises and falls, associated
with a ‘jerky’ rhythm created by the interruptions ([ti] + [m]), fuel the perception of
atypicality in this speech sample.
The mean rating of this sample in the perception test was 3.7. In addition, 20
test subjects provided a written description, most of which refer to the speaker’s
way of speaking as ‘somehow odd’, without specifying the ‘oddness’ in more detail.
Three subjects (15.0%) mention the ‘speech rhythm’ as a salient feature, three the
creaky voice, two the ‘intonation’ (10.0%), and two the ‘stress’. One of the subjects
suggests that it sounds as if the speaker is not a native speaker of Finnish, and one
mentions the interrupted word as a salient feature. Thus, even if the test subjects’
written descriptions do not give a very clear image of the features of this sample
that they considered ‘atypical’, the acoustic analyses combined with their responses
indicate that the ‘bouncing pitch’ pattern combined with a ‘jerky’ speech rhythm
created by recurrent interruptions were responsible for the perception of atypicality
in this case.
2.2 Prosodic Characteristics of Preadolescent Boys with ASD … 27

30
24
Pitch (semitones re 100 Hz)

12

100 Hz
0

-12
no: se oli oikestaan tartei mä sanon (.) kiinnostava asia koska mua(0.5) jos mä

well it was actually (no meaning) I say (.) interesting thing because I if I

0 5
Time (s)

Fig. 2.4 Example of a disconnected speech rhythm by a boy with ASD

2.2.3 Disconnected Speech Rhythm

The third example concerns a boy with ASD whose utterance was disconnected in
terms of rhythm and syntactic structure. In addition, as Fig. 2.4 shows, he produces
an unclear word, tartei, at the beginning of the extract. This word has no meaning in
Finnish. The fact that the syntactic structure of the utterance is significantly discon-
nected in this sample is a typical feature in the speech of preadolescents with ASD,
and tends to create comprehension problems in their interaction (Wiklund & Laakso,
2020). The extract includes two pauses (after mä sanon, ‘I say’, and koska mua,
‘because I’) and a lengthened syllable (carried by the word no, ‘well’), which create
an impression of disconnected speech rhythm. The first syllable of the word oikestaan
(‘actually’) is also marked by a sudden, remarkably large pitch rise, which is not
typical of Finnish intonation.6 Otherwise, the pitch is rather typical in this example:
the mean f 0 is 179.4 Hz, which is a normal value for a preadolescent boy, and the
standard deviation is 43.7 Hz (or 3.7 st), which falls within the moderate range.
In the perception test, the mean rating for this sample was 3.6, and 20 subjects
also gave a written explanation. Again, it is noteworthy that, according to seven of
them (35.0%), it sounded as if the utterance was produced by someone who was not a
native Finnish speaker, even if all the informants of this study are monolingual Finnish
speakers. Six test subjects (30.0%) mentioned the word ‘unclear’ in their responses,
two (10.0%) wrote that the boy’s ‘pronunciation’ was atypical and two considered
the ‘speech rate’ to be atypical. Interestingly, the speech rate in this example is 4.8
syllables per second, which is very typical of spoken Finnish.7 According to our

6 The size of the pitch rise could not be measured because the syllable carrying it is immediately
preceded by a sequence produced with a creaky voice, breaking the f 0 curve. In addition, at the end
of the extract, the word asia, ‘thing’, is produced with a creaky voice.
7 According to Koskela (2013), the mean speech rate of Finnish-speaking adults is 4.96 syllables

per second.
28 2 Prosody

analyses, perceptions of atypicality, a lack of clarity, and a foreign accent arose from
the disconnected speech rhythm (manifested in pauses, a lengthened syllable, and a
sudden, remarkably large pitch movement), the disconnected syntactic structure of
the utterance, and the occurrence of an unintelligible word.

2.2.4 Large Pitch Excursions

The fourth example is of the speech of a boy with ASD that includes recurring salient
pitch excursions. This phenomenon is reminiscent of the pattern described above (see
Sect. 2.2.1), but in this case the pitch movements—albeit large and salient—are not
‘sing-song’. As Fig. 2.5 shows, the pitch first rises to 11.3 st during the production
of the vocalisation ää (‘er’) and immediately afterwards falls to 14.4 st.8 It rises
again suddenly during the production of the second syllable ([tAin]) of the word
jotain (‘something’), which is atypical given the normal stress in Finnish on the first
syllable of words. In addition, even stressed syllables in Finnish do not typically
carry pitch rises of this size: here the rise is as much as 12.1 st, which is highly
marked in Finnish. The pitch falls immediately afterwards by as much as 9.9 st. It
rises again strongly at the end of the utterance: the size of the rise occurring between
the first ([tA]) and the last ([tu]) syllable of the word tapahtu (‘happened’) is 11.7
st. After a 1.2-s pause, the speaker again produces a vocalisation ää (‘er’) with a
rising pitch. The mean f 0 value in this sample is 198.3 Hz, which is typical for a
preadolescent boy, and the standard deviation is 65.4 Hz (or 5.5 st), corresponding
to a moderate range. The extract also includes one clear grammatical mistake: the
wrong case ending of the word kolmen (‘three’ + a genitive ending).9
In the perception test, the mean rating for this sample was 3.6. In addition, 18
subjects provided a written explanation of their ratings: six of them (33.3%) thought
the speech sounded as if it had been produced by a non-native speaker of Finnish, six
(33.3%) wrote that the ‘speech rhythm’ was atypical, and five (27.8%) found it diffi-
cult to understand what was being said. In addition, three subjects (16.7%) mentioned
‘intonation’ as a salient feature, and three (16.7%) noted a grammatical mistake
(wrong case ending). According to our analyses based on the acoustic features of the
sample as well as on the test subjects’ written explanations, the perception of atypi-
cality here arose from large pitch excursions combined with the grammatical mistake
(wrong case ending). The recurring large pitch movements create the impression not
only of an atypical, non-native-like pitch, but also of a disconnected speech rhythm,
strengthened by the long pause and the two ää vocalisations.

8 The first short part of the f 0 curve occurring above the annotation ‘ää’ in Fig. 2.5 is caused by an
external noise. The same applies to the first two short parts of the f 0 curve at the end of the extract,
above the second ‘ää’ annotation.
9 Wrong case endings and other morpho-syntactic problems are somewhat typical of the speech of

Finnish-speaking preadolescents with ASD (see Chap. 3) (Wiklund & Laakso, 2020).
2.2 Prosodic Characteristics of Preadolescent Boys with ASD … 29

30
24
Pitch (semitones re 100 Hz)

12

100 Hz
0

-12
ää mul on ainakin kolmen viikos jotain tapahtu (1.2) ää

er I have at least in three weeks something happened er

0 5
Time (s)

Fig. 2.5 Example of large pitch excursions by a boy with ASD

30
24
Pitch (semitones re 100 Hz)

12

100 Hz
0

-12
.hhh mut h. muuta en sitten ph. .mth muis- muilta tai en kiinnittänykään niin paljon huo-

.hhh. but h. other things I didn’t then ph. .mth rem- from others or I didn’t pay so much att-

0 5
Time (s)

Fig. 2.6 Example of a flat pitch by a boy with ASD

2.2.5 Flat Pitch

The speech of persons with ASD is often described as monotonous (see, e.g., Nakai
et al., 2014), even if—as the above examples show—this is not always the case. This
last example is of the flat pitch in the speech of a boy with ASD.10 As Fig. 2.6 shows,
the pitch movements are very small during the production of this extract: indeed, the
standard deviation is only 12.9 Hz (or 2.0 st), which corresponds to a small range.
The mean pitch is 114.5 Hz, which is rather low for a preadolescent boy.

10 An example of a flat pitch occurring in the speech of a French-speaking boy with ASD is presented
in Wiklund and Vainio (2019).
30 2 Prosody

The sample also includes salient features other than the flat pitch. It starts with
a long, clearly audible inhalation through the mouth (.hhh), quickly followed by
an audible exhalation through the mouth (h.). There is a clearly audible exhalation
through the nose (ph.) in the middle of the extract, followed by a smack of the lips
(.mth). The speech is also morpho-syntactically disfluent due to the occurrence of
false starts and the disconnected syntactic structure.
In the perception test, the mean rating for this sample was 3.5. In addition, 10 test
subjects gave an explanation for their ratings. Six of them (60%) observed that it was
difficult to understand what was being said, two (20.0%) mentioned the ‘lethargic
articulation’ as a salient feature, two others called attention to the clearly audible
breathing, and two thought that the speaker had a speech disorder, even though none
of the informants of this study had been diagnosed with such a disorder. Only one
subject mentioned ‘flatness’ as a salient feature. As noted above, small pitch intervals
and narrow variations in pitch movements are typical of Finnish intonation (Iivonen,
1998). Thus, a ‘flat pitch’ may not be as easily considered salient in Finnish as in
a number of other languages. The perception of atypicality in this case could be
attributed to a combination of features: a flat pitch associated with morpho-syntactic
disfluency, lethargic articulation, and recurrent clearly audible breathing.

2.3 The Role of Prosodic Features in the Creation

of Comprehension Problems

As already mentioned, in the studies presented in this book, comprehension problems

are materialised by other-initiated repairs in which the repair process is started by
the recipient of the problematic utterance (Schegloff et al., 1977; SIL Glossary of
Linguistic Terms, 2021). Comprehension problems have been studied mainly through
the Finnish data of the project (see Sect. 1.2.1) (Lehtinen, 2012; Wiklund & Laakso,
2019; Wiklund & Stevanovic, 2018; Wiklund & Vainio, 2019; Wiklund, 2016). The
Finnish data include 72 other-initiated repairs. Most (86.1%) of the other-initiated
repairs occurring in the data are repair initiations in which a repair procedure is started
by an interlocutor, but the outcome of the repair is left to the speaker who produced
the trouble-source turn. The remaining cases (13.9%) are outright corrections in
which the interlocutor not only indicates that something said earlier needs to be
corrected, complemented, or explained, but also presents another version (which he
or she finds more correct or more complete) of the contents of the trouble-source
turn (Schegloff et al., 1977; Sorjonen, 1997). Most of the repair sequences of the
data occur between the therapists and the boys with ASD (84.7%). Sometimes, repair
sequences also occur between the boys (11.1%), even occasionally between the two
therapists (4.2%).
A large majority of the repair initiations are produced by one of the therapists
(80.6%), whereas half of the outright corrections are produced by one of the boys
(Wiklund & Stevanovic, 2018; Wiklund, 2016). The fact that most of the repair
2.3 The Role of Prosodic Features in the Creation of Comprehension Problems 31

initiations are produced by one of the therapists and not by one of the boys might
be related to the educational role of the therapists. In other words, the fact that the
therapists frequently produce repair initiations might be related to the fact that they
are probably engaged in encouraging the boys with ASD to articulate things more
clearly. Thus, in these data, all the repair sequences that occur do not necessarily
correspond clearly to a lack of comprehension or hearing; some of them may play
an educational role (Wiklund & Stevanovic, 2018; Wiklund, 2016).
Concerning the types of repair initiations, the data include both direct repair
initiations and candidate understandings. In direct repair initiations, an interlocutor
indicates that there is a need for repair (for example by saying: “What?” or “Sorry,
I didn’t understand.”) but does not offer a possible solution. In candidate under-
standings, an interlocutor not only indicates that there is a need for repair in the
interaction, but also offers a possible interpretation of the prior speaking turn (Sche-
gloff et al., 1977; Sorjonen, 1997). According to a study by Leskelä (2012), candidate
understandings are often used to resolve comprehension problems between mentally
disabled persons and their therapists. Kurhila (2012), in turn, has studied repair
initiations in conversations between native and non-native Finnish speakers. In her
data as well, candidate understandings are much more frequent than direct repair
initiations. Interestingly, in the data from group therapy sessions involving pread-
olescent boys with ASD, direct repair initiations are much more frequent (90.3%)
than candidate understandings (9.7%) (Wiklund, 2016). At first, this might seem a
bit surprising, because the data with participants with ASD are similar to Leskelä’s
and Kurhila’s data in the sense that all these data sets consist of ‘asymmetrical’ inter-
actions in which one party is linguistically more competent than the other. However,
the fact that the therapists of the group therapy sessions with participants with ASD
mainly use direct repair initiations instead of candidate understandings is in line
with the instructions for easy language designated for spoken communication with
persons who have lowered interaction skills (Kartio, 2009). Indeed, according to the
easy language instructions, candidate understandings incur a risk of so-called ‘yea-
saying’. In other words, especially in asymmetrical conversations, there is a risk that
the linguistically less competent participant who has produced a trouble-source turn
(that is, a ‘problematic’ speaking turn creating a repair sequence) accepts a candidate
understanding suggested by the more competent participant of the interaction even
if the contents of the candidate understanding do not correspond to the intended
meaning of the trouble-source turn (Kartio, 2009; Wiklund, 2016).
Table 2.1 presents the frequencies of different prosodic features occurring in
trouble-source turns produced by the Finnish-speaking boys with ASD.11 As the
table shows, creaky voice (35.3%) and quiet voice (31.4%) are the most common
features in these data. Pitch excursions (23.5%) are also relatively frequent.
The fact that a creaky voice is the most frequently occurring prosodic feature
in the trouble-source turns of the data probably reflects the fact that this feature is,
generally speaking, very common in spoken Finnish. According to Ogden (2001), a
creaky voice is often used as an interactional cue indicating the end of a speaking

11 The contents of Table 2.3 have been published in Wiklund (2016).

32 2 Prosody

Table 2.1 Occurrences of

Prosodic feature Occurrences in trouble-source turns
prosodic features in
(%)
trouble-source turns produced
by the Finnish-speaking boys Creaky voice 35.3
with ASD Quiet voice 31.4
Pitch excursions 23.5
Stretched syllables 17.6
Jerky speech rhythm 15.7
Bouncing pitch 11.8
Accelerated speech 11.8
Slowed speech 7.8
Loud speech 7.8
Changes in volume 3.9
Change of speech rate 3.9
Nasal voice 3.9
Lack of pauses 1.9
Growling voice 1.9

turn in Finnish talk-in-interaction. It is also common in Finnish speech that a whole

utterance is produced mostly, or even entirely, with a creaky voice (Aho, 2010).
In the data with participants with ASD, a creaky voice does not seem to cause
comprehension problems (Wiklund, 2016), and it does not cause a perception of
atypicality (Wiklund et al., forthcoming). Nevertheless, it is a feature that rather
often occurs in trouble-source turns.
Example 2.1 presents an extract in which the last word of a trouble-source turn
is produced with a creaky voice, and part of the word is also produced with a quiet
voice. The extract comes from the session of Group A. The boys have shared their
news, and now the group is discussing bullying at school with the help of a series
of drawn pictures. In the picture at hand, a little girl is being bullied by a group
of bigger girls. The male therapist (‘MT’) asks what could follow when the little
girl goes to talk to her friends and older siblings (lines 01–03). Heikki immediately
gives an answer (no sitten noille kävis huonosti, “well then things would go badly
for them”, line 05). The therapist does not understand Heikki’s answer and reacts to
his turn with the direct repair initiation mitä se tarkoittaa (“what does it mean”, line
07).
2.3 The Role of Prosodic Features in the Creation of Comprehension Problems 33

Example 2.1:

01 MT: .mthhh mitäs te luu #lette että# kun se menee, (.) .hh

what do you think when she goes

02 kave reilleen tai s- (.) #m-# vanhemmille sisa ruksilleen

to her friends or to her older siblings

03 #kertoon niin#, (.) mitäs siitä vois #seurata#.

to tell so what could follow from it

04 (.)

05 Heikki: no sitten noille kävis #huo nosti #.

well then things would go badly for them

06 (0.8)

07 MT: mitä se #tarkottaa#. (.)

what does it mean

08 Heikki: no ainaki:, mhhh (.) .hh no, mhhh

well at least well

09 Rikhard: mhhhhhh

10 (0.4)

11 Heikki: no=ehkä? (0.8) no ehkä se, mhhhh no ehkä nei

well maybe well maybe she well maybe they wouldn’t

12 varmaa, (0.4) .hhhh jos sill=ois vaikka vanhempi

probably if she had for example an older

13 sisko ni ehkä sitte t'ta, (0.8) ehkä sitte ö- 0.6)

sister maybe then like maybe then er

14 öö, (.) mhhhhh .mt (0.6) ehkä s' sitte ottas, (.) öö,

er maybe t- then she would grab

15 täl lee niit molempii; (.) molempii kiinni >ha han'h

like this both of them would grab both of them like

16 että:,< (.) .hh ette enää KOS kes siihen tai muute käy

like you don’t touch her anymore or else things will go

34 2 Prosody

17 huo nommin . ((kats. tätiä))

worse for you

18 (.)

19 MT: [ mm ] m?

20 FT: [ mm, ]

Figure 2.7 illustrates the acoustic shape of the trouble-source turn (line 5)
including a word produced with a creaky voice.
As Fig. 2.7 shows, the word sitten (‘then’) carries a raised pitch, and the pitch
level falls immediately after during the production of the word noille (‘to them’).
The last word of the utterance, huonosti (‘badly’), is produced with a creaky voice
on a low pitch level. The first syllable, [huo], is stressed. In addition, the last two
syllables, [nos] and [ti], are produced with a quiet voice. These prosodic features do
not impair the intelligibility of the utterance and thus do not cause the comprehension
problem occurring in this example. The comprehension problem is rather related to
the fact that Heikki’s comment (“well then things would go badly for them”, line
05) is surprising. The male therapist probably expects another kind of answer to his

0.7994

-0.5605
0 2.45
Time (s)

30
Pitch (semitones re 100 Hz)

24

12

100 Hz
0

-12
no sitten noille kävis huonosti

’well then things would go badly for them’

0 2.45
Time (s)

Fig. 2.7 Acoustic shape of the trouble-source turn including a word produced with a creaky voice
2.4 The Use of Prosody in Questions and Statements 35

question (“what could follow from it”, line 03), and therefore he finds it difficult to
understand Heikki’s turn (line 05). The male therapist reacts to Heikki’s turn with
the direct repair initiation mitä se tarkoittaa (“what does it mean”, line 07). Heikki
reacts to the therapist’s repair initiation by explaining what he meant in the trouble-
source turn (lines 8 and 11–17). His turns are disfluent, but he is able to resolve
the comprehension problem with his turns: both therapists react to his latter turn
with the dialogue particle mm (lines 19 and 20), which indicates here that they have
heard and understood what has been said (Hakulinen et al., 2004: § 798), but they do
not take a stance towards the contents (in other words, their reactions do not imply
that they necessarily agree with the boy). It is noteworthy that Heikki is able to reply
adequately to the therapist’s repair initiation and resolve the comprehension problem
with his own action. This is in line with Volden’s (2004) results, according to which
the repair abilities of children with ASD are in many respects similar to those of
children without ASD.
Even if trouble-source turns produced by the boys with ASD often have certain
prosodic features (such as a creaky and/or a quiet voice and/or large pitch excursions),
it is noteworthy that in very few cases do prosodic features seem to be the main cause
of the comprehension problem (Lehtinen, 2012; Wiklund, 2016). However, prosodic
features may contribute to the creation of comprehension problems in spontaneous
interaction between preadolescent boys with ASD and their neurotypical therapists.

2.4 The Use of Prosody in Questions and Statements

How persons with ASD use prosodic features in spontaneous interaction is little
studied. Indeed, most previous studies on the prosody of persons with ASD have been
either purely phonetic investigations or neurophysiological studies using mismatch
negativity experiments. How prosody is used in spontaneous interaction has been
studied in the Finnish data of the project (see Sect. 1.2.1) (Wiklund et al., 2021).
More precisely, the studies carried out within the framework of this project have
focused on the use of pitch in questions and statements from a qualitative point of
view. The objective has been to examine the use of rising pitch in questions and state-
ments compared with what is found in the previous research literature. The research
questions were the following: (1) Do speaking turns of boys with ASD recurrently
carry a rising pitch contour? (2) If rising pitch occurs, does its use correspond to
what is described in previous research literature concerning Finnish? (3) Do the
participants use other prosodic features as interactional resources? (Wiklund et al.,
2021).
36 2 Prosody

According to Iivonen (1998: 317), a falling pitch contour concerns both questions
and statements in Finnish, generally speaking. The final pitch level is the same in
questions and statements: it is low in both cases (Iivonen, 2009: 73). Unvoicing and
a creaky voice are common in Finnish at the end of utterances (Iivonen, 2006, 2009;
Ogden, 2001). The initial pitch level is high in questions, whereas in statements, it
is considerably lower. The pitch range of statements is narrow, whereas in questions
it is very wide (Iivonen, 2009: 73). The global shape of statements is descending.
In questions, the pitch rises until the last word and then falls. The stress is weak in
statements, whereas in questions, the last word is strongly stressed (Iivonen, 2009:
73).
Unlike in many other languages, an utterance-final pitch rise has not grammat-
icalised as a sign of interrogativity in Finnish (Kallioinen, 1968; Hirvonen, 1970;
Iivonen, 1978, 1998; Routarinne, 2003: 179–183). Rising pitch has traditionally
been considered problematic in studies focused on Finnish intonation. According
to the established view, final pitch rises are rare, and they do not occur systemat-
ically (Iivonen, 1998, 2001). Utterances can, however, also end in a pitch rise in
Finnish (Iivonen, 1978), and in spontaneous speech, intonation contours are, gener-
ally speaking, variable (Iivonen et al., 1987). Ogden and Routarinne (2005: 160) also
point out that most research on Finnish intonation has been based on monologous or
read-aloud speech, and that in spontaneous interaction utterance-final pitch rises are
actually common.
Iivonen (1998: 326) and Anttila (2008) have found that pitch rises are common,
especially at the end of tag questions and elliptic questions in Finnish. Anttila’s
(2009) study based on spontaneous speech revealed that in fact final pitch rises are
common in various types of questions. The author mentions that this could be caused
for example by interference from other languages, such as English (Anttila, 2009:
176), or it could be a ‘new manner’ to appeal to the listener (Anttila, 2009: 176) or
to express politeness (Anttila, 2008: 90).
According to Iivonen (1978), utterance-final pitch rises often indicate continuity
in Finnish. The findings of Routarinne (2003) concerning the speech of young girls
are similar. The study reveals that utterance-final pitch rises typically occur in narra-
tive sequences and indicate that the speaker intends to continue speaking (see also
Lehtinen, 2010b). The phenomenon implies that the line of narration will continue,
and it often occurs in utterances that direct the attention of the listener towards the
continuation of the story. An utterance-final pitch rise is also an interactional resource
that calls for the reaction of the interlocutor (Routarinne, 2003: 170) and indicates in
this manner that the speaker is taking the interlocutor into account (p. 183). Indeed,
Ogden and Routarinne (2005: 173) have shown that utterance-final pitch rises often
lead to a minimal response (such as ‘mm’) with the help of which the interlocutor
can mark himself or herself as the recipient of what is being said.
The findings of Routarinne (2003: 169) show that the majority (74%) of utterances
marked with an utterance-final pitch rise are statements of parts of them. Utterances
ending in a pitch rise generally describe a state of affairs that is expected to be shared
by the interlocutors. These utterances typically occur in orientative parts of the story
in which the speaker describes the circumstances of the actual events (Routarinne,
2.4 The Use of Prosody in Questions and Statements 37

2003: 183–188). The final pitch rise indicates where the background information
should be related (Routarinne, 2003: 187). In other words, a final pitch rise projects
the continuation of the line of narration as well as a climax or another continuation,
for the understanding of which the information included in the utterance marked with
the final pitch rise is particularly important.
As already mentioned (see Sect. 2.1), Le Normand et al. (2008) have studied
prominence and prosodic contours related to speech acts among French-speaking
children with ASD. Their data came from recorded free play sessions during which
8 children with ASD (6 boys and 2 girls) played with a speech pathologist. The
ages of the children ranged from 4 to 6 years, and they were assessed as being in
the mild-moderate range of the autism spectrum. The findings of the study revealed
that the children with ASD had difficulties producing prosodic prominence: low
prominence was highly frequent in their speech compared with typically developing
children, which seems to point to more monotonous prosodic patterns. This feature
is further strengthened by the fact that the children with ASD also produced many
words with flat contours. On the communicative level, the children with ASD also
had difficulties in matching the form and function of the prosodic contours. Declar-
atives were frequently produced with a flat contour. Exclamations, in turn, were less
atypical; they consisted mainly of screams out of pitch range. Concerning questions,
no clear pattern of prosodic contours emerged in the study, which could have been
attributable to the low number of questions produced by the children with ASD in
the data studied by the authors (Le Normand et al., 2008).
As already mentioned above, rising pitch has previously been studied using
Conversation Analysis (CA) in Finnish data in the speech of young girls (Ogden &
Routarinne, 2005; Routarinne, 2003). As all the informants of these data are young
boys, in addition to contributing to the field of research on ASD, the study also
contributes new and interesting information on Finnish prosody in general (Wiklund
et al., 2021).
The data were studied both quantitatively and qualitatively. First, the total number
of speaking turns of each boy was counted. After that, the total number of speaking
turns ending in a pitch rise was counted. After these calculations, the speaking turns
were analysed in their contexts of occurrences using methods of CA. The objective
of the qualitative analyses consisted of finding out why a rising pitch occurred in
each context. In addition to pitch, we also considered other prosodic features, such
as voice quality. After the qualitative analyses, we counted the number of speaking
turns in which a rising pitch occurred in a question, and the number of other types of
speaking turns in which a rising pitch occurred. This made it possible to discover if a
rising pitch was used in the data only in questions or also in other types of speaking
turns (Wiklund et al., 2021).
Table 2.2 presents the total number of speaking turns for each speaker as well as
the number of turns produced with a final pitch rise.12

12
The contents of Tables 4 and 5 have been published in Finnish with different pseudonyms in
Wiklund, Ihaksinen and Vainio (2021).
38 2 Prosody

Table 2.2 Total number of speaking turns and turns ending in a pitch rise in the Finnish data on
participants with ASD
Number of turns ending in a rising pitch Total number of turns Rising pitch (%)
Heikki 73 215 34.0
Mikael 30 157 19.1
Rikhard 4 167 2.4
Toni 23 105 21.9
Kalle 21 122 17.2
Jaakko 41 216 19.0
Teppo 25 98 25.5
Total 217 1080 20.1

Table 2.3 Total number of question turns and question turns ending in a pitch rise in the Finnish
data with participants with ASD
Number of question turns ending Total number of question turns Rising pitch (%)
in a rising pitch
Heikki 10 22 45.5
Mikael 1 6 16.7
Rikhard 1 3 33.3
Toni 3 4 75.0
Kalle 3 5 60.0
Jaakko 1 4 25.0
Teppo 3 4 75.0
Total 22 48 45.8

As Table 2.2 shows, there is a lot of individual variation between the participants:
in the speech of one of the boys (‘Heikki’), the proportion of speaking turns ending
in a pitch rise is as much as 34.0%, whereas in the speech of another boy (‘Rikhard’),
the proportion is only 2.4%. On average, 20.1% of the speaking turns occurring in
the speech of the boys with ASD end on a rising pitch (Wiklund et al., 2021). Thus,
the majority (79.9%) of their turns end on a falling pitch or on a flat pitch. It is
however noteworthy that occurrences of final pitch rises are recurrent in the data,
which is in line with the study of Ogden and Routarinne (2005: 160), according to
which utterance-final pitch rises are common in spontaneous Finnish speech.
The total number of question turns for each speaker as well as the number of
question turns produced with a final pitch rise can be found in Table 2.3.
The variation between individuals is remarkable also with regard to the propor-
tions of question turns ending in a pitch rise in these data. As Table 2.3 shows, the
proportion of question turns ending in rising pitch is 75.0% in the speech of two boys
(‘Toni’ and ‘Teppo’), whereas in the speech of one boy (‘Mikael’), the proportion
is only 16.7%. On average, 45.8% of the question turns end on a pitch rise in the
2.4 The Use of Prosody in Questions and Statements 39

speech of the Finnish-speaking boys with ASD (Wiklund et al., 2021). This result
is in line with the findings of Anttila (2009), according to which rising pitch occurs
recurrently at the end of various types of questions in spontaneous Finnish speech. It
is however noteworthy that the total number of question turns occurring in the data
is relatively modest (48). Thus, the results of the study cannot be generalised.
The term ‘question’ refers here to an utterance which belongs to the epistemic
field of the recipient, and which forms the first part of an adjacency pair consisting
of a question and a response (Huhtamäki, 2015: 18–22). An ‘utterance’ refers to
a prosodic and actional unit which has a certain function, and which represents a
conversational action of some kind. A speaking turn includes at least one utterance.
Grammatically, an utterance can be a word, phrase, clause, etc. (Hakulinen et al.,
2004: § 1003). The term ‘statement’ is used here in a very broad sense to refer
to all types of speaking turns other than question turns. In the phonetic analyses,
changes in the fundamental frequency (f 0 ) were measured, as well as the f 0 mean,
the minimum f 0 , and the maximum f 0 during the production of the speaking turn.
The phonetic analyses were carried out with the help of the speech analysis program
Praat (Boersma & Weenink, 2021). The f 0 curves have been verified and corrected
manually with the help of the spectrogram. For example, the mistakes in the curve
caused by a creaky voice have been corrected or explained in the text. Semitones
(st) have been used to describe the changes, because a semitone scale corresponds
better to auditory perception than a hertz scale, and using semitones also makes it
possible to compare different speakers. A ‘rising pitch’ was defined on the basis of
the auditory perception of the utterance: if the perception of an utterance-final pitch
movement was clearly rising, it was considered a ‘rising pitch’. Thus, a specific
threshold value was not used.
The second example illustrates question turns that end on a rising pitch.13 There
are three turns of this type (lines 04, 14, and 24) in this extract. In addition, the extract
includes other speaking turns that end on a rising pitch. The example comes from
the beginning of the session of Group B. Jaakko has just shared his news, and Kalle
says that he can start asking questions (line 01). The turn is a pre-start with which
Kalle volunteers as the next speaker (Sacks et al., 1974). The turn ends on a slightly
rising pitch (,?).

13The same extract has been analysed in Finnish using different pseudonyms in Wiklund, Ihaksinen
and Vainio (2021: 60–64).
40 2 Prosody

Example 2.2:

01 Kalle: mä voin alottaa kysymykset,?

I can start the questions

02 MT: m-h?

03 (0.7)

04 Kalle: minkä elokuvan kävit kattomassa ?

which movie did you go to see

05 Jaakko: polv- (.) poutapil viä (.) ja lihapullakuuroja.

clo- Cloudy with a Chance of Meatballs

06 ((MT nods))

07 (2.1)

08 Kalle: <ookoo;>

ok

09 Toni: no olik[se hyvä;]

well was it good

10 Jaakko: [sitä]hän ei (.) periaattees koskaan tapahtunu

in principle it never happened

11 >mut siinä oli lihapullia.< ((shows the shape of a

but there were meatballs

12 meatball with his fingers))

13 (1.0)

14 Toni: no m- mitä sä teit koneella,?

well w- what did you do with the computer

15 Jaakko: nää (.) mä vaan katoin videoita peleistä? (1.7) yks

x I just watched videos of games

16 erit täin hyvä tulikin (0.4) #viime viikol tais

a very good one was published was it last week

17 olla#,? (1.4) oon nyt kiirees- (.) ti kattonu sen

2.4 The Use of Prosody in Questions and Statements 41

I have now quickly watched it

18 (.)kuinka se on vedetty (.) l:äpi ((shows quotation marks

how it has been played through

19 with his fingers) (0.4) plus jotakin netissä ja #tänään

plus something on the internet and today

20 oon just viimeistelly#,?

I have just finalised

21 (2.7)

22 FT: ni-hi;

yeah

23 (1.0)

24 Teppo: no: (.) pidätsä luistelusta?

well do you like ice skating

25 Jaakko: e:n (1.4) mä vihaan #oikeestaan sitä# (1.0) ((Jaakko

no I actually hate it

26 blows out))

The male therapist (’MT’) reacts to Kalle’s pre-start with the particle turn ‘m-h’
(line 02) ending in a rising pitch. The rising pitch at the end of the pre-start indicates
that Kalle expects the therapist who is in charge of leading the conversation to react
to his turn. Thus, it indicates that the speaker seeks a reaction on the part of the
recipient, which is typical of utterance-final pitch rises in Finnish (Routarinne, 2003:
170). In addition, the final pitch rise can be related to the indication of continuity in
this context: on the one hand, Kalle looks for a reaction on the part of the recipient,
but on the other hand, Kalle’s turn also implies that he will continue talking himself
after the reaction he is seeking.
After a pause, Kalle presents his actual question, “which movie did you go to
see” (line 04). The last word, kattomassa (‘to see’) has been produced with a quieter
voice than the preceding speech, which implies that the turn is ending. However, the
question ends in a rising pitch. The pitch rise can be related to the fact that the question
occurs in a ‘queue of questions’. In other words, the speaker knows that each boy will
ask one question, and he is himself the first one to ask. It is also possible that even
if an utterance-final pitch rise has not grammaticalised as a sign of interrogativity in
Finnish (Kallioinen, 1968; Hirvonen, 1970; Iivonen, 1978, 1998; Routarinne, 2003:
179–183), a rising pitch emphasises the interrogativity of a question and appeals to
the listener more strongly than a question ending in a flat or falling pitch in certain
contexts. These contexts may include certain institutional contexts, such as the one in
42 2 Prosody

this study, where asking a question is an action that is predetermined by the situation.
In this manner, the speaker can emphasise that he is acting as required by the situation
and asking a question on his own turn. Nevertheless, as mentioned above, in these
data only 45.8% of the questions produced by the boys with ASD end in a pitch
rise. Thus, the phenomenon does not occur systematically or even dominantly in
these data. The proportion is, however, rather large, when one takes into account the
fact that all questions do not end in a rising pitch even in such languages as French
(Delais-Roussarie et al., 2015).
Jaakko answers Kalle’s question: “clo-Cloudy with a Chance of Meatballs” (line
05). The response is prosodically interesting, because the turn includes large pitch
excursions: first, the first (and only) syllable of the interrupted word polv- (~ ‘clo-’)
is produced at a remarkably high pitch level. The interrupted word is followed by a
micropause after which the first syllable of the word poutapilviä (~ ‘Cloudy’) is again
produced at a remarkably high pitch level. The end of the second part of the compound
(-viä) is, however, produced at a remarkably low pitch level. These changes create
a sing-song pitch. Such large pitch excursions are typical of this speaker across the
whole data (Wiklund & Vainio, 2019). The male therapist reacts to Jaakko’s turn by
nodding (line 06), after which there is a 2.1 s-long pause in the conversation (line
07). After this, Kalle says ‘ok’ with a slow speech rate and indicates in this manner
that he has heard and understood Jaakko’s turn (line 08).
After Kalle’s response, Toni asks a spontaneous follow-up question, “well was it
good” (line 09). This question ends in a slightly falling pitch. In this context, the fact
that the question ends in a falling pitch can be related to the fact that it is a follow-up
question. Thus, Toni does not present his question in the ‘queue of questions’ as an
independent question but as a follow-up question related to Kalle’s question. Jaakko
starts to talk in overlap with Toni and gives more information about the movie he
saw (lines 10–11). He does not answer Toni’s question but explains his own previous
turn (line 05).
After a 1.0 s-long pause, Toni presents his actual question: “well w- what did you
do with the computer” (line 14). This question ends in a slightly rising pitch. The
turn starts with the dialogue particle no (‘well’), which indicates here two types of
transitions (Hakulinen et al., 2004: § 1036): on the one hand, a topical transition
from the ‘movie’ to the ‘computer’, and on the other hand, a sequential transition
from Kalle’s question and the discussion related to it to Toni’s question turn. On the
one hand, the rising pitch here can be related to the fact that the question occurs in
a position predetermined by the institutional situation (i.e., as one question in the
‘queue of questions’). On the other hand, the occurrence of the rising pitch can be
related to the ‘recycling’ of pitch. In other words, the speaker places his question as
a part of the ‘queue of questions’ by recycling the rising pitch used by Kalle in his
previous turn (cf. for example Szczepek Reed, 2009).
Jaakko gives a long response to Toni’s question (lines 15–20). This turn also
includes large pitch rises and falls at the beginning and in the middle of words (line
16: erittäin, ‘very’; line 17: nyt, ‘now’; line 19: netissä, ‘on the internet’). In addition,
the turn includes several utterance-final pitch rises (lines 15, 17, 20). A creaky voice
2.4 The Use of Prosody in Questions and Statements 43

occurs on lines 16–17 (“was it last week”) as well as at the end of the turn (lines 19–
20, “and today I have just finalised”). The beginnings of certain words carry strong
stress (line 18: läpi, ‘through’; line 19: netissä, ‘on the internet’; line 20: viimeistelly,
‘finalised’). In this turn, Jaakko uses stress to emphasise words and a creaky voice
as a prosodic device indicating the end of the turn. However, the large number of
prosodic changes inside the same turn creates an impression of disfluency in his
speech (Wiklund & Laakso, 2019, 2020).
During Jaakko’s turn (lines 15–20), the first utterance ending in a pitch rise (“I
just watched videos of games”, line 15) is orientative: it orients the recipients to
the continuation of the line of narration and indicates where the continuation should
be related (Routarinne, 2003). The second turn ending in a pitch rise (“was it last
week”, lines 16–17) gives background information concerning the actual events being
described in the turn (Routarinne, 2003: 183–188). The last utterance that ends in
a rising pitch (“and today I have just finalised”, lines 19–20) calls for the reaction
of the interlocutors (Routarinne, 2003). In addition, the last utterance is produced
with a creaky voice, which implies in Finnish that the end of the turn is approaching
(Ogden, 2001). It is noteworthy that Jaakko shows with the recurrent utterance-final
pitch rises that he takes his interlocutors into account (Routarinne, 2003: 183).
Jaakko’s turn is followed by a 2.7 s-long pause (line 21), after which the female
therapist produces a minimal response (‘ni-hi’, line 22) with a slightly falling pitch.
After this, there is a 1.0-s-long pause (line 23), followed by Teppo’s question “well
do you like ice skating” (line 24). Teppo’s turn also starts with the dialogue particle
no (‘well’) that indicates transition (Hakulinen et al., 2004: § 1036). The dialogue
particle is lengthened. The turn ends in a rising pitch (see Fig. 2.8). The transition
indicated by the particle no (‘well’) is both topical and sequential, as in the case of
Toni’s turn (line 14). Here too, the final pitch rise is probably related to the fact that
the question occurs in the ‘queue of questions’ predetermined by the institutional
situation, as well as to the ‘recycling’ of prosody—in other words, to the fact that
two previous speakers have produced their questions with a final pitch rise. Jaakko
answers “no I actually hate it” (line 25) and blows out (lines 25–26). The first word of
the turn, en (‘no’), is strongly stressed, and it is followed by a 1.4 s-long pause. Also,
the word vihaan (‘I hate’) starts with a strong stress. The end of the turn, oikeestaan
sitä (‘actually it’), is again produced with a creaky voice. Thus, in this turn as well,
Jaakko uses prosodic features to emphasise words as well as to indicate the end of
the turn.
Figure 2.8 shows the changes of pitch (fundamental frequency, f 0 ) during Teppo’s
question turn “well do you like ice skating” (line 24). The mean pitch during the
production of the question is 216.9 Hz. The minimum pitch is 169.8 Hz, and the
maximum pitch is 256.4 Hz. The standard deviation of the f 0 is 1.6 semitones (st),
which is a low value. In other words, the pitch is generally monotonous in this
speaking turn, even if it ends on a rising pitch. The size of the rising pitch movement
is 5.6 st (measured between the penultimate syllable, [lus], and the last syllable, [ta]).
A rise of this size is perceived as a clearly rising pitch movement.
As Example 2.2 shows, the Finnish-speaking boys with ASD in this study are
able to ask relevant questions about the topic that is being discussed, as well as
44 2 Prosody

0.02307

-0.03339
0 1.814
Time (s)

30
24
Pitch (semitones re 100 Hz)

12

100 Hz
0

-12
no (.) pidät sä luistelusta

’well do you like ice skating’

0 1.814
Time (s)

Fig. 2.8 Question turn ending on a pitch rise produced by a Finnish-speaking boy with ASD

answer them (Kremer-Sadlik, 2001; Wiklund, 2016; Wiklund et al., 2021; Wiklund &
Määttä, 2021). One of the boys also shows in this example, with recurring final pitch
rises, that he takes his interlocutors into account (Routarinne, 2003: 183). This is
remarkable because, as already mentioned, reciprocal interaction and taking the
interlocutor’s perspective into account are generally difficult for people with ASD
(APA, 2013). The analyses also strengthen the findings of previous studies, according
to which children with ASD are able to produce narrative sequences in conversations
(Solomon, 2004; Wiklund, 2012). In addition, as Example 2.2 shows, the boys with
ASD in this study are able to emphasise words with prosody as well as to use and
interpret prosodic features indicating the end of a turn. The analyses also provide more
evidence that preadolescents with ASD are able to correctly interpret the interactional
meanings of combinations of dialogue particles and prosodic features (Wiklund &
Stevanovic, 2018; Wiklund & Vainio, 2019; Wiklund, 2012). The use of prosody
as an interactional resource could possibly be practiced even more in therapy for
children and teenagers with ASD. Indeed, at least in these data, preadolescents with
ASD seem to be able to use prosody as an interactional resource; thus, this ability
could be strengthened more in the future through conscious training (Wiklund et al.,
2021).
2.5 Conclusion 45

2.5 Conclusion

The results of the study by Wiklund et al. (forthcoming) show that it can be proved
both experimentally and statistically that neurotypical Finnish-speaking adults find
the prosody of preadolescent, Finnish-speaking boys with ASD more atypical than
the prosody of age- and gender-matched controls. In general, the speech samples of
the ASD individuals were rated as significantly more atypical than the samples of
the control group. It is, however, noteworthy that the sample size is small, which is a
limitation of the study. However, the data come from naturally occurring, spontaneous
interactions. The data are also in Finnish, whereas most previous studies concerning
the prosody of persons with ASD are based on English data. Therefore, the study
provides valuable new information even if the sample size is small. In the future, the
study could be replicated with a larger sample size.
The findings of Wiklund et al. (forthcoming) concerning the perceived atypicality
of speech prosody among children with ASD are in line with the results obtained
by Peppé et al. (2006) and Redford et al. (2018), in the case of English. Thus,
given that Finnish and English are distant languages both prosodically and typo-
logically, the results provide preliminary evidence for the possible universality of
the phenomenon. Naturally, more languages should be studied from this perspective
to support the universality hypothesis. According to the results of Wiklund et al.
(forthcoming), potential causes of the perception of atypicality include the following
prosodic features: sing-song pitch, bouncing pitch, disconnected speech rhythm,
large pitch excursions, and flat pitch. It is noteworthy that these features surprisingly
often gave the impression of a non-native accent. As mentioned, Finnish intonation
is, generally speaking, characterised by small pitch intervals and a narrow range
of pitch movements (Iivonen, 1998), hence a sing-song pitch and large pitch excur-
sions create a strong impression of atypicality, whereas a flat pitch is not perceived as
salient. In addition to the above-mentioned prosodic features, occurrences of morpho-
syntactic problems (such as wrong case endings and a disconnected syntactic struc-
ture), unintelligible words, and lethargic articulation may strengthen the impression
of atypicality. Creaky voice, which was common in the speech samples, did not
attract the attention of the test subjects, which probably relates to the fact that creaky
voice is very common in spoken Finnish (Aho, 2010; Ogden, 2001). The test subjects
frequently found it difficult to understand what the informants with ASD were saying
in the speech samples,14 which highlights the importance of addressing prosody and
other expressive language-related difficulties in these individuals.
The stimuli assessed as atypical by the neurotypical test subjects included both
large and small pitch ranges (albeit some speakers exhibited a moderate range). This
finding is in line with previous studies reporting a tendency among children with
ASD to have a very narrow or a very wide pitch range (Baltaxe, 1984; Wiklund &
Vainio, 2019). Diehl and Paul (2013), Fosnot and Jun (1999) as well as Nadig and
Shaw (2012) have also reported a wide range in the speech of children with ASD.

14 Naturally, other features of speech than prosodic features also contribute to the creation of
intelligibility problems in these data.
46 2 Prosody

Lindström’s (2019) results based on Finnish data show evidence of impairment in

the processing of natural speech prosody among children with ASD on various levels
of information processing, including the aberrant discrimination of prosodic changes.
Thus, the occurrence of atypical prosodic features on the level of prosodic production
could relate to the fact that children with ASD find it difficult to self-monitor prosodic
changes on the level of perception.
The study by Wiklund et al. (forthcoming) also has clinical implications. Interven-
tion to remedy atypical prosody among persons with ASD is not yet systematically
organised, and research is limited (Holbrook & Israelsen, 2020). It is worth pointing
out that, having studied the comprehension of problems and the supporting of social
interaction skills in group therapy sessions involving Finnish-speaking preadoles-
cents with ASD, Wiklund and Stevanovic (2018) (see Sect. 6.2) found that prosodic
skills are often overlooked, and that the focus tends, implicitly or explicitly, to be
on the general norms of social interaction and embodied social interaction. Thus,
the results of Wiklund et al. (forthcoming) could increase awareness of prosodic
deficiencies among preadolescents with ASD, and therefore be of use in developing
therapeutic practices. The examples that were given in this chapter show that atyp-
ical prosodic features can be acoustically characterised, which supports the use of
technology-based instruction in intervention. Moreover, accurate characterisation of
the prosody profiles could facilitate better targeting of interventions.
This chapter has also discussed the role of prosodic features in the creation of
comprehension problems (Wiklund, 2016), materialised by other-initiated repairs in
which the repair process is started by the recipient of the problematic utterance (Sche-
gloff et al., 1977; SIL Glossary of Linguistic Terms, 2021). It was stated that most of
the other-initiated repairs (86.1%) in the data were ‘repair initiations’, whereby the
repair procedure is started by a participant, but the outcome of the repair is left to the
speaker who produced the trouble-source turn. The remaining cases (13.9%) were
‘outright corrections’ (Schegloff et al., 1977; Sorjonen, 1997). Most of the repair
sequences occurred between the therapists and the boys (84.7%). A large majority
of the repair initiations were produced by the therapists (80.6%), whereas in half the
cases the outright corrections were produced by the boys. The fact that most of the
repair initiations were produced by one of the therapists and not by one of the boys
might be related to the educational role of the therapists: that is, to the fact that the
therapists were probably engaged in encouraging the participants to articulate things
more clearly. Thus, all the repairs occurring in these data may not indicate a break-
down of intersubjectivity: some of them may have an educational role (Wiklund,
2016).
Concerning the types of repair initiations, it was shown in this chapter that the
data included both ‘direct repair initiations’ and ‘candidate understandings’—that
is, repair initiations in which the speaker not only indicates that there is a need
for repair (as in ‘direct repair initiations’), but also offers a possible solution, a
candidate understanding, of the prior turn (Schegloff et al., 1977; Sorjonen, 1997;
Wiklund, 2016). Direct repair initiations were much more frequent (90.3%) than
candidate understandings (9.7%) in the data, even if other studies have shown that
candidate understandings are often used in ‘asymmetrical’ interactions (Kurhila,
2.5 Conclusion 47

2012; Leskelä, 2012). However, the fact that the therapists mainly used direct repair
initiations instead of candidate understandings is in line with the instructions of plain
language designated for spoken communication (Kartio, 2009).
Half (50.0%) of the direct repair initiations in the data were formulated with the
help of a question word (Wiklund, 2016). Another typical type (25.0%) consisted
of open class repair initiators, such as mitä (‘what’) and its variants tä(h) and hä(h)
as well as anteeks (‘sorry’), etc. (Drew, 1997; Haakana, 2011). Repair initiations in
which the trouble-source turn is partially or completely repeated in the repair initia-
tion represented 14.3% of the direct initiations. Sometimes, the repetition was asso-
ciated with a question word or with the interrogative clitic -kO. The most uncommon
repair initiation type in these data consisted of an explicit verbalisation of the under-
standing problem (10.7%). This type was used only by therapists, and other studies
have shown that it is also rare in other data gathered from asymmetrical institutional
situations (Kurhila, 2012; Lindholm, 2012).
As for prosody, the most common feature related to trouble-source turns in these
data is creaky voice, which occurs in more than one-third (35.3%) of the cases. Quiet
voice is also relatively frequent; it appeared in nearly one-third of the trouble-source
turns (31.4%). Large pitch excursions (23.5%) occurred in about one-quarter of the
instances. Stretched syllables (17.6%) and jerky speech rhythm (15.7%) appeared
in less than one-fifth of the cases. Bouncing pitch (11.8%) and accelerated speech
rate (11.8%) are other prosodic features that could be found in the trouble-source
turns. Previous studies had already shown that ASD persons’ speech often includes
deviant prosodic features (see Sects. 2.1 and 2.2). The results of the study presented
here (Wiklund, 2016) demonstrate that these features—especially creaky or quiet
voice and large pitch excursions—are also related to comprehension problems in the
interaction of persons with ASD. However, it is noteworthy that even if the trouble-
source turns often carry certain prosodic features in these data and these features
may contribute to the creation of the repair sequence, in very few cases do prosodic
features seem to be the main cause of the comprehension problem (Lehtinen, 2012;
Wiklund, 2016).
The study presented here (Wiklund, 2016) also provides evidence about the prag-
matic and interactional skills of persons with ASD. Indeed, it is noteworthy that the
data include passages in which the informants seem to have the ability to make certain
inferences about the mental states of others (i.e., to have a ‘theory’ of other ‘minds’).
This is remarkable, because it is well-known that the ability in question is impaired in
ASD (e.g., Cummings, 2009: 57). For example, the data include instances in which
the informants show an understanding of other people’s perspectives and/or in which
they are able to orient to implicit meanings, such as the interactional meanings of
discourse particles (see also Chap. 5; Wiklund, 2012).
The third part of this chapter was focused on the use of pitch in questions and
statements in the speech of Finnish-speaking preadolescents with ASD (Wiklund
et al., 2021). The results showed that final pitch rises occur frequently in the speech
of the informants. A final pitch rise is especially common in questions: 20% of all
turns of the boys end in a rising pitch, and 46% of question turns carry a final pitch
48 2 Prosody

rise. Individual differences concerning the frequencies of turns ending in a pitch rise
are, however, significant in these data.
The results of the study (Wiklund et al., 2021) also showed that the participants
with ASD can use utterance-final pitch rises as an interactional resource. Indeed, they
can call for other participants’ reactions and indicate that they take other participants
into account with the help of utterance-final pitch rises (Routarinne, 2003). They
can also ‘recycle’ prosody, in other words, repeat the prosody of a previous speaker.
In addition, the study found that the informants can produce and interpret correctly
prosodic features indicating finality in a spontaneous interaction. In other words, they
can for example produce and interpret such features as falling pitch, creaky voice,
and quiet voice as signs of the end of a turn (Ogden, 2004). They can also emphasise
words with the help of stress and changes in intonation.
The informants of these data seem to be able to use prosodic features as an
interactional resource even if reciprocal interaction is generally difficult for persons
with ASD (APA, 2013). The use of intonation could, however, be practiced even more
in therapy for children and adolescents with ASD. At least according to the study
presented here (Wiklund et al., 2021), preadolescents with ASD have a predisposition
to use prosody in interaction. Thus, these predispositions could be strengthened even
more in the future.

References

Aho, E. (2010). Spontaanin puheen prosodinen jaksottelu. [Doctoral dissertation]. University of

Helsinki.
American Psychiatric Association (APA). (2013). Diagnostic and statistical manual of mental
disorders (DSM-V) (5th ed.) American Psychiatric Association.
Anttila, H. (2008). The effect of interrogative function on intonation in spontaneous and read
Finnish. [Master’s thesis]. University of Helsinki.
Anttila, H. (2009): Interrogative intonation in spontaneous Finnish. In V. de Silva & R. Ullakonoja
(Eds.), Phonetics of Russian and Finnish. General description of phonetic systems. Experimental
studies on spontaneous and read-aloud speech (pp. 167–176). Peter Lang.
Asghari, S. Z., Farashi, S., Bashirian, S., & Jenabi, E. (2021). Distinctive prosodic features of people
with autism spectrum disorder: A systematic review and meta-analysis study. Scientific Reports,
11(1), 23093.
Asperger, H. (1944). Die “Autistischen Psychopathen” im Kindesalter. Archiv Für Psychiatrie Und
Nervenkrankheiten, 117, 76–136.
Baltaxe, C. (1984). The use of contrastive stress in normal, aphasic and autistic children. Journal
of Speech and Hearing Research, 27, 97–105.
Baltaxe, C., & Simmons, J. (1985). Prosodic development in normal and autistic children. In E.
Schopler & G. Mesibov (Eds.), Communication problems in autism (pp. 95–125). Plenum Press.
Baltaxe, C., & Simmons, J. (1992). A comparison of language issues in high-functioning autism
and related disorders with onset in children and adolescence. In E. Schopler & G. Mesibov (Eds.),
High-functioning individuals with autism (pp. 210–225). Plenum Press.
Boersma, P., & Weenink, D. (2021). Praat: Doing phonetics by computer. [Computer program].
http://www.praat.org/.
References 49

Chevallier, C., Noveck, I., Happé, F., & Wilson, D. (2009). From acoustics to grammar: Perceiving
and interpreting grammatical prosody in adolescents with Asperger Syndrome. Research in
Autism Spectrum Disorders, 3(2), 502–516.
Couper-Kuhlen, E. (2000). Prosody. In J. Verschueren, J.-O. Östman, & C. Bulcaen (Eds.),
Handbook of pragmatics (pp. 1–19). John Benjamins Publishing Company.
Crystal, D. (1980). A first dictionary of linguistics and phonetics. Deutsch.
Cummings, L. (2009). Clinical pragmatics. Cambridge University Press.
De Marchena, A., & Miller, J. (2017). “Frank” presentations as a novel research construct and
element of diagnostic decision-making in autism spectrum disorder. Autism Research, 10(4),
653–662.
Delais-Roussarie, E., Post, B., Avanzi, M., Buthke, C., Di Cristo, A., Feldhausen, I., Jun, S.-A.,
Martin, Ph., Meisenburg, T., Rialland, A., Sichel-Bazin, R., & Yoo, H.-Y. (2015). Intonational
phonology of French: Developing a ToBI system for French. In S. Frota & P. Prieto (Eds.),
Intonation in romance. Oxford University Press.
DePape, A.-M. R., Chen, A., Hall, G. B. C., & Trainor, L. J. (2012). Use of prosody and informa-
tion structure in high functioning adults with autism in relation to language ability. Frontiers in
Psychology, March 2012, Vol. 3, Article 72.
Diehl, J., & Paul, R. (2013). Acoustic and perceptual measurements of prosody production on the
profiling elements of prosodic systems in children by children with autism spectrum disorders.
Applied Psycholinguistics, 34(1), 135–161.
Drew, P. (1997). ‘Open’ class repair initiators in response to sequential sources of troubles in
conversation. Journal of Pragmatics, 28(1997), 69–101.
Fay, W., & Schuler, A. L. (1980). Emerging language in autistic children. University Park Press.
Fine, J., Bartolucci, G., Ginsberg, G., & Szatmari, P. (1991). The use of intonation to communicate
in pervasive developmental disorders. Journal of Child Psychology and Psychiatry, 32, 771–782.
Fosnot, S. M., & Jun, S.-A. (1999). Prosodic characteristics in children with stuttering or autism
during reading and imitation. In: J. J. Ohala, Y. Hasegawa, M. Ohala, D. Granville & A. C.
Bailey (Eds.), Proceedings of the 14th International Congress of Phonetic Sciences (ICPhS), San
Francisco, CA, USA.
Globerson, E., Amir, N., Kishon-Rabin, L., & Golan, O. (2014). Prosody recognition in adults
with high-functioning autism spectrum disorders: From psychoacoustics to cognition. Autism
Research, 8(2), 153–163.
Green, H., & Tobin, Y. (2008). Intonation in hebrew-speaking children with high-functioning autism:
A case study. In P.A. Barbosa, S. Madureira & C. Reis (Eds.), Speech prosody 2008, fourth interna-
tional conference, campinas, Brazil. https://www.isca-speech.org/archive_v0/sp2008/sp08_237.
html.
Green, H., & Tobin, Y. (2009). Prosodic analysis is difficult…but worth it: A study in high
functioning autism. International Journal of Speech-Language Pathology 11, 308–315.
Grossman, R. B., Bemis, R. H., Plesa Skwerer, D., & Tager-Flusberg, H. (2010). Lexical and
affective prosody in children with high-functioning autism. Journal of Speech, Language, and
Hearing Research, 53, 778–793.
Haakana, M. (2011). “Mitä” ja muut avoimet korjausaloitteet [“Mitä” ‘what’ and other repair
initiations]. Virittäjä, 115(1), 36-67.
Hakulinen, A., Vilkuna, M., Korhonen, R., Koivisto, V., Heinonen, T. R. & Alho, I. (2004). Iso
suomen kielioppi. SKS.
Hirvonen, P. (1970). Finnish and English communicative intonation. University of Turku.
Holbrook, S., & Israelsen, M. (2020). Speech Prosody Interventions for Persons with Autism Spec-
trum Disorders: A Systematic Review. American Journal of Speech-Language Pathology, 29(4),
2189–2205.
Huhtamäki, M. (2015). En fråga om prosodi? Prosodiska drag hos frågor i Helsingforssvenska
samtal. [Doctoral dissertation]. University of Helsinki.
Iivonen, A. (1978). Is there interrogative intonation in Finnish? In E. Gårding, G. Bruce & R.
Bannert (Eds.), Nordic Prosody. Papers from a Symposium (pp. 43–53). Lund University.
50 2 Prosody

Iivonen, A. (1998). Intonation in Finnish. In D. Hirst & A. Di Cristo (Eds.), Intonation systems: A
survey of twenty languages (pp. 311–327). Cambridge University Press.
Iivonen, A., Nevalainen, T., Aulanko, R., & Kaskinen, H. (1987). Puheen intonaatio. Gaudeamus.
Iivonen, A. (2001). Intonation in Finnish questions. In: Nordic Prosody. Proceedings of the VIII
Conference, Trondheim 2000, W. Van Dommelen & T. Fretheim (Eds.). Frankfurt am Main: Peter
Lang.
Iivonen, A. (2006). Creaky Voice as a prosodic feature in Finnish. In G. Bruce & M. Horne (Eds.),
Nordic Prosody. Proceedings of the IX Conference, Lund 2004 (pp. 137–146). Peter Lang.
Iivonen, A. (2009). Finnish sentence accent and intonation. In V. de Silva & R. Ullakonoja (Eds.),
Phonetics of Russian and Finnish. General description of phonetic systems. Experimental studies
on spontaneous and read-aloud speech (pp. 67–73). Peter Lang.
Kallioinen, V. (1968). Suomen kysymyslauseen intonaatiosta. Virittäjä, 72, 35–54.
Kanner, L. (1943). Autistic disturbances of affective contact. The Nervous Child, 2, 217–250.
Kartio, J. (Ed.). (2009). Selkokieli ja vuorovaikutus. Kehitysvammaliitto.; Ke, F., Whalon, K., &
Yun, J. (2018). Social skill interventions for youth and adults with autism spectrum disorder: A
systematic review. Review of Educational Research, 88(1), 3–42.
Korpilahti, P., Jansson-Verkasalo, E., Mattila, M.-L., Kuusikko, S., Suominen, K., Rytky, S., Pauls,
D. L., & Moilanen, I. (2007). Processing of affective speech prosody is impaired in Asperger
syndrome. Journal of Autism and Developmental Disorders, 37(8), 1539–1549.
Koskela, A. (2013). Aikuisten puhe-ja artikulaationopeus sekä artikulaationopeuden yhteys
oraalimotorisiin taitoihin. [Master’s thesis]. University of Oulu.
Kremer-Sadlik, T. (2001). How children with autism and Asperger syndrome respond to questions:
An ethnographic study. [Doctoral dissertation]. University of California.
Kurhila, S. (2012). Kun ymmärtäminen on vaakalaudalla-avoimet korjausaloitteet kakkoskielisessä
keskustelussa. In L. Leskelä & C. Lindholm (Eds.), Haavoittuva keskustelu. Keskustelunanalyyt-
tisia tutkimuksia kielellisesti epäsymmetrisestä vuorovaikutuksesta (pp. 145–183). Kehitysvam-
maliitto.
Lau, J. C. Y., Patel, S., Kang, X., Nayar, K., Martin, G. E., Choy, J., Wong, P. C. M., & Losh, M.
(2022). Cross-linguistic patterns of speech prosodic differences in autism: A machine learning
study. PLoS ONE, 17(6), e0269637.
Lehtinen, M. (2010a). Observations on the prosodic characteristics of Finnish-speaking youngsters
with Asperger Syndrome. In: Botinis, A. (Ed.), Proceedings of ‘ExLing 2010a—ISCA Tutorial
and Research Workshop on Experimental Linguistics’ (pp. 93–96). ISCA/University of Athens.
Lehtinen, M. (2010b). The recategorisation of the rheme and the structure of the oral paragraph in
French and in Finnish. Discours, 7.
Lehtinen, M. (2012). “Aiheesta toiseen”—Ymmärrysongelmatilanteita Asperger-nuorten kuntou-
tuskeskusteluissa. In L. Leskelä & C. Lindholm (Eds.), Haavoittuva keskustelu. Keskustelu-
nanalyyttisia tutkimuksia kielellisesti epäsymmetrisestä vuorovaikutuksesta (pp. 235–277).
Kehitysvammaliitto.
Leskelä, L. (2012). Selkokielinen vuorovaikutus kehitysvammaisen keskustelijan tukena. In L.
Leskelä & C. Lindholm (Eds.), Haavoittuva keskustelu. Keskustelunanalyyttisia tutkimuksia
kielellisesti epäsymmetrisestä vuorovaikutuksesta (pp. 184–234). Kehitysvammaliitto.
Lindholm, C. (2012). Vuorovaikutuksen haasteita ja mahdollisuuksia—tapaustutkimus muisti-
sairaiden päivätoiminnasta. In L. Leskelä & C. Lindholm (Eds.), Haavoittuva keskustelu.
Keskustelunanalyyttisiä tutkimuksia kielellisesti epäsymmetrisestä vuorovaikutuksesta (pp. 57–
98). Kehitysvammaliitto.
Lindström, R. (2019). Neural and perceptual processing of emotional speech prosody in typi-
cally developed children and in children with autism spectrum disorder. [Doctoral dissertation].
University of Helsinki.
Lohi, N. (2020). Prosody of adolescent Finnish boys diagnosed with ASD: Variation within the
group and comparison with neurotypically developed adolescents. [Master’s thesis]. University
of Helsinki.
References 51

Nadig, A., & Shaw, H. (2012). Acoustic and perceptual measurement of expressive prosody in
high-functioning autism: Increased pitch range and what it means to listeners. Journal of Autism
and Developmental Disorders, 42(4), 499–511.
Nakai, Y., Takashima, R., Takiguchi, T. & Takada, S. (2014). Speech intonation in children with
autism spectrum disorder. Brain and Development, 36(6), 516–522.
Neurobehavioral Systems. (2021). Presentation. [Computer program]. https://www.neurobs.com/.
Le Normand, M.-T., Boushaba, S., & Lacheret-Dujour, A. (2008). Prosodic disturbances in autistic
children speaking French. In P. A. Barbosa, S. Madureira & C. Reis (Eds.), Proceedings of
the Speech Prosody 2008, Campinas, Brazil. https://www.isca-speech.org/archive/sp2008/pap
ers/sp08_195.pdf.
Ogden, R. (2001). Turn transition, creak and glottal stop in Finnish talk-in interaction. Journal of
the International Phonetic Association, 31, 139–152.
Ogden, R. (2004). Non-modal voice quality and turn-taking in Finnish. In: Sound patterns in inter-
action, E. Couper-Kuhlen & C. E. Ford (Eds.), (pp. 29–62). Amsterdam / Philadelphia: John
Benjamins Publishing Company.
Ogden, R. & Routarinne, S. (2005). The communicative functions of final rises in Finnish intonation.
Phonetica 62, 160–175.
Olivati, A. G., Assumpção Jr, F. B., & Nunes Misquiatti, A. R. (2017). Acoustic analysis of speech
intonation pattern of individuals with autism spectrum disorders. CoDAS 2017, 29(2), e20160081.
Paul, R. (1987). Communication. In D. Cohen & A. Donnellan (Eds.), Handbook of autism and
pervasive developmental disorders (pp. 61–84). Wiley.
Paul, R., Augustyn, A., Klin, A., & Volkmar, F. R. (2005a). Perception and Production of Prosody
by Speakers with Autism Spectrum Disorders. Journal of Autism and Developmental Disorders,
35(2), 205–220.
Paul, R., Shriberg, L., Mcsweeney, J., Cicchetti, D., Klin, A., & Volkmar, F. R. (2005b). Brief
report: relations between prosodic performance and communication and socialization ratings in
high functioning speakers with autism spectrum disorders. Journal of Autism and Developmental
Disorders, 35(6), 861–869.
Peppé, S., Martinez Castilla, P., Lickley, R., Mennen, I., Mccann, J. O’hare, A., & Rutherford, M.
(2006). Functionality and perceived atypicality of expressive prosody in children with autism
spectrum disorders. In R. Hoffmann & H. Mixdorff (Eds.), Proceedings of the Speech Prosody
2006, Dresden, Germany. https://www.isca-speech.org/archive/sp2006/papers/sp06_060.pdf.
Redford, M. A., Kapatsinski, V., & Cornell-Fabiano, J. (2018). Lay listener classification and
evaluation of typical and atypical children’s speech. Language and Speech, 61(2), 277–302.
Rosenblau, G., Kliemann, D., Dziobek, I., & Heekeren, H. R. (2017). Emotional prosody processing
in autism spectrum disorder. Social Cognitive and Affective Neuroscience, 12(2), 224–239.
Routarinne, S. (2003). Tytöt äänessä. Parenteesit ja nouseva sävelkulku kertojan vuorovaiku-
tuskeinoina. SKS.
Sacks, H., Schegloff, E., & Jefferson, G. (1974). A simplest systematics for the organization of
turn-taking for conversation. Language, 50(4), 696–735.
Schegloff, E., Jefferson, G., & Sacks, H. (1977). The preference for self-correction in the
organization of repair in conversation. Language, 53(2), 361–382.
Shriberg, L., Paul, R., McSweeney, J., Klin, A., Cohen, D., & Volkmar, F. R. (2001). Speech and
prosody characteristics of adolescents and adults with high functioning autism and Asperger
syndrome. Journal of Speech, Language and Hearing Research, 44, 1097–1115.
SIL Glossary of Linguistic Terms. (2021). https://glossary.sil.org/.
Solomon, O. (2004). Narrative introductions: Discourse competence of children with autistic
spectrum disorders. Discourse Studies, 6(2), 253–276.
Sorjonen, M.-L. (1997). Korjausjäsennys. In L. Tainio (Ed.), Keskustelunanalyysin perusteet
(pp. 111–137). Vastapaino.
Szczepek Reed, B. (2009). Prosodic orientation: A practice for sequence organization in broadcast
telephone openings. Journal of Pragmatics, 41(6), 1223–1247.
52 2 Prosody

Tager-Flusberg, H. (2000). Understanding the language and communicative impairments in autism.

International Review of Research on Mental Retardation, 23, 185–205.
Thorson, J. C., Usher, N., Patel, R. & Tager-Flusberg, H. (2016). Acoustic analysis of prosody in
spontaneous productions of minimally verbal children and adolescents with autism. In J. Barnes,
A. Brugos, S. Shattuck-Hufnagel & N. Veilleux (Eds.), Proceedings of Speech Prosody (Vol. 8).
Van Bourgondien, M. E., & Woods, A. V. (1992). Vocational possibilities for high functioning
adults with autism. In E. Schopler & G. Mesibov (Eds.), High functioning individuals with autism
(pp. 227–242). Plenum Press.
Volden, J. (2004). Conversational repair in speakers with autism spectrum disorder. International
Journal of Language & Communication Disorders, 39(2), 171–189.
Wiklund, M. (2012). Gaze behavior of pre-adolescent children afflicted with Asperger Syndrome.
Communication & Medicine, 9(2), 173–186.
Wiklund, M. (2016). Interactional challenges in conversations with autistic preadolescents: the role
of prosody and nonverbal communication in other-initiated repairs. Journal of Pragmatics, 94,
76–97.
Wiklund, M., & Laakso, M. (2019). Ungrammatical utterances and disfluent speech as causes
of comprehension problems in interactions of preadolescents with highly functioning autism.
Clinical Linguistics & Phonetics, 33(7), 654–676.
Wiklund, M., & Laakso, M. (2020). Comparison of disfluent and ungrammatical speech of pread-
olescents with and without ASD. Journal of Autism and Developmental Disorders, 51(8),
2773–2789.
Wiklund, M., & Määttä, S. (2021). Therapists’ response strategies in a group session involving
French-speaking children with autism spectrum disorder. Journal of Interdisciplinary Voice
Studies, 6(1), 109–130.
Wiklund, M., & Stevanovic, M. (2018). Ymmärrysongelmat ja vuorovaikutustaitojen opet-
taminen lievästi autististen varhaisnuorten kuntoutuskeskusteluissa [Understanding problems
and the teaching of social interaction skills in group therapy sessions of preadolescents with
high-functioning autism]. Psykologia, 53(05–06), 421–451.
Wiklund, M., & Vainio, M. (2019). Pitch-related features in the speech of Finnish- and French-
speaking boys with autism in data coming from group therapy sessions. In H. E. H. Lenk, J.
Härmä, B. Sanromán Vilas, & E. Suomela-Härmä (Eds.), Studies in Comparative Pragmatics
(pp. 45–63). Cambridge Scholars Publishing.
Wiklund, M., Ihaksinen, K., & Vainio, M. (2021). Autismikirjon poikien intonaation käyttö
kysymys-ja kerrontavuoroissa spontaanissa vuorovaikutuksessa [The Use of Intonation in Ques-
tions and Narrative Turns in the Spontaneous Interaction of Boys Afflicted with Autism Spectrum
Disorder]. Puhe Ja Kieli, 41(1), 43–70.
Wiklund, M., Vainio, L., Saalasti, S. & Vainio, M. (forthcoming). Puheen prosodian havaittu
epätyypillisyys suomenkielisillä autismikirjon varhaisnuorilla [Perceived Atypicality of Speech
Prosody of Finnish-Speaking Preadolescents with Autism Spectrum Disorder]. Puhe ja kieli.
Zhou, P., Ma, W., & Zhan, L. (2019). A deficit in using prosodic cues to understand communicative
intentions by children with autism spectrum disorders: An eye-tracking study. First Language,
40(1), 41–63.
Chapter 3
Disfluencies

Abstract Speech disfluencies interrupt the continuous flow of speech and may thus
present challenges for conversational interaction. In fluent speech, there are few
interruptions, and listeners can concentrate on the contents of the speech without
being disturbed by the manner in which the speech is produced. ‘Disfluency’ refers
to the interruption of ongoing speech and exhibits features such as silent and filled
pauses, sound prolongations, repetitions, and cut-off utterances. According to the
studies presented in this chapter, the durational mean proportions of disfluencies
and ungrammatical expressions are greater in the speech of participants with ASD
(26.4%) than in the control group (15.5%) (Wiklund & Laakso, 2020). Individual
variation in the amount of disfluency is high within both groups. In addition, a qual-
itative difference can be found: The boys with ASD produce long and complex
disfluent turns with word searches, self-repairs, false starts, fillers, sound prolonga-
tions, inconsistent syntactic structures, and grammatical errors, whereas the control
group members mainly produce fillers and sound prolongations. In the conversa-
tional interactions of the speakers with ASD, disfluencies also cause comprehension
problems; the control group members, in turn, do not experience comprehension
problems (Wiklund & Laakso, 2019, 2020).

Keywords Disfluencies · Ungrammatical utterances · Disfluent speech ·

Comprehension problems · Interaction · ASD · Autism · Conversation analysis

3.1 Overview of Previous Research on Disfluencies

in the Speech of Persons with ASD1

Speech disfluencies interrupt the continuous flow of speech and may thus present
challenges for conversational interaction. When speech is fluent, speech produc-
tion flows smoothly both in terms of sound and information (e.g., Manning, 2010).
In fluent speech, there are few interruptions, and listeners can concentrate on the
contents of the speech without being disturbed by the manner in which the speech

1 The literature review is adapted from the ones presented in Wiklund and Laakso (2019, 2020).

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 53
M. Wiklund, Speech and Interaction of Preadolescents with Autism Spectrum Disorder,
Prosody, Phonology and Phonetics, https://doi.org/10.1007/978-981-19-8117-3_3
54 3 Disfluencies

is produced. ‘Disfluency’ refers to the interruption of ongoing speech and exhibits

features such as silent and filled pauses, sound prolongations, repetitions, and cut-off
utterances (e.g., Shriberg, 2001). Disfluencies have been connected to speech plan-
ning and processing, and they are relevant to listeners’ comprehension of speech (e.g.,
Wiklund & Laakso, 2019, 2020). It is noteworthy that normally fluent speakers also
present disfluencies: for example, they hesitate, repeat, and interrupt their speech in
order to formulate their expressions (e.g., McDougall & Duckworth, 2017). Speakers
also display considerable individual variation in the rate and types of disfluency
features they produce. Disfluencies of speech not only assist speakers in planning
and articulating their speech, but also give listeners time to understand what was
said.
Disfluencies of speech can also be more disturbing and can be divided into
stuttering-like disfluencies (SLD) and other disfluencies (OD) (Ambrose & Yairi,
1999; Byrd et al., 2012; Tumanova et al., 2014). Ambrose and Yairi (1999: 899)
include the following features in the category of SLD: (a) part-word repetitions
(‘b-but,’ ‘thi-thi-this’), (b) single-syllable word repetitions (‘you,’ ‘and and’), (c)
disrhythmic phonations, such as prolongations (‘mmmmy,’ ‘cooookie’), blocks
(‘#toy’), and broken words (‘o#pen’). The OD group, in turn, includes such features
as (d) interjections (‘um’), (e) revisions/abandoned utterances (“Mom ate/Mom fixed
dinner,” “I want/Hey look at that”), and (f) multisyllable/phrase repetitions (“because
because,” “I want I want to go”).
Disfluencies have been defined by different authors in varying ways using different
terminology; researchers have not always had similar views on what can be defined as
‘stuttering-like’ and what can be defined as belonging to the category of ‘other disflu-
encies’. Several studies have demonstrated that many speakers with ASD produce
disfluent speech (e.g., Lake et al., 2011; Scaler Scott et al., 2014; Scott et al., 2006;
Shriberg et al., 2001; Sisskin, 2006; Wiklund & Laakso, 2019, 2020). The defini-
tions of disfluencies differ in various studies, but persons with ASD have been found
to produce numerous stuttering-like disfluencies such as sound, syllable, and word
repetitions, as well as atypical disfluencies such as breaks, insertions, and word-final
sound prolongations.
Shriberg et al. (2001) studied a group of 15 speakers with High-Functioning
Autism (HFA) and a group of 15 speakers with Asperger Syndrome (AS) and
compared the groups to one another as well as to the profiles of 53 control speakers.
All participants were male, and their ages ranged from 10 to 50 years. According to
the results—differing from the controls—67% of the individuals with AS and 40% of
the individuals with HFA had inappropriate or non-fluent phrasing, including sound,
syllable, or word repetitions and single-word revisions in more than 20% of their
utterances. In a study by Lake et al. (2011), the speech of 13 adults with ASD (4
of whom were diagnosed with AS) and 13 controls was studied. On the one hand,
a higher number of silent pauses and disfluent repetitions were found in the ASD
group compared with the control group. On the other hand, fewer ‘listener-oriented
disfluencies’ such as filled pauses and revisions occurred in the ASD group’s speech
than in the control group’s speech.
3.1 Overview of Previous Research on Disfluencies in the Speech of Persons … 55

Scott et al. (2006) found that the speech of two young adults with AS included
stuttering-like disfluencies (part-word repetitions and blocks), as well as non-
stuttering-like disfluencies (phrase repetitions and interjections). Sisskin’s (2006)
study, in turn, examined the speech of two persons with AS (aged 7 and 17 years). Both
informants exhibited within-word stuttering-like disfluencies (part and whole-word
repetitions and blocks) and between-word non-stuttering-like disfluencies (phrase
repetitions, revisions, and interjections). Stuttering-like disfluencies were either mid-
syllable insertions (defined as a short exhalation resembling the production of /h/) or
word-final disfluencies (repetitions in which the repetition forms a rhyme by omitting
the initial consonant(s) or syllable of the target word, e.g., ‘train-ain’) that are not
typical of developmental stuttering.
A study by Scaler Scott et al. (2014) was directed towards the speech of groups
of 11 school-aged children with AS, 11 matched children who stutter (CWS), and
11 matched control children with no diagnosis (ND). The results reveal that the
three participant groups displayed differences that were statistically significant in
terms of the percentage of words that contained stuttering-like disfluencies. The
group of CWS produced a higher percentage of stuttering-like disfluencies than the
other groups (41% for CWS; 21% for children with AS), while the children with
AS produced a higher percentage of word-final disfluencies2 (5% for children with
AS and 1% for CWS). Word-final disfluencies were present in the speech samples
of eight out of 11 children with AS. By comparison, word-final disfluencies were
present in four out of 11 children of the CWS group and three out of 11 children of
the ND group.
The definitions and classifications of disfluencies differed in the studies presented
above, but all studies found that speakers with AS frequently have disfluent speech
with at least repetitions, and some studies also suggested that speakers with AS are
more disfluent than the controls. Several studies also indicate that speakers with AS
revise and self-repair their speech (Shriberg et al., 2001; Sisskin, 2006; Wiklund &
Laakso, 2019, 2020), although contradictory findings have also been reported (Lake
et al., 2011). Insertions (Lake et al., 2011), interjections (Scott et al., 2006; Sisskin,
2006), pauses (Lake et al., 2011), and word-final prolongations (Scaler Scott, 2014;
Sisskin, 2006) have also been found to occur. Stuttering-like disfluencies (e.g., part- or
whole-word repetitions and blocks) were observed in three studies (Scaler Scott et al.,
2014; Scott et al., 2006; Sisskin, 2006). School-aged children with AS nevertheless
differed from children who stutter.
Engelhardt et al. (2017) compared the disfluencies in the speech of groups of 13
adults with HFA, 13 age- and gender-matched controls and 13 randomly selected
controls. The results showed that adults with HFA produced more repair disfluencies
than the controls, and that the tendency to produce repairs is likely ‘speaker-oriented’.
With respect to repetitions, significant differences were not found between the groups.
The authors also observed differences concerning the occurrences of unfilled pauses:

2 Word-final disfluencies were defined in this study as part-word repetitions of final sounds or
syllables of words (e.g., ‘chair-air’ and ‘light-t’) or prolongations of the final sounds of words (e.g.,
‘thissss’).
56 3 Disfluencies

the unmatched controls produced fewer unfilled pauses compared to matched controls
and the persons with HFA.
According to a study by Park et al. (2018), school-aged children with ASD are
disfluent in their verbal production—except when reading. The study included seven
children with ASD, six children who stutter, and eight control children. The partic-
ipants ranged in age from 6 to 8 years, and they were matched for age and vocab-
ulary ability. The participants’ speech disfluencies were studied in reading, story
retelling, and picture description tasks. Disfluency characteristics found in the data
were identified and classified by the authors as ‘stuttering-like disfluencies’ or ‘other
disfluencies’. The results of the study showed that the three groups of informants did
not differ in the frequencies of other disfluencies across the three speaking tasks. In
the group of children with ASD, the total amount of disfluencies was significantly
higher in the story retelling task than in the reading task. Similarly, among the chil-
dren with ASD, the frequencies of other disfluencies in the story retelling task and the
picture description task were significantly higher than in their reading performance.
The frequencies of stuttering-like disfluencies, in turn, did not change significantly
across tasks in the ASD group.
Pirinen et al. (2021) have studied disfluencies of speech among young Finnish-
speaking adults with ASD (age range: 19–29 years, n = 30) and their controls (n =
30). The informants were shown seven video clips from socio-pragmatically chal-
lenging situations. After having seen the video clips, the informants were asked to
speak about what they had just seen in their own words. Typical disfluencies such
as word and phrase repetitions, filled pauses, rephrasings, and interrupted words and
clauses were analysed. More disfluencies could be found in the speech of the infor-
mants with ASD than in the speech of the control group members, and the difference
was statistically significant. When the amounts of different types of disfluencies were
compared between the groups, it was found that persons with ASD produced more
disfluencies than the controls, regardless of the type of disfluency. The difference was
greatest concerning filled pauses and interrupted words and clauses. Considerable
individual variation in the amounts of disfluencies could be found in both groups.
In addition, Pirinen (2022: 70) found that disfluencies in the speech of persons with
ASD may reflect their challenges in linguistic, cognitive, and senso-motoric skills.
Ungrammatical expressions also create an impression of disfluency in the speech
of people with ASD. A study by Wiklund and Laakso (2019) demonstrated that
incorrect case endings, ambiguous pronominal references, grammatically incoherent
syntactic structures, and inaccurate tenses are features that make speech flow disfluent
and difficult to follow. Indeed, syntactic impairments have been found in individuals
with ASD (Cummings, 2014a). For example, children with ASD may align their
use of syntactic structures to that of a conversational partner (Allen et al., 2011).
These children also encounter difficulties in their grammatical comprehension of
instructions (Saalasti et al., 2008), and they tend to map verbs onto causative actions
(Naigles et al., 2011). McGregor et al. (2012) report in their study that only those
children with ASD who do not have syntactic deficits in their speech demonstrate an
age-appropriate knowledge of words. The word learning of children with ASD has
3.2 Disfluencies in the Speech of Preadolescent Boys with and Without ASD 57

also been found to be compromised due to their reduced sensitivity to the social infor-
mation of gaze cues (Norbury et al., 2010). Thus, studies have concluded that children
with ASD may have more profound linguistic difficulties underlying their challenges
in discourse and social interaction (Saalasti et al., 2008; Wiklund & Laakso, 2019,
2020).

3.2 Disfluencies in the Speech of Preadolescent Boys

with and Without ASD3

The main research objective of the study presented in this sub-chapter was to compare
the quantity of disfluencies occurring in the speech of two groups: (1) Finnish-
speaking preadolescent boys with ASD (see Sect. 1.2.1)4 and (2) Finnish-speaking
age- and gender-matched controls with typical development (see Sect. 1.2.3)
(Wiklund & Laakso, 2020). The analyses were quantitative and qualitative. First, the
durations of all speech extracts including disfluencies and ungrammatical expres-
sions were measured. Then the extracts were qualitatively analysed using methods
of Conversation Analysis (e.g., Hutchby & Wooffit, 2008). The results of the qual-
itative analyses are presented in Sect. 3.3. Ungrammatical expressions have been
included in the study, because they also create an impression of disfluent speech
and are typical of the informants with ASD in this study (Wiklund & Laakso, 2019,
2020).
The speech analysis program ‘Praat’ (Boersma & Weenink, 2021) was used to
measure the durations of disfluent and ungrammatical extracts of speech in the data.
The percentages for the durations of disfluent and fluent speech were then compared
between individuals and groups. After that, the ‘Atlas.ti’ program (ATLAS.ti Scien-
tific Software Development GmbH, 2021) was used to classify and code the occur-
rences of extracts of disfluent and ungrammatical speech. For the coding, the speaking
turns for each participant (ASD group and control group) were identified. The turns
were subsequently classified as (1) turns including disfluencies, (2) turns including
ungrammatical expressions, and (3) turns including both. The frequencies were calcu-
lated in order to determine the proportion of the speaking turns affected. Due to the
small sample size, the durations and occurrences in turns per participant were exam-
ined and analysed using descriptive statistics (mean, median, and range of variation)
to compare the proportions of speech disfluencies and ungrammatical expressions in
the ASD group with those of the control group.

3 This sub-chapter presents an overview of a study described in more detail in Wiklund and Laakso
(2020).
4 Two boys without an ASD diagnosis who participated in the group therapy sessions of these data

have been excluded from this study. Thus, the number of informants with ASD is five (n = 5) in
the study described in this sub-chapter.
58 3 Disfluencies

As disfluencies, the following features have been taken into account: word
searches, self-repairs, false starts, fillers, prolongations, inconsistent syntactic struc-
tures, blocks, and grammatical errors. As grammatical errors, such features as incor-
rect case endings, inconsistent syntactic structures, unclear references, and erroneous
lexical choices have been considered. Table 3.1 provides examples of the different
types of disfluencies and grammatical errors considered in the study (Wiklund &
Laakso, 2020). In these examples, the talk is depicted in three lines: the first line
is the original utterance in Finnish, the second line offers a word-by-word gloss in
English, and the third line provides an English translation.5
The preadolescents with ASD produced more disfluencies and ungrammatical
expressions than the neurotypical controls in these data. The mean proportion of
disfluencies and ungrammatical expressions in the total duration of their speech was
26.4% for the ASD group and 15.5% for the control group (Wiklund & Laakso,
2020). The group means of fluent and disfluent speech and distributions for each
participant are presented in Figs. 3.1 (ASD group) and 3.2 (Control group).6
The disfluency percentages in the speech of the participants with ASD ranged from
35% by A1 to 15% by A4, with a mean of 26.4% and a median of 29% (Fig. 3.1).
In the control group, the range was from 24% by C3 to 6% by C4, with a mean of
15.5% and a median of 15% (Fig. 3.2). Although the means and medians of the ASD
and control groups clearly differed, considerable individual variation occurred in the
percentage of disfluency within the groups (Wiklund & Laakso, 2020).
In total, the group with participants with ASD produced more speaking turns (412
speaking turns in 30 min) than the control group (192 speaking turns in 30 min).
When the number of speaking turns affected by disfluency were examined, the mean
proportion of disfluent speaking turns was greater in the control group (44.7%) than
in the data from the ASD group (35%) (Wiklund & Laakso, 2020). In the group with
participants with ASD, the percentage of disfluent turns ranged from 52.7 to 20.4%
with a median of 26.7%, whereas in the control group, the range was from 70.3
to 30% with a median of 42.2% (Wiklund & Laakso, 2020). Although the controls
produced more speaking turns with disfluencies, their disfluencies were mainly short
pauses and single fillers or sound prolongations which did not disturb the overall
flow of speech, whereas the ASD group produced long, complex disfluent turns with
word searches, self-repairs, false starts, fillers, prolongations, inconsistent syntactic
structures, and grammatical errors (Wiklund & Laakso, 2020). The disfluencies and
ungrammatical expressions occurring in the interactions of the participants with ASD
also caused comprehension problems (Wiklund & Laakso, 2019, 2020). These will
be discussed in greater detail in the next Sect. 3.3.

5 The contents of Table 3.1 have been published in Wiklund and Laakso (2020).
6 The contents of Figs. 3.1 and 3.2 have been published in Wiklund and Laakso (2020).
3.2 Disfluencies in the Speech of Preadolescent Boys with and Without ASD 59

Table 3.1 Examples of different types of disfluencies and grammatical errors (PRT =
particle/interjection; (.) = micro pause; (0.6) = measured pause of 0.6 s; # = creaky voice; ↑
= rising shift in intonation; : = sound prolongation; <> = slow pace; > < = fast pace)
Type of disfluency Example
Word searches

Self-repairs

False starts

Fillers

Sound prolongations

Blocks There were no blocks in the data

(continued)
60 3 Disfluencies

Table 3.1 (continued)

Type of disfluency Example
Grammatical errors Wrong case endings:

Inconsistent syntactic structures:

Unclear references:

Erroneous lexical choices:

3.2 Disfluencies in the Speech of Preadolescent Boys with and Without ASD 61

100%
90%
80%
70%
60%
Disfluent
50%
40% Fluent

30%
20%
10%
0%
A1 A2 A3 A4 A5 mean

Fig. 3.1 Percentages of fluent and disfluent speech of participants with ASD

100%
90%
80%
70%
60%
50% Disfluent

40% Fluent
30%
20%
10%
0%
C1 C2 C3 C4 C5 C6 mean

Fig. 3.2 Percentages of fluent and disfluent speech of control participants

62 3 Disfluencies

3.3 The Role of Disfluencies in the Creation

of Comprehension Problems7

The objective of the studies presented in this sub-chapter is to discuss the role
of disfluencies in the creation of comprehension problems in the interactions
of (1) Finnish-speaking preadolescent boys with ASD (see Sect. 1.2.1) and (2)
Finnish-speaking age- and gender-matched controls with typical development (see
Sect. 1.2.3) (Wiklund & Laakso, 2019, 2020). More precisely, the analyses focus on
turns that include ungrammatical utterances and/or disfluent speech and the role of
these turns in the creation of comprehension problems between the participants. The
study further aims to discover how ungrammatical and disfluent turns are addressed
by the participants of the conversation. That is, how do the other participants and
the speaker himself try to resolve the comprehension problem? (Wiklund & Laakso,
2019).
The analyses have been carried out using methods of Conversation Analysis (CA).
As already mentioned, in the studies presented in this book, ‘comprehension prob-
lems’ are materialised by other-initiated repairs in which the repair process is started
by the recipient of the problematic utterance (Schegloff et al., 1977; SIL Glossary
of Linguistic Terms, 2021). When studying comprehension problems from the point
of view of CA, it is the subsequent action that shows what the trouble is or whether
something in the previous speaker’s turn is oriented to as troublesome, problematic,
repairable, or ungrammatical, and thus correctable (Schegloff et al., 1977). Thus,
the analyses presented in this sub-chapter will focus on finding out whether the
ungrammatical and disfluent turns of speakers with ASD are treated as problem-
atic (Wiklund & Laakso, 2019), and whether differences can be found between the
speakers with ASD and the neurotypical controls (Wiklund & Laakso, 2020).
Example 3.1 illustrates the most common sequential trajectory related to compre-
hension problems caused by disfluencies found in these data (Wiklund & Laakso,
2019). In this trajectory, there is a therapist’s other-initiated repair after a disfluent
and morpho-syntactically complex turn produced by a boy with ASD. The extract
comes from the session of Group B. One of the boys, Toni, has just shared his
latest news. He mentions that it has been boring at school, and the female therapist
(‘FT’) reacts to this by saying that when she and the other therapist were kids they
thought that going to school was fun. This leads to a discussion about the quality of
food offered at school at that time. Morpho-syntactic disfluency is clearly present
in Jaakko’s speaking turns in lines 07–11, and in line 13 the male therapist (‘MT’)
requests clarification.

7 This sub-chapter is based on Wiklund and Laakso (2019, 2020).

3.3 The Role of Disfluencies in the Creation of Comprehension Problems 63

Example 3.18 :

01 Kalle: [no ] (.) silloin oli varmaan kou#luruokakin parem°paa°#;

well at that time food at school was probably also better

02 Toni: nii;

yeah

03 Jaakko: ((looks at Toni)) njää,

n-yeah

04 FT: ((bends towards Toni)) =mikä mikä oli,?

what what was

05 Toni: <kouluruoka>

food at school

06 FT: (smack) [njaa-a] ((turns to look at Jaakko))

um I see

07 Jaakko: [#ää no] silloin# silloin perusruoka #sil

well er at that time at that time basic food(-NOM) at that

well er at that time at that time basic food at that

08 tasolla oli alhasempi ku nykyään joten# (0.4) krh (0.4)

level was lower than nowadays so krh

level was lower than nowadays so

09 kouluruoka ei tartte #°olla ni°# .hh (0.4) niin hy- (0.3) °hy-°

the food at school(-NOM) no need to be that that go- go-

the food at school doesn’t need to be that that go- go-

10 oliv- (.) #ei tarttenu olla niin hyvää >koska se on samaa

wa- no needed be so good because it is same(-PAR)

wa- didn’t need to be so good because it is on the same

8The extract has been analysed in Lehtinen (2012) (in Finnish) as well as in Wiklund (2016)
and Wiklund and Laakso (2019) (in English). (As already mentioned, Lehtinen is the previous
surname of the author of this book.).
64 3 Disfluencies

11 luokkaa kuin ne< (2.0) >jos joku just tajus mitä mä sanoin<.

level-PAR as them if someone just understood what I said

level as them if anyone just understood what I said

12 (2.4)

13 MT: ((looks upwards)) samaa luokkaa kuin; ((turns to look at Jaakko)

on the same level as

14 Jaakko: .mth #no kun al- (.) (>kun oltii a jas<) taakseppäin#

well when lo- back in the old days

15 ruoka oli vähän < alhaisempaa>. (0.9)

food was a bit lower

16 #silleen (0.5) mh silleen kuing (.) °hy- pa-° (0.3) <makusta> ?

I mean like like how go- ba- taste

17 (0.3) ja > kouluruoka on< parempaa joten ne on samal linjalla

and the food at school is better so they are on the same level

18 (0.5) johtuen ajastansa. (1.4)

due to their time

19 FT: ((looks at Jaakko and nods)) hmm-m?

uh um

20 Jaakko: <eli sil periaatteella (0.8) silloin oli °parempaa°.>

so on that principle at that time it was better

21 (2.0)

22 MT: mm

um

23 FT: ((looks at Jaakko))(smack) jaaha

um I see
3.3 The Role of Disfluencies in the Creation of Comprehension Problems 65

In this extract, Jaakko has difficulty verbalising what he wants to say as he keeps
on repeating his words and reformulating his utterance. The trouble-source turn
includes several morpho-syntactic problems, such as wrong case endings, ambiguous
pronominal references, and cut-off syntactic structures (lines 07–11). For example,
he uses the nominative case instead of the genitive case (line 07: perusruoka →
perusruoan; line 09: kouluruoka → kouluruoan). The first utterance (lines 07–08)
also contains the ambiguous pronominal reference sil (‘that’, line 07). It is not clear
what this pronoun refers to, but it is possible that the speaker is using the adessive
form of the demonstrative pronoun se ‘it’ (sillä shortened sil) to replace the genitive
case ending of the word perusruoka (‘basic food’, line 07). The structure of the
second utterance, starting with the word kouluruoka (‘food at school’, line 09), is
remarkably disfluent. It includes false starts (hy-, hy-, and oliv-, lines 09–10) and
pauses. At the beginning, there is also a tense problem: the speaker uses the present
tense (ei tartte olla, “doesn’t need to be”, line 09) instead of the past tense. This is,
however, corrected by the speaker soon after (ei tarttenu olla, “didn’t need to be”, line
10), which makes the utterance more comprehensible. Naturally, the occurrences of
the wrong cases and cut-off syntactic structures are interrelated. That is, when the case
ending is not correct, the overall syntactic structure of the utterance does not work.
These morpho-syntactic problems seem to be caused by problems in the planning
of speech. When the speech is analysed element by element, it can be seen that the
speaker is changing the utterance all the time. Consequently, the choice of words,
case endings used, and the overall grammatical form of the utterance do not match,
and speech becomes disfluent and difficult to understand. This leads to a noticeable
silence (2.4 secs, line 12) and then to the therapist’s request for clarification (line
13).
The main cause of the comprehension problem is the ambiguous pronominal refer-
ence ne (‘them’, line 11) at the end of Jaakko’s second utterance. The clarification
request made by the male therapist (line 13) clearly targets this element as trouble-
some: the therapist repeats the expression samaa luokkaa kuin (‘on the same level
as’) used by Jaakko just before the ‘problem element’ of the turn, the demonstrative
pronoun ne (‘them’, line 11).
It is noteworthy that Jaakko is able to respond to the therapist’s request for clar-
ification in this case. This is in line with previous research results that show that
adolescents with ASD attempt to repair when confronted with a clarification request
(Geller, 1998; Ohtake et al., 2011; Volden, 2014). In lines 14–18, Jaakko gives a long
clarification of his previous turn. However, this turn is also remarkably disfluent. All
the utterances of the turn are grammatically incoherent. The turn also includes false
starts (al-, line 14; hy- and pa-, line 16) and a tense problem (line 20): the speaker
again uses the present tense (ja kouluruoka on parempaa joten ne on samal linjalla,
“and the food at school is better so they are on the same level”), instead of the past
tense. In this turn, Jaakko also seems to have more obvious difficulties in finding
words (line 18): the food cannot be ‘lower’ (although its level can) (cf. Norbury
66 3 Disfluencies

et al., 2010). This is also in line with previous observations that adolescents with
ASD may use unconventional forms in repairing their utterances (cf. Ohtake et al.,
2011).
Despite these problems, the female therapist nods and produces the discourse
particle ‘hmm-m’ with a rising pitch (line 19). Jaakko interprets the interactional
meaning of the discourse particle as a continuer and continues, adding more infor-
mation to clarify his point. This interactional strategy is successful, as both thera-
pists produce discourse particles, indicating—at least superficially9 —that they have
understood what Jaakko has tried to say (lines 22–23). Thus, even if the boy’s speech
includes many morpho-syntactic problems, he resolves the comprehension problem
with his own action. It is also noteworthy that Jaakko is himself orientated to the
potential comprehension difficulty of his recipients. Indeed, at the end of his turn
(line 11), after a 2 s pause with no responses from his recipients, Jaakko says jos joku
just tajus mitä mä sanoin (“if anyone just understood what I said”) thus displaying
his awareness of his problems with expression (Wiklund, 2016; Wiklund & Laakso,
2019).
Other sequential trajectories related to comprehension problems caused by disflu-
encies could also be found in the data involving Finnish-speaking participants with
ASD (Wiklund & Laakso, 2019). These trajectories include cases in which (1)
morpho-syntactic disfluency in speech production leads to a clarification request
by another group member with ASD, (2) morpho-syntactic disfluency in speech
production leads to a self-repair and to a therapist’s intervention, and (3) a turn is
understood in context despite the morpho-syntactic disfluency in speech production
that it includes (Wiklund & Laakso, 2019).
Disfluencies that occur in the control group data are mainly fillers and sound
prolongations, which do not cause comprehension problems (Wiklund & Laakso,
2020). In Example 3.2, the group is talking about their hobbies and free time, and
one of the boys, Pekka, states that he does not have much free time due to his frequent
training sessions and games (lines 01–03).

9 As the therapists do not comment at all on the contents of what the boy has just said, however,
it is not certain that they have completely understood it. Indeed, the discourse particles that the
therapists produce here constitute ‘displays of understanding’ rather than ‘proofs of understanding’
(Sacks 1992).
3.3 The Role of Disfluencies in the Creation of Comprehension Problems 67

Example 3.210 :

01 Pekka: mulkaan ei oo silleen niin

I-CLI don’t have PRT so

I don’t really have so

02 paljo (.) sitä (.) ku mul on treenei

much that because I have training sessions

much that [free time] because I have training sessions

03 ja pelei (0.8) koko ajan.

and games all the time

and games all the time

04 Teacher: >nii kun sun< (0.4) oli jalkapallo

yes football was your

05 sun laji eiks niin (.) kuinka usein

sport wasn’t it so how often

06 sulla on niitä.

do you have them

07 Pekka: no mulla on niinku: (0.6) öö neljä

well I have PRT er four

well I have like: (0.6) er four

08 (0.2) kertaa niinku viikossa >sit

times PRT a week then

times a week and then

10 The extract has been analysed in Wiklund and Laakso (2020).

68 3 Disfluencies

09 viikonloppusin< on pelei (0.6) °aina,°

on weekends are games always

on weekends there are always games

10 Teacher: no on se aika (0.6) aika usein ja Miikalla

well it is quite often and Miikka

11 on taas sulla on (0.4) sekä jääkiekko että

has you have both ice hockey and

12 jalkapallo.

football

13 Miikka: joo:,

yeah

14 Teacher: no ei viel ei tarvi valita.

well you don’t need to choose yet

15 Miikka: e:i,

no:

16 Teacher: mi-milloin sun pitää (0.4) milloin se hetki

whe- when do you when does the moment

17 tulee et nyt pitää jompaankumpaan panostaa.

come that now one has to invest one’s effort only in one of them

18 Miikka: mä veikkaan et niinku (0.8) viä seuraava kausi.

I guess that PRT still next season

I guess that like (0.8) still the next season

19 Teacher: mm-m,

mm-m

20 Miikka: ja sit pitää päättää.

and then one has to decide

3.3 The Role of Disfluencies in the Creation of Comprehension Problems 69

In this extract (Wiklund & Laakso, 2020), the speakers, Pekka and Miikka,
produce some fillers and pauses. Pekka first uses the filler silleen (a particle which
could be roughly translated as ‘that way’, line 01). This filler is typically related to
speech planning (Hakulinen et al., 2004: § 861) and also occurs here in a syntactically
well-positioned place. The general impression of his turn is that it is fluent, with no
cut-off utterances, such as searching, revision, and restarting, and the occurrence of
the filler and short pauses do not lead to comprehension problems. On the contrary,
the teacher reacts to Pekka’s turn with a follow-up question (lines 04–06) the contents
of which demonstrate that she has understood the preceding turn: as the boy has said
that he does not have much free time due to his training sessions and games (lines
01–03), the teacher asks him how often he has them.
During his second turn (lines 07–09), Pekka first produces the filler niinku: (a
particle which could be roughly translated as ‘like’ or ‘kind of’), which is prolonged
(line 07). This filler is also typically related to the planning of speech (Hakulinen
et al., 2004: § 861). Thus, the prolongation that it carries, as well as the pause (0.6 s)
and vocalisation öö (‘er’), occurring after it, emphasise this function. In addition, here
the filler and vocalisation are syntactically well positioned to display to the listener
that the speaker is thinking about continuing his utterance, thus enabling him to
maintain his speaking turn during planning. Shortly thereafter (line 08), there is yet
another occurrence of the niinku filler, which reflects similar speech planning. These
fillers do not create any comprehension problems: the contents of the beginning of
the turn produced by the teacher (no on se aika usein, “well it is quite often”, line
10) indicate that she has understood the preceding turn.
The teacher then addresses another participant, Miikka (lines 10–12), by stating
that Miikka plays both ice hockey and football. The boy answers affirmatively with
joo (‘yeah’) (line 13). The teacher states that Miikka does not yet need to choose
between the two sports (line 14), and Miikka answers e:i, (‘no:’) (line 15). After
that, the teacher asks Miikka when he has to make the choice between these two
sports (lines 16–17). The boy answers mä veikkaan et niinku (0.8) viä seuraava
kausi (“I guess that like (pause) still the next season”, line 18). This turn includes an
occurrence of the filler particle niinku that is followed by a rather long (0.8 s) pause,
and both indicate speech planning (Hakulinen et al., 2004: § 861). The filler and the
pause do not cause comprehension problems in this context: The teacher produces a
turn consisting of the particle mm-m (‘mm-m’, line 19), which indicates that she is
listening (Hakulinen et al., 2004: § 798), after which the boy continues ja sit pitää
päättää (“and then one has to decide”, line 20).
To conclude, Example 3.2 illustrates how the preadolescents of the control group
use some filler particles and pauses while planning their speech, but their speaking
turns are nevertheless fluent and easily understood, and they do not need support
from their teacher to answer the questions coherently.
70 3 Disfluencies

3.4 Conclusion

According to Wiklund and Laakso (2020), the durational mean proportions of disflu-
encies and ungrammatical expressions are greater in the speech of participants with
ASD (26.4%) than in the control group (15.5%). Thus, the study confirms the find-
ings by Lake et al. (2011), Scaler Scott et al. (2014), and Shriberg et al. (2001) that
more disfluencies occur in the speech of individuals with ASD than in the speech
of matched controls. Individual variation in the amount of disfluency is high within
both groups, which is in line with the results reported on large individual ranges of
speech disfluency in adult speakers (e.g., McDougall & Duckworth, 2017).
Furthermore, a qualitative difference can be found between the ASD group and the
control group (Wiklund & Laakso, 2020): the speakers with ASD produce long and
complex disfluent turns with word searches, self-repairs, false starts, fillers, sound
prolongations, inconsistent syntactic structures, and grammatical errors, whereas the
control group members mainly produce fillers and sound prolongations. Previously,
a qualitative difference has been reported by Lake et al. (2011), who demonstrated
that adult speakers with AS produce fewer listener-oriented disfluencies such as
filled pauses and revisions and more silent pauses and repetitions that reflect the
speaker’s speech processing. This result is in line with the findings presented in this
chapter in the sense that the disfluencies produced by the Finnish-speaking informants
with ASD also seem to reflect the speaker’s own grammatically disturbed speech
processing, whereas the disfluencies (fillers and sound prolongations) produced by
the control group members are more listener-oriented (Wiklund & Laakso, 2020).
Control group members produce more disfluent speaking turns with one or two
mild disfluencies such as the filler particle niinku (appr. ‘like’) and the vocalisation
öö (‘er’ or ‘um’) than preadolescents with ASD, who produce more speaking turns
with grammatical errors and complex disfluencies (Wiklund & Laakso, 2020). A
qualitative difference in the types of disfluencies has also been observed from 4- to 8-
year-old children with ASD and their age-matched controls: in a study by MacFarlane
et al. (2017), control children produced more fillers (e.g., um, uh) than children with
ASD. It is also important to note that in the studies by Wiklund and Laakso (2019,
2020), disfluencies and grammatical incoherence tended to co-occur in the speech
of preadolescents with ASD, whereas similar co-occurrences were not observed in
the control group (Wiklund & Laakso, 2020).
In addition, due to the quantitative and qualitative differences, the disfluencies and
the ungrammatical expressions that occurred in the interactions of the participants
with ASD also caused comprehension problems during the conversations, whereas
the control group did not experience comprehension problems (Wiklund & Laakso,
2019, 2020). The control group’s disfluencies were similar to the hesitations that
normally fluent speakers (c.f., McDougall & Duckworth, 2017) produce in order
to plan expressions and to help listeners anticipate a continuation. The participants
with ASD in this study often produced a combination of disfluencies and grammatical
errors, and this caused more profound problems for the listeners to understand what
was said. As a consequence, the disfluency of participants with ASD had more serious
References 71

effects on the interactional flow of their conversations in comparison to the age- and
gender-matched controls.
The speech of the boys with ASD in these data was also characterised by morpho-
syntactic problems such as incorrect case endings, ambiguous pronominal refer-
ences, disconnected syntactic structures, and verb tense problems. These difficul-
ties in constructing fluent utterances can also be connected to the previous find-
ings of problems in higher level cognitive processes such as attention, working
memory, and executive function, which have been observed by individuals with
high-functioning autism (e.g., Joseph et al., 2005; Kenworthy et al., 2008). Thus,
future research could also address the connections between these cognitive abili-
ties and the lexico-grammatical construction of conversational speech by individuals
with ASD (Wiklund & Laakso, 2020).
The studies presented in this chapter (Wiklund & Laakso, 2019, 2020) point out
that it would be very important to study the conversational interactions of partici-
pants with ASD more extensively. Further studies may reveal how the interactional
difficulties arise from the morpho-syntactic level of constructing coherent speaking
turns and finding correct lexical items, challenging the view of ASD as a disorder
of the pragmatic use of language. In the future, it would also be important to study
mixed groups including participants with and without ASD to examine whether the
quantitative and qualitative differences discovered in the studies presented in this
chapter can also be found in mixed groups (Wiklund & Laakso, 2020).

References

Allen, M. L., Haywood, S., Rajendran, G., & Branigan, H. (2011). Evidence for syntactic alignment
in children with autism. Developmental Science, 14, 540–548.
Ambrose, N. G., & Yairi, E. (1999). Normative disfluency data for early childhood stuttering.
Journal of Speech, Language & Hearing Research, 42, 895–909.
ATLAS.ti Scientific Software Development GmbH. (2021). Atlas.ti. [Computer program]. https://
atlasti.com/.
Boersma, P., & Weenink, D. (2021). Praat: doing phonetics by computer. [Computer program].
http://www.praat.org/.
Byrd, C. T., Logan, K. J., & Gillam, R. B. (2012). Speech disfluency in school-age children’s
conversational and narrative discourse. Language, Speech & Hearing Services in Schools, 43,
153–163.
Cummings, L. (2014). Pragmatic disorders. Springer.
Engelhardt, P. E., Alfridijanta, O., McMullon, M. E. G., & Corley, M. (2017). Speaker-versus
listener-oriented disfluency: A re-examination of arguments and assumptions from autism
spectrum disorder. Journal of Autism and Developmental Disorders, 47, 2885–2898.
Geller, E. (1998). An investigation of communication breakdowns and repairs in verbal autistic
children. The British Journal of Development Disabilities, 44(87), 71–85.
Hakulinen, A., Vilkuna, M., Korhonen, R., Koivisto, V., Heinonen, T. R., & Alho, I. (2004). Iso
suomen kielioppi. SKS.
Hutchby, I., & Wooffitt, R. (2008). Conversation analysis (2nd edn.). Polity Press.
Joseph, R., McGrath, L., & Tager-Flusberg, H. (2005). Executive function and its relation to language
ability in verbal school-aged children with autism. Developmental Neuropsychology, 27, 361–379.
72 3 Disfluencies

Kenworthy, L., Yerys, B., Anthony, L., & Wallace, G. (2008). Understanding executive control
in autism spectrum disorders in the lab and in the real world. Neuropsychological Review, 18,
320–338.
Lake, J. K., Humphreys, J. R., & Cardy, S. (2011). Listener versus speaker oriented aspects of
speech: Studying the disfluencies of individuals with autism spectrum disorders. Psychonomic
Bulletin and Review, 18, 135–140.
Lehtinen, M. (2012). “Aiheesta toiseen” – Ymmärrysongelmatilanteita Asperger-nuorten kuntou-
tuskeskusteluissa. In L. Leskelä & C. Lindholm (Eds.), Haavoittuva keskustelu. Keskustelu-
nanalyyttisia tutkimuksia kielellisesti epäsymmetrisestä vuorovaikutuksesta (pp. 235–277).
Kehitysvammaliitto.
MacFarlane, H., Gorman. K., Ingham, R., Presmanes Hill, A., Papadakis, K. Kiss, G., & van
Santen, J. (2017). Quantitative analysis of disfluency in children with autism spectrum disorder
or language impairment. PLOS one.
Manning, W. (2010). Clinical decision making in fluency disorders. Delmar.
McDougall, K., & Duckworth, M. (2017). Profiling fluency: An analysis of individual variation in
disfluencies in adult males. Speech Communication, 95, 16–27.
McGregor, K. K., Berns, A. J., Owen, A. J., Michels, S. A., Duff, D., Bahnsen, A. J., & Lloyd, M.
(2012). Associations between syntax and the lexicon among children with or without ASD and
language impairment. Journal of Autism and Developmental Disorders, 42, 35–47.
Naigles, L. R., Kelty, E., Jaffery, R., & Fein, D. (2011). Abstractness and continuity in the syntactic
development of young children with autism. Autism Research, 4, 422–437.
Norbury, C. F., Griffiths, H., & Nation, K. (2010). Sound before meaning: Word learning in autistic
disorders. Neuropsychologia, 48, 4012–4019.
Ohtake, Y., Wehmeyer, M., Nakaya, A., Takahashi, S., & Yanagihara, M. (2011). Repair strategies
used by verbal students with autism during free play. Focus on Autism and Other Developmental
Disabilities, 26(1), 3–14.
Park, H., Park, H., Won, D., Kim, H., Song, S., & Sim, H. S. (2018). Disfluency characteristics of
school-age children with high-functioning autism spectrum disorders. Communication Sciences &
Disorders, 23(2), 436–450.
Pirinen, V., Dindar, K., Hurtig, T., Ebeling, H., & Loukusa, S. (2021). Puheen tyypilliset sujumatto-
muudet autismikirjon henkilöillä ja heidän verrokeillaan. Oral presentation at XLVII Kielitieteen
päivät 2021 (The XLVII Annual Conference on Linguistics 2021), Tampere (Finland).
Pirinen, V. (2022). Puheen sujumattomuudet autismikirjon henkilöillä. In S. Loukusa, T. Hautala &
A.-K. Tolonen (Eds.), Sujuvaa vai sujumatonta? Puheen ja kielen sujuvuutta tutkimassa. [Puheen
ja kielen tutkimuksen yhdistyksen julkaisuja 54] (pp. 64–74). Puheen ja kielen tutkimuksen
yhdistys.
Saalasti, S., Lepistö, T., Toppila, E., Kujala, T., Laakso, M., Nieminen-von Wendt, T., von Wendt,
L., & Jansson-Verkasalo, E. (2008). Language abilities of children with Asperger syndrome.
Journal of Autism and Developmental Disorders, 38, 1574–1580.
Sacks, H. (1992). Lectures on conversation (Vol. 1 & 2). Blackwell.
Scaler Scott, K., Tetnowski, J. A., Flaitz, J. R., & Yaruss, J. S. (2014). Preliminary study of disflu-
ency in school-aged children with autism. International Journal of Language & Communication
Disorders, 49(1), 75–89.
Schegloff, E., Jefferson, G., & Sacks, H. (1977). The preference for self-correction in the
organization of repair in conversation. Language, 53(2), 361–382.
Scott, K. S., Grossman, H. L., Abendroth, K. J., Tetnowski, J. A., & Anddamico, J. S. (2006).
Asperger syndrome and attention deficit disorder: clinical disfluency analysis. In J. Au-Yeung &
M. M. Leahy (Eds.), Research, treatment, and self-help in fluency disorders: New horizons.
Proceedings of the fifth world congress on fluency disorders (pp. 273–278). International Fluency
Association.
Shriberg, L., Paul, R., McSweeney, J., Klin, A., Cohen, D., & Volkmar, F. R. (2001). Speech and
prosody characteristics of adolescents and adults with high functioning autism and Asperger
syndrome. Journal of Speech, Language and Hearing Research, 44, 1097–1115.
References 73

SIL Glossary of Linguistic Terms. (2021). https://glossary.sil.org/.

Sisskin, V. (2006). Speech disfluency in Asperger’s Syndrome: Two cases of interest. Perspectives
on Fluency and Fluency Disorders, 16(2), 12–14.
Tumanova, V., Conture, E. G., Lambert, E. W., & Walden, T. A. (2014). Speech disfluencies of
preschool-age children who do and do not stutter. Journal of Communication Disorders, 49,
25–41.
Volden, J. (2004). Conversational repair in speakers with autism spectrum disorder. International
Journal of Language & Communication Disorders, 39(2), 171–189.
Wiklund, M. (2016). Interactional challenges in conversations with autistic preadolescents: The
role of prosody and nonverbal communication in other-initiated repairs. Journal of Pragmatics,
94, 76–97.
Wiklund, M., & Laakso, M. (2019). Ungrammatical utterances and disfluent speech as causes
of comprehension problems in interactions of preadolescents with highly functioning autism.
Clinical Linguistics & Phonetics, 33(7), 654–676.
Wiklund, M., & Laakso, M. (2020). Comparison of disfluent and ungrammatical speech of pread-
olescents with and without ASD. Journal of Autism and Developmental Disorders, 51(8),
2773–2789.
Chapter 4
Comprehension Problems

Abstract The studies on comprehension problems carried out within the framework
of this project (Lehtinen, 2012; Wiklund, 2016; Wiklund & Laakso, 2019, 2020)
confirm the results of previous studies (e.g., Geller, 1998; Keen, 2003), according to
which children with ASD often experience communication breakdowns in their inter-
actions. Unlike previous studies, however, in these data, the comprehension prob-
lems did not lead to problem behaviour (Brady et al., 1995; Keen, 2003; Wetherby
et al., 1998). The results are in many respects in line with those of Geller (1998).
Indeed, according to Geller (1998), most of the communication breakdowns in the
interactions of school-aged children with ASD are caused by a lack of referential
clarity, which is also typical of the data studied within the framework of this project
(Lehtinen, 2012; Wiklund, 2016). In addition, ‘topical discontinuities’ constitute
one of the most common causes of comprehension problems in these data (Lehtinen,
2012; Wiklund, 2016). ‘Ungrammatical utterances’ and ‘disfluencies’ cause a certain
number of comprehension problems between the participants as well (Wiklund &
Laakso, 2019, 2020). The most common cause of comprehension problems in these
data was ‘overly literal understanding of speech’. In 84.1% of instances, there was
no eye contact between the speaker producing a trouble-source turn and the person
initiating the repair sequence (or making an outright correction). Sometimes the lack
of eye contact was associated with overlapping speech (38.1%).

Keywords Comprehension problems · Topical discontinuities · Overly literal

understanding · Non-verbal communication · Conversation analysis · Interaction ·
ASD · Autism

4.1 Overview of Previous Research on Conversational

Repairs in the Interaction of Persons with ASD

It is well known that communication breakdowns occur frequently in the interactions

of persons with ASD. Indeed, according to Keen (2003), there is evidence to suggest
that children with ASD experience communication breakdowns more frequently
than their typically developing peers due to joint attention deficits, impaired ‘theory
of mind’, and reliance on prelinguistic forms of communication. Communication

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 75
M. Wiklund, Speech and Interaction of Preadolescents with Autism Spectrum Disorder,
Prosody, Phonology and Phonetics, https://doi.org/10.1007/978-981-19-8117-3_4
76 4 Comprehension Problems

breakdowns may also cause problem behaviour (Brady et al., 1995; Keen, 2003;
Wetherby et al., 1998).
Studies on the communication breakdowns and ‘conversational repairs’—with the
help of which the participants of the interaction try to solve the breakdown of inter-
subjectivity—in the interactions of children with ASD are still scarce. Some studies
have, however, been carried out. Geller (1998), for example, studied communica-
tion breakdowns and repairs in the interactions of five school-aged, 7–12 year-old
children with ASD. The data came from free play sessions involving spontaneous
adult–child interactions with toys. The results of the study show that the utterances
of children with ASD often involve problems of form, such as lack of intelligibility
or lack of audibility. Errors in syntax or morphology, in turn, did not lead to the
listeners’ comprehension problems in Geller’s (1998) study. Most of the communi-
cation breakdowns were caused by a lack of referential clarity. Pragmatic errors such
as unclear coding of intentionality and linguistically unmarked topical shifts (i.e., lack
of explicitness when shifting topics) also caused some of the children’s communica-
tion breakdowns. Different patterns combining these problems (for example: ‘lack
of referential clarity’ + ‘lack of explicitness when shifting topics’) also occurred
recurrently.
When Geller (1998) analysed the adults’ requests for clarification in her data,
she found that 40% of the requests were confirmation requests, 33% were specific
requests for information, and 23% were indefinite requests. When the children’s
reactions to these requests for clarification were examined, it was found that 53%
of the children’s repairs fell into the categories of indeterminate, ambiguous, and/or
no attempts to repair the message. The rest of the cases (47%) were coded as infor-
matives, acknowledgements, and/or non-language repairs. In sum, Geller’s (1998)
results show that in 73% of the instances, the children’s reactions contained some
attempt to resolve and/or respond to the adult’s clarification request. Thus, the
study suggests that children with ASD attempt to repair most of the communication
breakdowns that occur in spontaneous adult–child interactions.
Volden (2004: 173) remarks that both linguistic and social–cognitive skills are
required from a speaker to accomplish a repair appropriately. Firstly, it is necessary
to understand linguistically the request for clarification produced by the listener,
as well as how it relates to the speaker’s previous utterance. Secondly, the speaker
needs to have expressive language skills in order to be able “to modify successfully
the linguistic signal once a repair strategy has been selected” (p. 173). Thirdly, on
the social–cognitive level, the speaker needs to be able to evaluate what informa-
tion was missing from his/her previous utterance that created the comprehension
problem. This process requires “an ability to judge another person’s state of knowl-
edge” (p. 174), which, in turn, requires, according to the author (p. 174), “an intact
‘theory of mind’”. ‘Theory of mind’ is defined by Baron-Cohen (2001: 169) as being
“able to reflect on the content of one’s own and others’ minds”. It is well known that
an impaired ‘theory of mind’ constitutes a fundamental deficit in ASD.
Volden (2004) studied the conversational repairs of nine school-aged children with
ASD and their matched controls. The communication breakdowns were simulated
by an unfamiliar examiner. The results of the study show that children with ASD
4.2 Causes of Comprehension Problems in Group Therapy Sessions … 77

behave similarly to age-matched controls in responding to requests for clarification.

The children with ASD use a variety of repair strategies, and they are able to add
information if the communication breakdown persists. However, according to the
results (Volden, 2004), children with ASD are significantly more likely than the
controls to respond in an inappropriate way to a request for clarification.
Keen (2005) studied the repair strategies of six children with ASD aged 2–5 years
and with fewer than 10 words or signs. The data came from naturally occurring
mother–child dyads videotaped at home. According to the results, in most inter-
actions with their mothers, the informants attempted to repair the communication
breakdowns that occurred. Most informants used both non-problematic and problem-
atic behaviours in their repairs. The children were also more likely to use repetitions
and substitutions than augmentations as a repair strategy.
Philip (2008), in turn, studied the conversational repair responses of three adoles-
cents with ASD to communication breakdowns that occurred naturally in conversa-
tions. Although the informants also differed from each other concerning the repairs,
the results show that all three participants of the study had communication break-
downs caused by the presence of ambiguous referents, inadequate information, and
irrelevant information (p. 114). All three participants were, however, able to respond
to clarification requests presented by their listeners (p. 120), which is in line with
Geller’s (1998), Keen’s (2005), and Volden’s (2004) results, for example.
Abbas (2021) studied talk-in-interactions between neurotypical persons and chil-
dren with ASD in both institutional and home settings. His results show that
behavioural problems, lack of attention, and lack of linguistic knowledge or problems
in expressing knowledge adequately were the most common causes of communica-
tion breakdowns between neurotypical persons and children with ASD (p. 140).
‘Repetition’ was the most frequent strategy used by the therapists to repair a
communication breakdown. The second most frequent strategy was ‘modification’,
consisting of repeating the same question with alternations (p. 141). According to
the study by Abbas (2021: 141), these strategies were successful in some cases.
Nevertheless, some communication breakdowns remained unrepaired.

4.2 Causes of Comprehension Problems in Group Therapy

Sessions Involving Preadolescent Boys with ASD

The objective of the studies (Lehtinen1 , 2012; Wiklund, 2016) presented in this
sub-chapter is to shed light on the causes of comprehension problems in group
therapy sessions involving Finnish-speaking boys with ASD (see Sect. 1.2.1). As
already mentioned, in the studies presented in this book, comprehension problems
are materialised by other-initiated repairs in which the repair process is started
by the recipient of the problematic utterance (Schegloff et al., 1977; SIL Glos-
sary of Linguistic Terms, 2021). Repair in the sense of CA was first defined by

1 As already mentioned, Lehtinen is the previous surname of the author of this book.
78 4 Comprehension Problems

Schegloff et al. (1977) as “the set of practices whereby a participant interrupts the
ongoing course of action to attend to possible trouble in speaking, hearing, or under-
standing the talk” (Kitzinger, 2014: 229) (for a more detailed description of repairs in
the sense of CA, see Sect. 1.3). According to the results carried out within the frame-
work of this project (Lehtinen, 2012; Wiklund, 2016), overly literal understanding
of speech (see Sect. 4.2.1) and topical discontinuities (Sect. 4.2.2) are by far the
two most common causes of the comprehension problems occurring in the data. In
addition, comprehension problems are also caused in these data by, e.g., deficiencies
of verbal expression (such as disfluencies, morpho-syntactic mistakes, ambiguous
pronominal references, and disconnected syntactic structures) (Wiklund & Laakso,
2019) (see Sect. 3.3) as well as by prosodic features (Wiklund, 2016) (see Sect. 2.3)
and speech overlaps (Lehtinen, 2012; Wiklund, 2016). In most cases, there is no
eye contact between the speaker producing a trouble-source turn and the participant
initiating the repair sequence (Wiklund, 2016) (see Sect. 4.2.3).

4.2.1 Overly Literal Interpretation of Speech

As already mentioned (see Sect. 2.3), the group therapy sessions with Finnish-
speaking participants with ASD include 72 other-initiated repairs. Most (86.1%) of
them are repair initiations in which a repair procedure is started by an interlocutor,
but the outcome of the repair is left for the speaker who produced the trouble-source
turn. The remaining cases of the other-initiated repairs (13.9%) are outright correc-
tions. In an outright correction, the interlocutor not only indicates that something
said earlier in the conversation needs to be corrected, complemented, or explained,
but he or she also presents another version (which he or she finds more correct or
more complete) of the contents of the trouble-source turn (Schegloff et al., 1977;
Sorjonen, 1997). Most of the repair sequences of these data occur between the ther-
apists and the boys (84.7%), but sometimes repair sequences also occur between the
boys (11.1%)—occasionally even between the two therapists (4.2%).
Overly literal understanding of speech is the most common cause of compre-
hension problems in these data: it causes the comprehension problem in 33 cases,
which is 45.8% of all other-initiated repairs in these data (Lehtinen, 2012; Wiklund,
2016). Situations in which the informants with ASD interpret a turn produced by a
therapist literally and fail to notice an implicit cue that it includes are particularly
typical of these data. The following example illustrates this type of situation.2 In this
extract, the male therapist (‘MT’) tries to resolve a problem of non-verbal interaction
encountered by one of the boys, Heikki. Indeed, during the whole session, Heikki
has been staring at a mirror on the opposite side of the room instead of looking at
his interlocutors (Wiklund, 2012) (see 5.3). In Example 4.1, Heikki wants to take the
floor (line 02), and the therapist tries to make him look at his interlocutor with an
implicit cue included in his turn (line 04). The verbal cue is associated with pointing
hand gestures (line 05).

2 The same extract has been analysed in Finnish in Lehtinen (2012: 255–258).
4.2 Causes of Comprehension Problems in Group Therapy Sessions … 79

Example 4.1:

01 MT: onks ↑pojilla mitäs sitte ky↑syisitte?

do the boys have what would you ask then

02 ((Heikki raises his hand and stares at the mirror straight in 03

front of him; FT looks at MT))

04 MT: öö #k- m-# (0.4) ↑tääl↓lä on ↑kukas (0.4) >keneltä sä kysyt<,

er k- m- here is who who do you ask

05 ((FT looks at Heikki; MT glances at Rikhard and points at him 06

with his hand; Heikki glances at MT))

07 (.)

08 Heikki: öö[:,]

er

09 MK: [He]ikki ↑keneltä sä kysyt. ((Heikki and Mikael look

Heikki who do you ask

10 straight forwards, the others look at Heikki))

11 (.)

12 Heikki: öö ↑Rikhardilt?

er Rikhard

13 MT: >kat- ja #a-#< missä päin R[ikhard o]n. ((tries to make

loo- and a- in which direction Rikhard is

14 eye contact with Heikki, makes gestures towards Rikhard))

15 Rikhard: [khhhhh,]

16 (0.4)

17 Rikhard: ↓°tä[äl°],

here

18 Heikki: [ni]in tai ↑siis että:,

yes or well I mean

19 MT: ↑missä ↓päin, (.) £heh He(h)ikki [Rikhard on,£]

in which direction heh He(h)ikki Rikhard is

20 ((makes a gesture with his hand again, laughs))

21 Rikhard: [↑lännessä ] vai idässä:]

in the West or in the East

80 4 Comprehension Problems

22 FT: [missäs on R]ikhard ↑istuu]

where is Rikhard sitting

23 tässä pöy#dän ääressä#. ((looks at Heikki and leans towards him))

at this table

24 MT: katsotko, (.) ↑niin (.) ↑hänen #suuntaansa#;

would you look in his direction

25 ((Heikki glances in the direction of Rikhard and MT; MT makes a

26 gesture in the direction of Rikhard))

27 Heikki: ai [↑niis ] siis;

oh yes well

28 FT: [£↑mmh,£]

29 MT: £NII, (.) °nii,° katso ↑hänen£ suun#taansa ku sä# [ky]syt,

yes yes look in his direction when you ask

30 ((again makes a gesture towards Rikhard, puts a hand on his

31 hips, moves his legs))

32 Heikki: [ai,]

oh

33 MT: ni,

yeah

34 Heikki: >nii tota:<,? (.) .hhhh ö[ö:, ] ((keeps glancing at

yeah well er

35 Rikhard and the mirror in front of him))

36 MT: [mhh,]

37 (1.0)

38 Heikki: mi↑käs se, (.) mikäs se [paikan n]imi oli? ((looks straight ahead

at the

what was what was the name of the place

39 mirror))

40 MT: [° Heikki.°] ((tries to make eye

41 contact with Heikki, points at Rikhard, smiles; FT looks at

42 Heikki))
4.2 Causes of Comprehension Problems in Group Therapy Sessions … 81

43 (.)

44 Rikhard: ↑TEuva. ((leans forwards))

45 (0.4)

46 Heikki: mil↑lasta ↓siel on.

what is is like there

47 (0.4)

48 Rikhard: no se=om ↑Pohjam↓maalla sinne:, (.) ö kestää ajamiseen, (.)

well it is in Ostrobothnia to drive there it takes

49 joku:, (.) >s' o'< vähän niinkuj Jyväskylä ↑kaks,

something like it is a bit like Jyväskylä two

50 ((MT looks at Rikhard and sighs; FT looks at MT))

51 (0.4) ö puol ↓tuntii tai s'tten, (.) .hhh (0.6) ↑kaks tuntii ja

er half an hour or then two hours and

52 puol ↓tuntii, (.) j- se, h (.) ttai s'te kolme tun°tia°; (0.4)

a half an hour a- it h or then three hours

53 sama ku on Helsinki Jyv°äskylä°, ((leans backwards)) (2.0)

the same as Helsinki Jyväskylä

Heikki does not, however, understand the male therapist’s hint. First, he produces
a particle turn consisting of the hesitation marker öö (‘er’) (line 08), which can be
classified here as an open class repair initiator, because it marks the whole preceding
turn as problematic. The therapist reacts to this by repeating a part of his previous
turn: Heikki keneltä sä kysyt (‘Heikki who do you ask’) (line 09). Heikki understands
what is literally asked from him, but he does not understand the implicit hint included
in the turn. This can be deduced from the fact that he gives a literal answer to the
question (öö Rikhardilt, ‘er Rikhard’, line 12) but still does not turn to look at his
interlocutor. The male therapist reformulates his hint and adds a series of pointing
gestures to it: kat- ja a- missä päin Rikhard on (‘loo- and a- in which direction
Rikhard is’, lines 13–14). After a while, he repeats his hint with gestures (missä päin
heh He(h)ikki Rikhard on, ‘in which direction heh He(h)ikki Rikhard is’, lines 19–
20). Heikki still does not understand the hint. The hints produced by Rikhard (tääl,
‘here’, line 17; lännessä vai idässä, ‘in the West or in the East’, line 21) and the
female therapist (missäs on Rikhard istuu tässä pöydän ääressä, ‘where is Rikhard
sitting at this table’, lines 22–23) are not successful either.
82 4 Comprehension Problems

The comprehension problem is resolved only when the male therapist says explic-
itly what he has tried to hint (katsotko niin hänen suuntaansa, ‘would you look in his
direction’, line 24; nii nii katso hänen suuntaansa ku sä kysyt, ‘yes yes look in his
direction when you ask’, 29). The particle chain ai niin (‘oh yes’, line 27) occurring
in Heikki’s turn indicates that he has heard about this topic before (Hakulinen et al.,
2004: § 859). The connective particle siis (‘well’, line 27), in turn, typically starts
some conclusion based on what has been said before (Hakulinen et al., 2004: § 1132),
but the turn remains syntactically incomplete. The connective particle is produced
with a falling pitch, which may indicate that Heikki has stopped to see if the male
therapist will continue to speak. Thus, the comprehension problem does not seem
to be completely resolved at this stage. The dialogue particle ai (‘oh’), produced
with a flat pitch, occurring in Heikki’s next turn (line 32) also suggests this, because
the particle in question indicates that Heikki is still treating the male therapist’s turn
(line 29) as new information and that he would still give space for the male therapist
for an additional explanation (Hakulinen et al., 2004: § 798). However, the male
therapist marks himself as a recipient with the dialogue particle ni (‘yeah’, line 33)
(Hakulinen et al., 2004: § 798). The comprehension problem seems to be resolved,
because during his next turn (lines 34–35) Heikki glances several times at Rikhard.
The problem of non-verbal interaction remains unresolved, however, because Heikki
continues staring at the mirror straight in front of him: during his next turn (lines
38–39), his gaze is again fixed on the mirror.
It is noteworthy that the therapist supports his turns with pointing gestures, which
is in line with the instructions of easy language (Kartio, 2009: 23–24). He also
repeats what he has said and reformulates it, when he has noticed the creation of
a comprehension problem (Kartio, 2009: 20). However, the example clearly brings
up the fact that persons with ASD tend to have difficulties understanding implicit
messages in interaction. Because the informant in question faces particular challenges
with self-monitoring a feature of non-verbal interaction, this may make it more
difficult for him to understand the hint. The group situation may also make it more
challenging for the participant to concentrate on such things.

4.2.2 Topical Discontinuities

According to Drew (1997), understanding is mostly a sequential phenomenon that

concerns units that are larger than a turn. Comprehension problems are most often
caused not by syntactic or referential problems occurring inside a turn but by the
participants’ difficulty in understanding how something that is being said is related
to the topic being discussed or to the ongoing action (Drew, 1997). It has been
shown that children with ASD have a tendency to respond in a non-contingent (i.e.,
4.2 Causes of Comprehension Problems in Group Therapy Sessions … 83

off-topic) manner in conversation (Hale & Tager-Flusberg, 2005a). However, when

these children’s use of topic-related contingent utterances was studied by Hale and
Tager-Flusberg (2005b), the authors found that in the course of one year, children
with ASD made significant progress in the ability to maintain a topic of discourse.
In line with Hale’s and Tager-Flusberg’s (2005a) study, Geller (1998) found that
lack of explicitness when shifting topics causes a certain number of communication
breakdowns in the interactions of children with ASD.
Indeed, in the data studied within the framework of this project (see Sect. 1.2),
topical discontinuities also constitute a common cause of comprehension problems:
26 occurrences (36.1%) out of the 72 repairs that can be found in the Finnish data are
related to topical discontinuities (Lehtinen, 2012; Wiklund, 2016). These occurrences
include cases in which a child with ASD suddenly says something surprising, and the
other participants find it difficult to understand how the turn should be interpreted,
because its relationship with the ongoing topic and/or action is not obvious. For
example, the data include a passage in which Heikki suddenly asks nostetaaks me
nyt kädet ylös vai (‘do we now raise our hands up or’) (Lehtinen, 2012: 243–244).
The turn is linguistically and phonetically completely correct and intelligible, but it
is impossible to understand it without knowing that the boy refers to his question.
(Later the boy explains that he was referring to the two video cameras, which looked
like big guns to him.) In another passage, Kalle (see Sect. 3.3) suddenly brings up
‘food at school’ (kouluruoka) which is related to the topic under discussion, but
the sequential relevance of the referent is not obvious. In other words, it is not
obvious why ‘food at school’ is brought up at the moment it is mentioned, and how
it should be interpreted with regard to what has been said before. In a third passage,
Jaakko presents analyses on the nature of ‘pushing’ (whether pushing is generally
considered bullying) and a remark (se on meidän luokalta, ‘he is from our class’)
that are disconnected from the topical point of view and thus difficult for the other
participants to interpret (Lehtinen, 2012: 258–262). It is also typical in these data
that a boy with ASD presents a remark that is relevant to the topic or the ongoing
action, but the turn does not include enough contextualisation cues that would make
it possible for the others to interpret the turn. Example 4.2 illustrates a case of this
type.3

3 The same extract has been analysed in Finnish in Lehtinen (2012: 263–265).
84 4 Comprehension Problems

Example 4.2:

01 MT: .hhhh tota ↑tota. (.) ↑nyt me päätettiin, (.) Reetan kanssa

well well now we decided with Reetta

02 että, (.) .hhh kum meillä on, (.) ↑usein ollun niitä, (0.4)
that as we have often had those

03 niitä, (.) .hhh niill on, #auöö#, joku ↑moniste ollum mitä me

those they have some handouts have had that we

04 ollaan k- käytyl läpi ja: täytetty ja, (.) sittek k[eskustelt]u

have gone through and filled in and then discussed

05 ↑siitä niin, .hh

it so
06 Rikhard: [hhhhhhhhh]

07 Heikki: t[ällä k]ertaa ei kyllä ole nyt?

this time now there is none

08 MT: [↑nyt, ]

now

09 MT: ↑täl↓lä ker↑taa ei ↓ole. hhhhmmm .mthhh ja:

this time there is none and

10 Heikki: ↑↑harmi[m paikka.]

it’s a pity

11 ((MT looks over his shoulder, collects the cards he had put on

12 the table and gets up))

13 Rikhard: [krhm] >tuo o seiskytluvu°lta°<; ((osoittaa yhtä

that one is from the ‘70s ((points at one of

14 the cards; Mikael looks at the cards in MT’s hands))

15 (0.6)

16 MT: on:↓ko=se seitkyt ↑mistäs [sä niim päättelet,] ((goes

is it from the ‘70s why do you think so

17 back to the bookshelf, leaves some of the cards on a chair

18 next to the bookshelf))

19 FT: [>eks se oo< vähäv £↑vaiht]el[ua£,]

isn’t it nice for a change
4.2 Causes of Comprehension Problems in Group Therapy Sessions … 85

20 ((smiles and looks at Heikki))

21 Rikhard: [vaat]tei°sta°.

because of the clothes

22 (0.4)

23 MT: aha?

aha

24 (.)

25 Rikhard: mhhm [ja ↑ka]mpauksista.

and the haircuts

26 Heikki: [(£↑↑>se=oli<£)]

it was

27 (0.6)

28 Heikki: ↑@hei,@

hey

29 MT: vaatteista ja kampauksista. .hh sä voit,? .hh [olla i]han

clothes and haircuts you can be quite

30 ((comes back with a few cards and sits down))

31 Rikhard: [krhhhh]

32 MT: ↑oi↓keassa että se on [suurim ]piirtein, hh .hh aika

right that it is approximately

33 ((MT ja FT smile at each other; Heikki looks at the

34 mirror straight in front of him))

35 FT: [mhe he,]

36 MT: lähellä seitkyt[£lukua näi]tten kuvien, hh .hh ↑kuvien

close to the ‘70s when these pictures pictures

37 ((hides part of the cards close to himself and holds one

38 card so that the others can see it))

39 FT: [mhe he he?]

40 MT: tekoaika.£ hh .hh mut, (.) ↑näätsä Heikki tota:, (.)

were made but do you see Heikki er

86 4 Comprehension Problems

In this extract, the male therapist has just taken out cards with drawn pictures on
the basis of which the group is supposed to discuss bullying at school. Rikhard points
at one of the cards in the therapist’s hand and says “that one is from the ‘70s’” (tuo
o seiskytluvulta, line 13). The speaker expresses clearly with a pointing gesture that
his attention is directed towards a certain card. However, the turn does not contain
a sufficient interpretive framework, because it does not include any background
information about what the remark is based on. Indeed, the male therapist asks
explicitly for clarification in his repair initiation (line 16): “is it from the ‘70 s why
do you think so’” (onko se seitkyt mistäs sä niim päättelet). After this, Rikhard gives
a relevant justification for his remark (line 21: ‘because of the clothes’, vaatteista;
line 25: ‘and the haircuts’, ja kampauksista), and the therapist expresses with his
response that the comprehension problem has been resolved (lines 29–40).
This example illustrates the difficulty faced by persons with ASD in taking the
perspective of others: in other words, the speaker finds it difficult to infer the number
of contextualisation cues that are required for the interlocutor to understand the
message (Gumperz, 1992). In line with the instructions of easy language (Kartio,
2009: 15–16), asking the speaker directly leads to the resolution of the comprehen-
sion problem. Formulating a candidate understanding would probably not have been
possible in this case, because the therapist receives Rikhard’s first argument (line 21:
‘because of the clothes’, vaatteista) as new information with the discourse particle
‘aha’ (aha, line 23) carrying a rising pitch (Hakulinen et al., 2004: 774).
The example also supports the hypothesis according to which high-functioning
individuals with ASD—such as the participants in these data—are able to manage the
processing work required by conversational repairs themselves. For this reason, direct
repair initiations seem to work better than candidate understandings in interacting
with high-functioning persons with ASD. Indeed, as Kartio (2009: 16) warns, the
problem with candidate understandings is that they may lead to so-called “yea-
saying”—that is, to accepting the suggestion presented by the interlocutor even if it
does not correspond to the meaning originally intended by the speaker.

4.2.3 Non-verbal Features Associated with Trouble-Source

Turns: The Role of Eye Contact4

People with ASD often exhibit atypical non-verbal behaviours, such as eye contact,
facial expressions, postures, and gestures (e.g., McPartland & Klin, 2006; Szatmari
et al., 1989; Tantam et al., 1993; Wiklund, 2012). Indeed, the lack of eye contact is
typically one of the first symptoms to draw adult’s attention in a child when ASD
is suspected (Hutt & Ounsted, 1966; Kylliäinen, 2007). Typically, human infants
develop the capacity to follow a gaze as early as three months of age (D’Entremont

4 This sub-chapter presents an overview of a study described in greater detail in Wiklund (2016). In
addition to the role of gaze, Wiklund (2016) considers prosodic features of repairs and repair types
in the interactions of preadolescents with ASD.
4.2 Causes of Comprehension Problems in Group Therapy Sessions … 87

et al., 1997). This makes it possible to establish joint attention with another individual,
and therefore, it helps in understanding other people’s thoughts and intentions (Baron-
Cohen, 1995; Frith & Frith, 2001). People with ASD appear to have a different way
of engaging in gaze following and an impairment of joint attention (e.g., Charman,
2004; Dawson et al., 2004; Loveland & Landry, 1986; Mundy, 1995; Mundy &
Newell, 2007).
Wiklund (2016) studied the relationship between non-verbal features and conver-
sational repairs. More precisely, the study focused on non-verbal features that
occurred during the production of trouble-source turns that launched other-initiated
repairs (repair initiations and outright corrections) in group therapy sessions. The data
consisted of the Finnish ASD groups’ sessions (see Sect. 1.2). The most important
finding of the study (Wiklund, 2016) was that in 84.1% of instances, there was no eye
contact between the speaker producing a trouble-source turn and the one initiating the
repair sequence (or making an outright correction). It is noteworthy that the lack of
eye contact was sometimes associated with overlapping speech (38.1%). Most often,
the overlaps occurred between the boys with ASD and their therapists (62.5%). In the
remaining cases, the overlaps occurred between the therapists (20.8%) or between the
boys (16.7%). Sometimes there were other non-verbal features related to the trouble-
source turn, but these cases were far less frequent than the cases including a lack
of eye contact. For example, in a few instances, the speaker producing the trouble-
source turn was smiling (4.8%), pointing at something with his finger (4.8%), or
making gestures with his hands (4.8%).
Example 4.35 presents a situation of interaction including two trouble-source turns
followed by other-initiated repairs. The extract is from the first part of the session
of Group B. Toni has just shared his latest news with the others. He has mentioned
that school has been boring, and the female therapist reacts to this by saying that
when she and the other therapist were kids, they thought that going to school was
fun. The speaker’s gaze is indicated at the top of each line in a notation system
formulated by Goodwin (1981). A full stop (.) means that the speaker turns to look
at an interlocutor. A line (__) is used to indicate that the speaker is looking at an
interlocutor, and a comma (,) means that the speaker is turning his/her gaze away. In
addition to these generally used conventions, a fourth sign has been added: a line of
x’s (xxxx) signifies that the speaker is looking away from the interlocutors.

5 The same extract is analysed from another point of view in Example 4.3 of this book. In addition,
the extract can be found in Lehtinen (2012), Wiklund (2016), and Wiklund and Laakso (2019).
88 4 Comprehension Problems

Example 4.3:

____________________________________________________

01 FT: ((looks at MT)) ja sithän me oltiin i- (.) kun olimme koululaisia ni

and then-CLI 35 we were when we were schoolchildren

_______________________________

02 meidän mielestä koulu oli kivaa,

we thought that going to school was fun

_________________________________________________

03 MT: ((looks at FT)) niin (.) ainakin aina silloin tällöin. hehee[heh]

yes at least every now and then (laughs)

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

04 Kalle: [no ] (.) silloin oli varmaan kou#luruokakin parem°paa°#;

well at that time food was probably also better at school

xxxx

05 Toni: nii;

yeah

_____

06 Jaakko: ((looks at Toni)) njää,

n-yeah

xxxxxxxxxxxxxxx

07 FT: ((bends towards Toni)) =mikä mikä oli,?

what what was

xxxxxxxxxxxx

08 Toni: <kouluruoka>

food at school

xxx . __

09 FT: (smack) [njaa-a] ((turns to look at Jaakko))

um I see
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

10 Jaakko: [#ää no] silloin# silloin perusruoka #sil

well er at that time at that time basic food at that

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
4.2 Causes of Comprehension Problems in Group Therapy Sessions … 89

11 tasolla oli alhasempi ku nykyään joten# (0.4) krh (0.4)

level was lower than nowadays so

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

12 ↑kouluruoka ei tartte #°olla ni°# .hh (0.4) niin hy- (0.3) °hy-°

the food at school doesn’t need to be that that go- go-

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

13 oliv- (.) #ei tarttenu olla ↑niin hyvää >koska se on samaa

wa- didn’t need to be so good because it is on the same

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

14 luokkaa kuin ↑ne< (2.0) >jos joku just tajus mitä mä sanoin<.

level as them if someone just understood what I said

15 (2.4)

xxxxxxxxxxxxxxxxxx .

16 MT: ((looks upwards)) samaa luokkaa kuin; ((turns to look at Jaakko))

on the same level as

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

17 Jaakko: .mth #no kun al- (.) (>kun oltii a↑jas<) taakseppäin#

well when lo- back in the old days

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

18 ruoka oli vähän <↑alhaisempaa>. (0.9)

food was a bit lower

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

19 #silleen (0.5) mh silleen kuing (.) °hy- pa-° (0.3) <makusta> ?

I mean like like how go- ba- taste

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

20 (0.3) ja >↑kouluruoka on< parempaa joten ne on samal ↑linjalla

and the food at school is better so they are on the same level

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

21 (0.5) johtuen ↑ajastansa. (1.4)

due to their time

______
90 4 Comprehension Problems

22 FT: ((looks at Jaakko and nods)) hmm-m?

uh um

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

23 Jaakko: <eli sil periaatteella (0.8) silloin ↑oli °parempaa°.>

so on that principle at that time it was better

24 (2.0)

xx
25 MT: mm

um

______ , xxxx

26 FT: ((looks at Jaakko))(smack) jaaha

um I see

According to Wiklund (2012), it is more difficult for preadolescents with ASD

to maintain eye contact when they are speaking than when they are listening to
others. The most important finding of Wiklund (2016) is probably that in 84.1% of
the trouble-source turns in the data there was no eye contact between the speaker
producing the turn and the one initiating the repair sequence (or making an outright
correction). Example 4.3 presents two occurrences of trouble-source turns where
there is no eye contact between the participants. The turn produced by Kalle (line
04) is characteristic: during the therapist’s previous turn (lines 01–02), Kalle looked
at her, but when he begins to speak himself (line 04), he turns his gaze to his own
hands and keeps it down during the duration of the whole turn (Wiklund, 2016).
Indeed, according to Wiklund (2012), looking at one’s own hands is one of the main
strategies for avoiding eye contact when speaking (see the next chapter in this book).
The second trouble-source turn in Example 4.3 is also produced without eye
contact between the speaker and the other participants (lines 10–14). The speaker,
Jaakko, looks straight ahead, which is another main strategy for avoiding eye contact
in these data (Wiklund, 2012). Jaakko also gestures with his hands. The gestures,
however, seem haphazard, and they do not help in interpreting his speech. The boy
probably gestures with his hands in order to help his concentration and/or to indicate
uncertainty (Wiklund, 2016).

4.3 Conclusion

The studies on comprehension problems carried out within the framework of this
project (Lehtinen, 2012; Wiklund, 2016; Wiklund & Laakso, 2019, 2020) confirm
the results of previous studies (e.g., Geller, 1998; Keen, 2003), according to which
children with ASD often experience communication breakdowns in their interactions.
4.3 Conclusion 91

Unlike previous studies, however, in these data, the comprehension problems did not
lead to problem behaviour (Brady et al., 1995; Keen, 2003; Wetherby et al., 1998).
The results are in line with those of Geller (1998) in many respects. Indeed,
according to Geller (1998), most of the communication breakdowns in the interac-
tions of school-aged children with ASD are caused by a lack of referential clarity,
which is also typical of the data studied within the framework of this project
(Lehtinen, 2012; Wiklund, 2016). In addition, pragmatic errors, such as unclear
coding of intentionality and linguistically unmarked topical shifts, also cause a certain
number of the children’s communication breakdowns in Geller’s (1998) data. This
is interesting, because ‘topical discontinuities’ constitute one of the most common
causes of comprehension problems in these data as well (Lehtinen, 2012; Wiklund,
2016). One difference can be found: in Geller’s (1998) study, errors in syntax or
morphology did not lead to the listeners’ comprehension problems, whereas in the
Finnish data of the project described in this book, ‘ungrammatical utterances’ and
‘disfluencies’ did cause a certain number of comprehension problems between the
participants (Wiklund & Laakso, 2019, 2020). However, the most common cause of
comprehension problems in these data was ‘overly literal understanding of speech’.
According to Volden (2004), children with ASD use a variety of repair strategies,
and they are able to add information if the communication breakdown persists. The
findings presented in Lehtinen (2012) and Wiklund (2016) are in line with Volden’s
(2004) results: the boys with ASD are often able to add information if the communi-
cation breakdown persists and to resolve the comprehension problem with their own
actions. It is noteworthy that in these data, the boys with ASD are also able to act
as a team when a comprehension problem occurs, and this teamwork is most often
successful.
The results of Lehtinen (2012) and Wiklund (2016) are also in line with Philip’s
(2008) findings. Philip (2008) studied the conversational repair responses of adoles-
cents with ASD to communication breakdowns that occurred naturally in conversa-
tions. The author found that although the informants also differed from each other
concerning the repairs, all three participants had communication breakdowns caused
by the presence of ambiguous referents, inadequate information, and irrelevant infor-
mation (p. 114). Ambiguous referents are also typical of the Finnish data of this
project, and the data also include instances where the boys with ASD provide inade-
quate or irrelevant information in their turns, and this leads to comprehension prob-
lems between the participants. Abbas (2021), in turn, studied interactions between
neurotypical persons and children with ASD in institutional and home settings. His
results show that ‘repetition’ was the most frequent strategy used by the therapists
to repair a communication breakdown, and the second most frequent strategy was
‘modification’, which consisted of repeating the same question with alternations
(p. 141). These same strategies are used also by the therapists of the Finnish data
of this project (Lehtinen, 2012; Wiklund, 2016). In the data of Abbas (2021), these
strategies were successful in some cases, but communication breakdowns also some-
times remained unrepaired. In the data of this project (Lehtinen, 2012; Wiklund, 2016;
Wiklund & Laakso, 2019, 2020), all communication breakdowns were finally at least
superficially repaired, but sometimes it was unclear if the therapists had understood
92 4 Comprehension Problems

what the boys said or if they had just produced a ‘display of understanding’ instead
of a ‘proof of understanding’ (Sacks, 1992).

References

Abbas, I. (2021): Unraveling conversations in neurotypical-neurodivergent dyads: a study of repair

in interactional loops. [Doctoral dissertation]. Muzaffarabad: University of Azad Jammu and
Kashmir.
Baron-Cohen, S. (1995). Mindblindness: An Essay on autism and theory of mind. MIT Press.
Baron-Cohen, S. (2001). Theory of mind and autism: A review. International Review of Research
in Mental Retardation, 23, 169–184.
Brady, N. C., McLean, J. E., McLean, L. K., & Johnston, S. (1995). Initiation and repair of intentional
communication acts by adults with severe to profound cognitive disabilities. Journal of Speech
and Hearing Research, 38, 1334–1348.
Charman, T. (2004). Why is joint attention a pivotal skill in autism? Philosophical Transactions of
the Royal Society of London, 358, 315–324.
D’Entremont, B., Hains, S. M. J., & Muir, D. W. (1997). A demonstration of gaze following in 3–6
month-olds. Infant Behavior and Development, 20(4), 569–572.
Dawson, G., Toth, K., Abbott, R., Osterling, J., Munson, J., Estes, A., & Liaw, J. (2004). Early
social attention impairments in autism: Social orienting, joint attention and attention to distress.
Developmental Psychology, 40(2), 271–283.
Drew, P. (1997). ‘Open’ class repair initiators in response to sequential sources of troubles in
conversation. Journal of Pragmatics, 28(1997), 69–101.
Frith, U., & Frith, C. (2001). The biological basis of social interaction. Current Directions in
Psychological Science, 10(5), 151–155.
Geller, E. (1998). An investigation of communication breakdowns and repairs in verbal autistic
children. The British Journal of Development Disabilities, 44(87), 71–85.
Goodwin, C. (1981). Conversational organization: Interaction between speakers and hearers.
Academic Press.
Gumperz, J. J. (1992): Contextualization revisited. In P. Auer & A. Di Luzio (Eds.), Contextualiza-
tion of language (pp. 39–53). John Benjamins Publishing Company.
Hakulinen, A., Vilkuna, M., Korhonen, R., Koivisto, V., Heinonen, T. R., & Alho, I. (2004). Iso
suomen kielioppi. SKS.
Hale, C. M., & Tager-Flusberg, H. (2005a). Brief report: The relationship between discourse deficits
and autism symptomatology. Journal of Autism and Developmental Disorders, 35, 519–524.
Hale, C. M., & Tager-Flusberg, H. (2005b). Social communication in children with autism: The
relationship between theory of mind and discourse development. Autism, 9(2), 157–178.
Hutt, C., & Ounsted, C. (1966). The biological significance of gaze aversion with particular reference
to the syndrome of infantile autism. Behavioral Science, 11(5), 346–356.
Kartio, J. (Ed.). (2009). Selkokieli ja vuorovaikutus. Kehitysvammaliitto.
Keen, D. (2003). Communicative repair strategies and problem behaviours of children with autism.
International Journal of Disability, Development and Education, 50(1), 53–64.
Keen, D. (2005). The use of non-verbal repair strategies by children with autism. Research in
Developmental Disabilities, 26, 243–254.
Kitzinger, C. (2014). Repair. In J. Sidnell & T. Stivers (Eds.), The handbook of conversation analysis
(pp. 229–256). Wiley Blackwell.
Kylliäinen, A. (2007). Face and gaze processing in children with autism. [Doctoral dissertation].
University of Tampere.
References 93

Lehtinen, M. (2012). “Aiheesta toiseen” – Ymmärrysongelmatilanteita Asperger-nuorten kuntou-

tuskeskusteluissa. In L. Leskelä & C. Lindholm (Eds.), Haavoittuva keskustelu. Keskustelu-
nanalyyttisia tutkimuksia kielellisesti epäsymmetrisestä vuorovaikutuksesta (pp. 235–277).
Kehitysvammaliitto.
Loveland, K. A., & Landry, S. H. (1986). Joint attention and language in autism and developmental
language delay. Journal of Autism and Developmental Disorders, 16(3), 335–349.
McPartland, J., & Klin, A. (2006). Asperger’s syndrome. Adolescent Medicine Clinics, 17(3),
771–788.
Mundy, P. (1995). Joint attention and social-emotional approach behavior in children with autism.
Development and Psychopathology, 7, 63–82.
Mundy, P., & Newell, L. (2007). Attention, joint attention and social cognition. Current Directions
in Psychological Science, 16, 269–274.
Philip, B. A. (2008). Conversational repair strategies in adolescents with autism spectrum disorders.
[Doctoral dissertation]. Bowling Green State University.
Sacks, H. (1992). Lectures on conversation (Vol. 1 & 2). Blackwell.
Schegloff, E., Jefferson, G., & Sacks, H. (1977). The preference for self-correction in the
organization of repair in conversation. Language, 53(2), 361–382.
SIL Glossary of Linguistic Terms. (2021). https://glossary.sil.org/.
Sorjonen, M.-L. (1997). Korjausjäsennys. In L. Tainio (Ed.), Keskustelunanalyysin perusteet
(pp. 111–137). Vastapaino.
Szatmari, P., Brenner, R., & Nagy, J. (1989). Asperger syndrome: A review of clinical features. The
Canadian Journal of Psychiatry, 34, 554–560.
Tantam, D., Holmes, D., & Cordess, C. (1993). Non-verbal expression in autism of Asperger type.
Journal of Autism and Developmental Disorders, 23(1), 111–133.
Volden, J. (2004). Conversational repair in speakers with autism spectrum disorder. International
Journal of Language & Communication Disorders, 39(2), 171–189.
Wetherby, A., Alexander, D., & Prizant, B. (1998). The ontogeny and role of repair strategies. In
A. Wetherby, S. Warren & J. Reichle (Eds.), Transitions in prelinguistic communication (Vol. 7,
pp. 135–159). Paul H Brookes.
Wiklund, M. (2012). Gaze behavior of pre-adolescent children afflicted with Asperger Syndrome.
Communication & Medicine, 9(2), 173–186.
Wiklund, M. (2016). Interactional challenges in conversations with autistic preadolescents: The
role of prosody and nonverbal communication in other-initiated repairs. Journal of Pragmatics,
94, 76–97.
Wiklund, M., & Laakso, M. (2019). Ungrammatical utterances and disfluent speech as causes
of comprehension problems in interactions of preadolescents with highly functioning autism.
Clinical Linguistics & Phonetics, 33(7), 654–676.
Wiklund, M., & Laakso, M. (2020). Comparison of disfluent and ungrammatical speech of pread-
olescents with and without ASD. Journal of Autism and Developmental Disorders, 51(8),
2773–2789.
Chapter 5
Gaze Behavior

Abstract People with ASD often exhibit atypical non-verbal behaviours, which are
often manifested by avoiding eye contact. Gaze constitutes an important interac-
tional resource, and a person’s tendency to avoid eye contact may affect the fluidity
of conversations and cause misunderstandings. It is therefore important to know
precisely the ways how this avoidance is done, and how it affects interaction. The
objective of this chapter (based on Wiklund, 2012) is to describe the gaze behaviour of
preadolescent children with ASD in institutional multiparty conversations. Method-
ologically, the study is based on conversation analysis and multimodal study of
interaction. The findings show that three main patterns are used for avoiding eye
contact: (1) fixing one’s gaze straight ahead, (2) letting one’s gaze wander around,
and (3) looking at one’s own hands when speaking. The informants of this study
do not look at their interlocutors at all at the beginning or the middle of their turn.
However, they sometimes turn to look at the interlocutors at the end of their turn.
This proves that these children are able to use gaze as a source of feedback. The infor-
mants of this study also show other remarkable interactional skills. For example, they
can launch narratives and interpret the implicit conversational meanings of combina-
tions of dialogue particles, prosodic features, and gestures correctly. Looking at the
speaker when listening also seems to be easier for them than looking at the listeners
when speaking themselves.

Keywords Gaze · Eye contact · Autism spectrum disorder · ASD · Non-verbal

communication

5.1 Introduction1

It is well established that typically developing human infants are able to follow the
gaze as early as the age of three months (D’Entremont et al., 1997). This makes
it possible to establish joint attention with another individual, which is extremely
important from the developmental point of view (Rossano, 2013: 310). Indeed,

1 This chapter is based on Wiklund (2012).

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 95
M. Wiklund, Speech and Interaction of Preadolescents with Autism Spectrum Disorder,
Prosody, Phonology and Phonetics, https://doi.org/10.1007/978-981-19-8117-3_5
96 5 Gaze Behavior

several studies have shown that establishing joint attention helps to understand other
people’s thoughts and intentions (e.g., Baron-Cohen, 1995; Frith & Frith, 2001).
Brooks and Meltzoff (2005) have also shown that establishing joint attention corre-
lates strongly with language acquisition. Therefore, as Rossano (2013: 310) points
out, “being able to detect the direction of another’s eyes is, from an ontogenetic point
of view, extremely important in the process of becoming social beings and acquiring
language”. Persons with ASD, however, often have deficiencies in following other
individuals’ gazes and in establishing joint attention (e.g., Charman, 2004; Dawson
et al., 2004; Loveland & Landry, 1986; Mundy, 1995; Mundy & Newell, 2007).
The crucial role played by eye contact in fluent interaction has long been acknowl-
edged (Argyle & Cook, 1976; Kendon, 1967). Gaze and other indications of atten-
tion are normal responses to speech and other communicative initiatives of the co-
participants (Hedge et al., 1978; Tantam et al., 1993). Gaze acts as a turn-taking signal
in interaction (Bavelas, 2005; Jokinen et al., 2009). It is also known that the gaze
constitutes an important source of feedback for the interlocutors (Goodwin, 2000)
and that it plays an important role in establishing and sustaining the focus of shared
attention (Trevarthen, 1984). One of the key features of normal gaze behaviour is that
listeners look at speakers (Goodwin, 1979, 1981; Tantam et al., 1993). According
to Tantam et al. (1993), persons with ASD are deficient in this ‘gaze response’. For
these authors, the deficiency arises from “the inborn tendency to orient to human
faces and human vocalisation is impaired in autism” (Tantam et al., 1993: 131).
Indeed, studies have shown that the difficulties in non-verbal communication faced
by people with ASD are often manifested particularly in avoiding mutual gaze (Szat-
mari et al., 1989; Tantam et al., 1993). The lack of eye contact is also typically one of
the first symptoms that draw adults’ attention to autistic children (Hutt & Ounsted,
1966; Kylliäinen, 2007).
When persons with ASD establish eye contact, it is often either evasive or intense
and staring (Gillberg & Gillberg, 1989; Nieminen-von Wendt, 2004). Some studies
suggest that the impression of abnormality in the interaction of persons with ASD
results mainly from a lack of integration of expression, speech, and gaze (Tantam,
1991; Tantam et al., 1993). Because gaze is such an important interactional resource,
the tendency of persons with ASD to avoid eye contact may affect the fluidity of
conversations and cause misunderstandings in situations where persons with ASD
are interacting with neurotypical people. For this reason, it is important to know
precisely how this gaze avoidance is performed, and how it affects the interaction.
The objective of the study (Wiklund, 2012) presented in this chapter consists of
describing the gaze behaviour of Finnish-speaking, 11–13 year-old boys with ASD in
institutional multiparty conversations (for a description of the data, see Sect. 1.2.1).
Methodologically, the study is based on Conversation Analysis (e.g., Hutchby &
Wooffit 2008; Sacks et al., 1974; Heritage & Maynard, 2006) and multimodal study
of interaction (Goodwin, 1979, 1981, 2000).
5.2 Three Main Patterns Used for Avoiding Eye Contact in Group Therapy … 97

5.2 Three Main Patterns Used for Avoiding Eye Contact

in Group Therapy Sessions Involving Preadolescent
Boys with ASD

It is sometimes said that it is easier for people with ASD to concentrate on producing
speech without having eye contact with the interlocutor and that this is why some of
them even talk with their eyes closed. On the other hand, neurotypical speakers do not
look at their interlocutors all the time either: according to eye-tracking experiments
by Jokinen et al., (2009: 307), the gaze is directed towards the interlocutors only
about 60% of the time while speaking. However, as the examples will show, the
gaze behaviour of the boys in the present data corpus is atypical for two reasons: (1)
the duration of the time they look at the interlocutors when speaking is remarkably
brief, and (2) they avoid eye contact at certain sequential positions where establishing
mutual gaze would be expected. Three main patterns for avoiding eye contact were
found and will be presented in the sub-sections below: fixing one’s gaze straight
ahead (Sect. 5.2.1), letting one’s gaze wander around (Sect. 5.2.2), and looking
at one’s own hands (Sect. 5.2.3). The analyses also aim to bring the participants’
interactional strengths to light (Wiklund, 2012).

5.2.1 Fixing One’s Gaze Straight Ahead

The most common pattern consists of fixing one’s gaze straight ahead when speaking.
This pattern is used by all the participants. It consists of choosing an empty spot where
no one is sitting and fixing one’s gaze at this spot. The first example illustrates this
phenomenon. The gaze of the speaker is indicated on top of each line using the
notation system presented by Goodwin (1981). A full stop (.) means that the speaker
turns to look at an interlocutor. A line (__) is used to indicate that the speaker is
looking at an interlocutor, and a comma (,) means that the speaker turns his gaze
away. Lines where the main speaker of each example has established eye contact
with someone are emphasised by an arrow at the beginning of the line. In addition
to these generally used conventions, a fourth sign has been added: a line of x-letters
(xxxx) signifies that the speaker looks away from the interlocutors. The same notation
system is used for the recipients’ gaze, except that the signs are placed under the
transcription of the speech.
The following extract has been taken from the beginning of the session of Group
A, when the first participant, Heikki, starts to tell the others his latest news. The
male therapist (MT) has just asked the boys who wants to start. Both therapists and
one of the boys (Rikhard) turn spontaneously to look at Heikki, because he has the
reputation of being very talkative and always willing to start. Heikki, who gazes
straight ahead, hesitates a moment (line 01) and then volunteers to start (line 06).
98 5 Gaze Behavior

Example 5.1:

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

01 Heikki: ö[ö: ] ((gaze fixed straight ahead))

er

02 MT: ((looks at Heikki))

________________

03 FT: ((looks at Heikki))

________________

04 Rikhard: ((looks at Heikki))

________________

05 Mikael: [m-,] hh ((looks at his hands))

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

06 Heikki: ↑minä voin alot°taa°. ((looks straight ahead))

I can start

07 MT: no ↑kerro vähän kuulumi°sia°.

well tell us some news

___________________________

08 FT: °↑mm°

_____

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

09 Heikki: jaa: ((looks straight ahead)) ↓niin t'ta:? (.) hhhh (0.6)

well so PRT

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

10 ↑hy↑vin ↓om mennyk koulussa ja::, (.) >ollu< ((glances away

it has gone well at school and has been

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

11 from the others; then looks again straight ahead)) tota: (.)

PRT
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
12 on yks kaveri jota mä en oot tavannup pitkää aikaan nii: (.) se

there is a friend that I haven’t seen for a long time so he

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
5.2 Three Main Patterns Used for Avoiding Eye Contact in Group Therapy … 99

13 ↑vihdoinkip pääs tänä viikolloppuna tapaa tuli meil↓le: (0.4) hh

could finally come this weekend to see me he came to our place

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

14 ja: (.) pelattii (.) >p'la-< #p'lattii# hyvii pele:j ja oli iha:

and we played pla- played good games and we had quite

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

15 (.) kivaa keskenää? ((looks straight ahead))

fun together

16 (0.8)

. ____________________________________, xxxxxxxxxx

17 → Heikki: turns to look at the MT and nods twice; then turns

xxxxxxxxxxxxxxxxxxxxxxxx

18 his gaze straight ahead))

19 MT: mm m? ((looks at Heikki and nods several times))

________________________________________________

20 FT: ↓kiva:? ((looks at Heikki))

that’s nice

___________________________

21 Rikhard: ((looks at Heikki))

__________________

22 Mikael: ((gaze wanders around))

xxxxxxxxxxxxxxxxxxxxxxx

23 (.)

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

24 Heikki: >tai< se:, (.) >se on sellanen ↑kaveri jota mä en oo< tavannum

or it it’s a friend that I haven’t seen

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

25 melkei yhteen ↑kuu↓kautee koska sillä o ollu niim paljom menoja.

for almost one month because he has been so busy

xxxxxxxxxxxxxxxxxxxxxxxx
100 5 Gaze Behavior

26 ((looks straight ahead))

27 (0.6)

. , xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

28 → Heikki: ((glances very briefly at MT, then turns his gaze straight

xxxxxx

ahead))

29 MT: mm m? .nsss ((looks at Heikki and nods))

_________________________________

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

30 Heikki: ja ↑sit tota: öö, mhhh (0.4) ((looks straight ahead))

and then PRT er

The male therapist gives Heikki permission to start (line 07), and Heikki takes
the floor (line 09). Heikki accepts the turn (line 06) without looking at MT, who is
responsible for allocating turns. This is made possible by the explicit formulation
of the turn-allocation (Kääntä, 2010). When Heikki takes the turn (line 09), he still
does not turn to look at the others. This, however, is not exceptional because speakers
typically look away at the moment of turn-taking (Jokinen et al., 2009: 307). On the
other hand, Heikki does not turn to look at the interlocutors at all during the entire
turn. According to Jokinen et al., (2009: 307), this is a peculiar feature because
speakers typically look at the recipients about 60% of the time.
Heikki’s gaze is intense, and it is fixed straight ahead on a big mirror placed in
front of him on the other side of the room. At the end of the first utterance (line
10), the speaker glances at a wall to his left, opposite to his interlocutors. Then
he turns his head back in the same direction as before and keeps looking straight
ahead until the end of his turn (line 15). However, he turns to look at MT after the
end of his turn during the pause that follows (line 16) and nods twice (line 17).
As MT is already looking at Heikki, they establish a mutual gaze easily. The male
therapist acknowledges Heikki’s nodding gesture by nodding back several times and
by producing a verbal minimal response (line 19).
He uses the dialogue particle mm, produced with a rising pitch. This indicates
that the recipient is listening and expects the current speaker to continue (Hakulinen
et al., 2004: § 798, § 800). The female therapist reacts to Heikki’s turn with a brief
evaluative comment (kiva, ‘that’s nice’, line 20). Heikki interprets these reactions as
signs of approval, turns back towards the mirror (straight ahead), and continues his
narrative (line 24). The same happens after his second turn (line 28): Heikki glances
at MT during the pause after his turn in order to elicit feedback, and MT reacts to
this with a nodding gesture associated with a verbal minimal response (mm, line 29).
5.2 Three Main Patterns Used for Avoiding Eye Contact in Group Therapy … 101

Heikki interprets this reaction as a sign of approval and as an implicit request to

continue speaking, because he turns his gaze away and takes the turn again (line 30).
This example not only confirms the results of Solomon (2004), according to which
children with ASD are able to launch narratives, but it is also proof of their non-verbal
competence in using mutual gaze as a source of feedback at relevant points of a
narrative sequence. It is noteworthy that although it has long been demonstrated that
people with ASD have difficulties in recognising non-verbal signs from their inter-
locutors (e.g., Scott, 1985), this example gives some opposite evidence: Heikki turns
to look at MT in order to elicit feedback and correctly interprets MT’s non-verbal
signs that encourage him to continue speaking. He is also able to interpret correctly
the interactional meaning of the dialogue particle mm, produced with a rising pitch.
Thus, even if it is well known that people with ASD have difficulties understanding
implicit messages in interaction (Cummings, 2009; Lehtinen, 2012; Lewis et al.,
2008; Martin & McDonald, 2004; Wiklund & Stevanovic, 2018; Wiklund, 2016),
this example proves that even when several interactional resources are being used
simultaneously (e.g., gaze, gestures, dialogue particles, and prosody), a preadoles-
cent child with ASD can interpret implicit interactional meanings. One of the boys,
Rikhard, looks at Heikki all the time, which allows the speaker to get visual feedback
from him (Goodwin, 1981: 57–59; Kallio & Ruusuvuori, 2011: 81). Thus, his gaze
behaviour as a listener is very competent. The other boy, Mikael, in turn, does not
look at the speaker at all. At the beginning (line 05), he looks at his hands. Later (line
22), his gaze starts to wander around. Consequently, it would be impossible for the
speaker to receive non-verbal feedback from him. The absence of gaze could also be
interpreted as a lack of interest in what is being said (Kallio & Ruusuvuori, 2011:
67; Trevarthen, 1984). Mikael’s gaze behaviour does not, however, affect the course
of the conversation, because Heikki does not try to elicit any visual feedback from
him, only from MT.

5.2.2 Letting One’s Gaze Wander Around

The next example illustrates another frequently occurring pattern: the face of the
speaker is turned either straight forward or slightly to one side (in both cases in a
direction where no one is sitting), and his gaze wanders around (Wiklund, 2012).
The extract comes from the session of Group B. Jaakko and Kalle have already told
their news, and now it is Toni’s turn.
102 5 Gaze Behavior

Example 5.2:

01 FT: ((turns to look at Toni))

. ______________________

02 MT: ((turns his gaze straight ahead))

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

03 Jaakko: ((turns to look at Toni))

. _______________________

04 Kalle: (((turns to look at Toni))

. ________________________

05 Teppo: ((turns his gaze straight ahead))

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

06 (6.2)

xxxxxxxxxxx . ____________________________

07 → Toni: ai kerronks mä ? ((turns to look at the FT)

er do I tell now

08 FT: ↑no joo. ((looks at Toni and nods))

well yes
___________________________________

, xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

09 Toni: ((turns his gaze straight ahead))

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

10 Toni: ((turns to look at his hands)) no mul nyt ei oo tapahtunu

well I have nothing

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

11 mitään muuta ((raises his gaze; looks slightly upward; gaze

special to tell now

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
12 wanders around)) #erikoista# (0.7) >paitsi et eilen ((looks at his

except that yesterday

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

13 hands)) oli< #ruotsinkokeet# (1.0) ((raises his gaze; gaze wanders

we had a Swedish test

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
5.2 Three Main Patterns Used for Avoiding Eye Contact in Group Therapy … 103

14 around)) °ne meni (.) iha° (1.1) ((looks at his hands)) ↑hyvin

it went quite well

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

15 >°#ainaki mun (.) mielestä ((raises his gaze; gaze wanders

at least that’s how I feel

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

16 around)) eikä nyt#°< ↑MUUTEN ei °#oo tapahtunu mitään# .hh

and otherwise nothing special has happened

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

17 ↑erikoista° ((turns his face slightly down; gaze wanders around))

18 (3.7)

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

19 Toni: ((face is turned straight forward; gaze wanders around))

20 Jaakko: ((looks straight ahead))

xxxxxxxxxxxxxxxxxxxxxxxxx

21 Kalle: ((looks at his shirt and fiddles with it with his both hands))

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

22 Teppo: ((looks at Toni diagonally))

____________________________

23 FT: ((looks at Toni))

________________

24 MT: ((looks at Toni))

_________________

25 MT: ((looks up)) >voitsä kertoo vähä mitä< ((turns to look at Toni))

can you tell a bit about

, xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx . _______________________

26 ↑TAvallista on tapahtunu.

the ordinary things that have happened

_________________________

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
104 5 Gaze Behavior

27 Toni: ↑TAvallista no menny kou↑luun ((turns his face slightly on the

ordinary things well I’ve gone to school

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

28 left, toward the MT; gaze wanders around)) (.) sit siel↑lä (.)

then there

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

29 opiskel↑lu (0.9) °#on ollu#° (.) tyl↑sää (0.3) °#koulussa#° (1.1)

I’ve studied it has been boring at school

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

30 sit lähteny ko#tiin# (2.8) #sieltä ja sit syöny (0.8) ja sit ollu

then I have left home from there and then I’ve eaten and then I’ve spent

.___________________________________________________

31 → vähän aikaa °koneella°#,? ((turns to look at the MT))

some time on my computer

32 MT: ((first looks at Toni; then turns his gaze away))

_____________________ , xxxxxxxxxxxxxxxxxxxxxxxxx

xxxxxxxxxxxxxxxxxxxxxxxx
33 FT: ((looks straight ahead))

xxxxxxxxxxxxxxxxxxxxxxxx
34 Jaakko: ((face is turned straight forward; rubs his eye with his

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

finger))

xxxxxxxx

35 Kalle: ((looks down on the floor on his left))

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

36 Teppo: ((looks at Toni diagonally))

____________________________
37 (3.5)

38 FT: ((turns to look at Toni)) ja sitten,

and then
. __________________________________

, xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
5.2 Three Main Patterns Used for Avoiding Eye Contact in Group Therapy … 105

39 Toni: ja sitten ((turns his face a bit more to the right; gaze wanders

and then

xxxxxxx
around))

40 Kalle: ((turns to look at Toni))

. _______________________
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

41 Toni: #menny ↑kotiin ja# (1.4) ((turns his gaze slightly down; gaze

I’ve gone home and

xxxxxxxxxxxxxxxx

wanders around))

42 Kalle: ((turns his gaze straight forward, looks down and raises his gaze

, xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx , xxxxxxxxxxxx

43 up again))

xxxxxxxxxx

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
44 Toni: #siellä# (1.2) °nukkunu°, ((during the word ‘nukkunu’ turns his

there I’ve slept

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

45 gaze straight ahead and fixes his gaze on the table))

46 (0.3)

47 FT: ((looks at Toni)) ↑mmm

PRT

______________________

Toni looks at the female therapist when he asks her if it is his turn to take the
floor (line 07). Kalle, who is sitting next to Toni, has just finished telling his news,
and now it is Toni’s turn. There is a long (6.2 s) pause (line 06), during which the
female therapist as well as Jaakko and Kalle turn to look at Toni significantly. It
takes some time before Toni understands the interactional meaning of the gaze of
the other participants, i.e., the fact that it constitutes an embodied turn-allocation in
this context. The interpretation of the embodied turn-allocation is complicated by
the lack of eye contact between Toni and FT, who is responsible for allocating turns
along with MT (Kääntä, 2010). However, when Toni perceives after a silence that FT
is looking at him, he looks at her and asks for permission to take the floor (line 07).
This proves that he has interpreted the embodied turn-allocation correctly. It is also
noteworthy that he spontaneously elicits visual feedback from FT before taking the
floor (Jokinen et al., 2009) and uses gaze in the construction of a question–answer
106 5 Gaze Behavior

adjacency pair (Rossano et al., 2009). On the other hand, this example also shows
the other boys’ ability to use gaze as a sign of turn-allocation: Jaakko and Kalle
(lines 03–04) turn to look at Toni during the pause in order to indicate that they have
already told their news and now it is Toni’s turn, whereas Teppo, whose turn will
come only after Toni’s (line 05), turns his gaze straight ahead, away from the others,
as a signal of not accepting the turn (Jokinen et al., 2009: 307).
When Toni takes the turn (line 10–17), he turns his gaze away, which is normal
at the moment of turn-taking (Jokinen et al., 2009: 307). However, what is atypical
is that, while speaking, Toni hardly looks at the interlocutors at all. During his first
turn (lines 10–17), Toni keeps turning his face up and down and lets his gaze wander
around. He does not turn to look at his interlocutors even at the end of his turn or
during the long pause that occurs after it (line 18); during this pause, his face is
turned straight ahead and his gaze wanders around (line 19). Thus, unlike common
practice (Jokinen et al., 2009; Kendon, 1967), the speaker does not look at the other
participants at the end of his turn. This probably explains the delayed response from
the interlocutors: MT reacts to Toni’s turn only after a 3.7 s pause (lines 25–26)
(Kendon, 1967). As speakers normally turn to look at the interlocutors at the end of
their turns in order to elicit feedback and possible turn-taking (Jokinen et al., 2009:
307), the fact that Toni does not do so falsely implies that he wants to continue.
However, the only boy with whom Toni could have a mutual gaze here is Teppo (line
22), because the two others are not looking at him (lines 20–21).
During his second turn (lines 27–31), Toni’s face is turned slightly to the left, in
the direction of MT, and his gaze wanders around until the production of the last three
words (line 31), when he turns to look at MT. This proves that he is also able to use
interlocutors’ gaze as a source of feedback. However, he only manages to establish a
mutual gaze with MT very briefly (lines 31–32). After that, MT turns his gaze away
and signals that he is not willing to take the turn (Jokinen et al., 2009: 307). In this
context, the shift of gaze constitutes a non-verbal request to Toni to continue his
narrative. The 3.5 s pause that follows (line 37) proves that Toni does not understand
the interactional meaning of this non-verbal sign. The female therapist apparently
notices this and puts the request into words (ja sitten, ‘and then’, line 38). Toni takes
the turn without looking at the FT or the other interlocutors (line 39). While speaking
(lines 39, 41, 44), he does not look at the interlocutors either: his gaze wanders around
until the end of the turn (line 44), when he fixes his gaze on the table (line 45). Kalle,
who is sitting next to Toni, turns to look at him during the turn (lines 40), but his gaze
does not evoke any response. Consequently, Kalle turns his gaze away (line 42).
At the end of his turn, Toni keeps his gaze fixed on the table (line 45); he does not
turn to look at any of the interlocutors. This time the lack of gaze does not, however,
cause a delayed response (lines 47). The fact that FT recognises the end of Toni’s turn
(line 44) even without eye contact is probably related to prosodic features. Firstly,
there is a rather long pause (1.2 s) just before the last word (nukkunu, ‘slept’), and
this word is produced with a very quiet voice. Secondly, the penultimate word (siellä,
‘there’) is produced with a creaky voice, which typically indicates the yielding of
the turn in Finnish (Ogden, 2001).
5.2 Three Main Patterns Used for Avoiding Eye Contact in Group Therapy … 107

5.2.3 Looking at One’s Own Hands When Speaking

One of the most frequent gaze patterns in this data corpus consists of looking at one’s
own hands. Three of the seven informants use it regularly: Rikhard and Mikael in
Group A and Kalle in Group B. This pattern often occurs alternately with the two
others (cf. Sects. 5.2.1–5.2.2). It is also the case in the following example, where
Rikhard (Group A) is telling his latest news.

Example 5.3:

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

01 Rikhard: ((looks at his hands)) vähä ↑kaikellais↓ta on ollu,

all kinds of things have happened

02 (0.5)

03 Mikael: ((looks at Rikhard))

___________________

04 Heikki: ((gazes straight ahead))

xxxxxxxxxxxxxxxxxxxxxxx

05 FT: ((looks at Rikhard))

___________________

06 MT: joo:? ((looks at Rikhard))

yes

___

07 (0.4)

08 MT: .hh mitäs ↑kouluun ↓kuuluu, ((turns his gaze straight forward))

well how is it going at school

, xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

09 ((turns to look at Rikhard again))

. _______________________________

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

10 Rikhard: ((looks at his hands)) noo, .hhhhh (.) iha o-, (.)

well quite

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
108 5 Gaze Behavior

11 okoo, (0.3) m'ta >meil on< ↑kokeit tulossa, .hh (.)

okay but we have tests approaching

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

12 ((turns his gaze up)) ((gaze wanders around))

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

13 >totanoi<, (.) >°onkse°< (0.3) ↑torstaina o hi:storiankoe,

er PRT is it on Thursday that there will be a History test

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

14 (0.9) ((turns to look at his hands)) ↑huomenna o:n, (1.3)

tomorrow there will be

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

15 ↑matikankoem mutta, (0.8) viime viikolla oli perjantaina,

a Math test but last week there was on Friday

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

16 (0.6) >°tota°< .hh >°mikä se nyt oli<, k-° .hh ↑ranskankoe?

PRT what was it again a French test

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

17 ((turns his gaze up; looks straight ahead))

18 (0.7)

19 MT: m↑hy? ((looks at Rikhard and nods several times)) ((turns his gaze

PRT
____ , xxxxxxxxxxxxxxx

20 straight forward))

xxxxxxxxxxxxxxxxxx

21 (0.8)

22 Mikael: ((looks at Rikhard))

____________________

23 Heikki: ((gazes straight ahead))

xxxxxxxxxxxxxxxxxxxxxxxx

24 FT: ((looks at Rikhard))

____________________
xxxxxxxxxxxxxxxxxxxxxx
5.2 Three Main Patterns Used for Avoiding Eye Contact in Group Therapy … 109

25 Rikhard: ((looks at his hands)) niin. mutta, (.) kuk-, (.)

yes but

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

26 MT: ((turns to look at Rikhard again))

._________________________________

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

27 Rikhard: ↑ärsyttävää on kun tuo ((turns his gaze up; looks straight
what is annoying is that this

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

28 ahead)) .hhhh (0.2) >°oks'° t'tanoin<, ((turns to look at his

is er PRT

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

29 hands)) (.) .hh ↑Tomas joka om meiäl luokalla se om mustalainen

Tomas who is in our class he is a Gypsy

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

30 ((turns his gaze up; looks straight ahead)) nii se tekee, (0.4)

PRT he does

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

31 ↑sil↓lee; (.) >↑Tomas Dansk< ((turns to look at his hands)) (.)

so that Tomas Dansk

xxxxxxxxxxxxxxxxxxxxxxxxx.___________________________________

32 → ni, (0.4) >sillee< et se SIIRtää ((turns his gaze up; glances

PRT so that he moves

_____________________________________________________________

33 → at the MT; then turns to look at the others)) muv VAATteet (.)

my clothes
__________________

34 → toiseen naulakkoo;

to another coat rack

35 MT: ((looks at Rikhard))

____________________
110 5 Gaze Behavior

36 (0.3)

37 MT: m↑hy? ((turns his gaze straight forward))

PRT

, xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

38 FT: ((looks at Rikhard))

___________________

39 Mikael: ((looks at Rikhard))

___________________
40 Heikki: ((gazes straight ahead))

xxxxxxxxxxxxxxxxxxxxxxxx

__________________

41 Rikhard: ja käy siirtelemäs nimilappuja; ((looks at the others))

and goes around moving name tags

_______________________________________________________

42 MT: ((looks at Rikhard))

___________________

43 MT: m↑hy? ((turns his gaze straight ahead and nods))

PRT

, xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

44 (1.3)

______________________________________________

45 Rikhard: °kai° ↑naula°koiltah°. ((looks at the others))

° PRT from the coat racks

46 MT: ((turns to look at Rikhard))

. __________________________

47 (0.3)

, xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
48 Rikhard: ((turns his gaze straight ahead; then turns to look

xxxxxxxxxxxxxx

49 at his hands))
5.2 Three Main Patterns Used for Avoiding Eye Contact in Group Therapy … 111

50 MT: joo:? ((nods and turns his gaze straight ahead))

yes

, xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

51 (2.8)

xxxxxxxxxxxxxxxxxxxxx

52 Rikhard: ((looks at his hands))

53 Mikael: ((looks straight ahead))

xxxxxxxxxxxxxxxxxxxxxxxx

54 Heikki: ((gazes straight ahead))

xxxxxxxxxxxxxxxxxxx

55 FT: ((looks at Rikhard))

___________________

56 MT: ↓haluutko kertoo vielä muuta? ((turns to look at Rikhard))

is there something else you would still like to tell

. ________________________________________________________

At the beginning (line 01), Rikhard is finishing a narrative. At the end of his turn,
he turns to look at his hands. As looking at one’s own hands makes eye contact
impossible, this gaze behaviour does not elicit feedback or possible turn-taking from
the interlocutors (Jokinen et al., 2009: 307). On the contrary, it implies that the speaker
still wants to continue. Consequently, MT reacts to Rikhard’s turn with the dialogue
particle joo (‘yes’) (line 06) which is lengthened and produced with a rising pitch.
This indicates that the recipient has heard and understood what has been said, but
still expects the speaker to continue (Hakulinen et al., 2004: § 798, § 800). However,
Rikhard does not seem to understand this implicit request to continue speaking,
because he does not take the turn spontaneously (line 07). Consequently, MT puts
the request into words by formulating a question (line 08). He also turns to look at
Rikhard (line 09). Rikhard is still looking at his hands and takes the turn without
looking at MT (line 10). The lack of eye contact does not, however, constitute an
atypical feature here, because it has been shown that the participant asking a question
looks at the recipient more often than vice versa and that in this type of sequential
position the lack of eye contact is not treated as problematic by the speaker (Rossano
et al., 2009).
Rikhard looks at his hands during the whole turn that follows (lines 10–16). At
the end of it (line 17), he turns his gaze straight ahead and does not try to establish
mutual gaze with the interlocutors—neither while speaking nor after the end of the
turn—even though all interlocutors except for Heikki are looking at him at the end of
the turn. This could cause uncertainty about the speaker’s willingness to continue his
narrative as well as about his interest in receiving feedback from others. After a 0.7-s
pause (line 18), MT reacts to what Rikhard has just said with the dialogue particle mhy
112 5 Gaze Behavior

produced with a rising pitch (line 19). This signals that the recipient treats the previous
turn as new information and expects the current speaker to continue (Hakulinen et al.,
2004: § 798, § 800). After this (lines 19–20), MT turns his gaze straight ahead, away
from Rikhard, as a sign of unwillingness to take the floor (Jokinen et al., 2009: 307).
Rikhard takes the turn without turning to look at MT (line 25) after a 0.8 s pause (line
21). Thus, even without eye contact, Rikhard interprets the interactional meaning of
MT’s reaction correctly: he compensates for the lack of eye contact by processing
other cues (dialogue particle, rising pitch, and pause that follows). During his turn,
he mostly looks at his hands or straight ahead, away from the others (lines 27–32).
However, during the last utterance (lines 32–34), which seems to be the climax of
the story, Rikhard first glances at MT and then turns to look at the others. This shows
that even if Rikhard mostly avoids eye contact, he knows how to use gaze as a source
of feedback.
The male therapist (MT) again reacts to Rikhard’s turn with the dialogue particle
mhy produced with a rising pitch (line 37). The reaction comes after a brief 0.3 s
pause (line 36). The quick reaction might be related to the fact that the speaker is
looking at the interlocutors at the end of his turn (lines 32–34) (Kendon, 1967).
Rikhard, who is still looking at the interlocutors, continues speaking immediately
after MT’s turn (line 41). MT also looks at Rikhard at this moment (line 42) and
reacts to his turn without delay (line 43). Indeed, this extract supports Kendon’s
(1967) results, according to which mutual gaze tends to reduce the amount of time
that passes between the end of the turn and the reaction of the interlocutor.
The male therapist again uses the dialogue particle mhy produced with a rising
pitch (line 43), associated with a nod and a shift of gaze away from Rikhard. Rikhard
is still looking at the interlocutors and interprets MT’s reaction correctly as an implicit
request to continue speaking (Hakulinen et al., 2004: § 798, § 800; Jokinen et al.,
2009: 307): after a 1.3-s pause, he takes the turn again (line 45). The fact that he
takes the turn only after a short delay might be related to the fact that he has already
been eliciting feedback with his gaze for some time at this point (lines 32–45) and
might expect someone else to make a comment. As no one wants to take the floor,
and as MT signals for him to continue, Rikhard takes the turn again (line 45). While
speaking, Rikhard still looks at the others to elicit feedback. As no one takes the turn,
and as Rikhard apparently has nothing to add, he turns his gaze away at the end of his
turn (lines 48–49). MT reacts to this with another dialogue particle, joo (‘yes’), which
is lengthened and carries a rising pitch (line 50). At the same time, MT nods and
turns his gaze straight ahead, away from Rikhard. The combination of these features
indicates that MT has heard and understood Rikhard’s turn but expects—or at least
wants to give the impression of expecting—him to continue talking about what has
happened (Hakulinen et al., 2004: §798, §800).2 Rikhard does not, however, take the
turn anymore; he just looks at his hands during the 2.8 s pause that follows (lines

2 Rikhard has been talking about a bully at school. It is typical for children with ASD to be bullied
at school, but they may not even necessarily notice it themselves. Bullying is also the main theme
of this therapy session. For these reasons, the therapist probably wants to hear the boy’s whole story
without interfering.
5.3 Conclusion 113

51–52). After this, MT turns to look at Rikhard and asks him explicitly if he still has
something else to add (line 56).

5.3 Conclusion

This chapter (based on Wiklund, 2012) has treated the gaze behaviour of pread-
olescent Finnish boys with ASD. All seven participants of the study avoid direct
eye contact with their interlocutors when producing spontaneous speech. Various
patterns are used to do this. The most frequent patterns consist of (1) fixing one’s
gaze straight ahead, (2) letting one’s gaze wander around, and (3) looking at one’s
own hands when speaking. According to Jokinen et al. (2009), neurotypical speakers
look at listeners about 60% of the time. The informants of this study look at the inter-
locutors much less than that: generally, they do not look at the interlocutors at the
beginning or the middle of their turn. However, it is noteworthy that they sometimes
do turn to look at the interlocutors (or at one of them) at transition-relevance places
(TRP) at the end of their turns.3 Even if establishing eye contact with the interlocutors
at TRPs is far from systematic in this data corpus, the fact that all participants of the
study do it at least occasionally suggests that—despite the general avoidance of eye
contact that characterises their interaction—they are able to use gaze as a source of
feedback (Goodwin, 2000). The subjects of this study also show other remarkable
interactional skills. For example, they can launch narratives and correctly interpret
the implicit conversational meanings of combinations of dialogue particles, prosodic
features, and gestures.
Previous studies have shown that mutual gaze is particularly important at the
ends of turns (Jokinen et al., 2009; Kendon, 1967). At the beginning of a turn, even
neurotypical speakers often look away from the interlocutors (Jokinen et al., 2009).
Therefore, this feature is not typical only of people with ASD. During the middle of
the turn, people with ASD may look at their interlocutors much less than neurotypical
persons do. However, this does not really seem to affect the fluidity of interaction.
Therefore, people with ASD should mostly be encouraged to turn to look at their
interlocutors at the TRPs of their speech.
Looking at the speaker when listening seems to be easier than looking at listeners
when speaking. All participants of the study turn to look at the speaker when listening
to a question at least some of the time. This gives proof of a certain level of non-verbal
competence and the participants’ ability to orientate to question–answer adjacency-
pair sequences (Kremer-Sadlik, 2004; Rossano et al., 2009). Some participants even
look spontaneously at the speaker all the time when listening to a narrative. Conse-
quently, the speaker can receive visual feedback from them (Goodwin, 1981: 57–59;
Kallio & Ruusuvuori, 2011: 81). The boys also show their ability to understand the
function of gaze as an embodied turn-allocation (Kääntä, 2010). Nevertheless, it
is also common for the boys to not look at the speaker at all when listening. This

3 That is, at places where the transition of the turn to another participant is sequentially relevant.
114 5 Gaze Behavior

could easily be interpreted as a lack of interest in what is being said and can affect the
fluidity of the conversation (Goodwin, 1981; Kallio & Ruusuvuori, 2011; Trevarthen,
1984).

References

Argyle, M., & Cook, M. (1976). Gaze and mutual gaze. Cambridge University Press.
Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind. MIT Press.
Bavelas, J. B. (2005). Appreciating face-to-face dialogue. In E. Vatikiotis-Bateson (Ed.), AVSP-
2005. http://www.isca-speech.org/archive_open/avsp05.
Brooks, R., & Meltzoff, A. N. (2005). The development of gaze following and its relation to
language. Developmental Science, 8(6), 535–543.
Charman, T. (2004). Why is joint attention a pivotal skill in autism? Philosophical Transactions of
the Royal Society of London, 358, 315–324.
Cummings, L. (2009). Clinical pragmatics. Cambridge University Press.
Dawson, G., Toth, K., Abbott, R., Osterling, J., Munson, J., Estes, A., & Liaw, J. (2004). Early
social attention impairments in autism: Social orienting, joint attention and attention to distress.
Developmental Psychology, 40(2), 271–283.
D’Entremont, B., Hains, S. M. J., & Muir, D. W. (1997). A demonstration of gaze following in
3–6 month-olds. Infant Behavior and Development, 20(4), 569–572.
Frith, U., & Frith, C. (2001). The biological basis of social interaction. Current Directions in
Psychological Science, 10(5), 151–155.
Gillberg, I. C., & Gillberg, C. (1989). Asperger syndrome—Some epidemiological considerations:
A research note. Journal of Child Psychology and Psychiatry, 30, 631–638.
Goodwin, C. (1979). The interactive construction of a sentence in natural conversation. In G. Psathas
(Ed.), Everyday language: Studies in ethnomethodology (pp. 97–121). Irvington Publisher.
Goodwin, C. (1981). Conversational organization: Interaction between speakers and hearers.
Academic Press.
Goodwin, C. (2000). Action and embodiment within situated human interaction. Journal of
Pragmatics, 32, 1489–1522.
Hakulinen, A., Vilkuna, M., Korhonen, R., Koivisto, V., Heinonen, T. R., & Alho, I. (2004). Iso
suomen kielioppi. SKS.
Hedge, B., Everitt, B., & Erith, C. (1978). The role of gaze in dialogue. Acta Psychologia, 42,
453–475.
Heritage, J., & Maynard, D. (Eds.). (2006). Communication in medical care: Interaction between
primary care physicians and patients. Cambridge University Press.
Hutchby, I., & Wooffitt, R. (2008). Conversation analysis (2nd edn.). Polity Press
Hutt, C., & Ounsted, C. (1966). The biological significance of gaze aversion with particular reference
to the syndrome of infantile autism. Behavioral Science, 11(5), 346–356.
Jokinen, K., Nishida, M., & Yamamoto, S. (2009). Eye-gaze experiments for conversation moni-
toring. In Proceedings of the 3rd IUCS (pp. 303–308). https://doi.org/10.1145/1667780.166
7843.
Kääntä, L. (2010). Teacher turn-allocation and repair practices in classroom interaction: A
multisemiotic perspective. [Doctoral dissertation]. University of Jyväskylä,
Kallio, A., & Ruusuvuori, J. (2011). Kertojan katse tarinankerronnassa. In P. Haddington & L.
Kääntä (Eds.), Kieli, keho ja vuorovaikutus. Multimodaalinen näkökulma sosiaaliseen toimintaan.
SKS.
Kendon, A. (1967). Some functions of gaze direction in social interaction. Acta Psychologica, 26,
22–63.
References 115

Kremer-Sadlik, T. (2004). How children with Autism and Asperger Syndrome respond to questions:
a ‘naturalistic’ theory of mind task. Discourse Studies, 6(2), 185–205.
Kylliäinen, A. (2007). Face and gaze processing in children with autism. [Doctoral dissertation].
University of Tampere.
Lehtinen, M. (2012). “Aiheesta toiseen”—Ymmärrysongelmatilanteita Asperger-nuorten kuntou-
tuskeskusteluissa. In: L. Leskelä & C. Lindholm (Eds.), Haavoittuva keskustelu. Keskustelu-
nanalyyttisia tutkimuksia kielellisesti epäsymmetrisestä vuorovaikutuksesta (pp. 235–277).
Kehitysvammaliitto.
Lewis, F. M., Woodyatt, G. C., & Murdoch, B. E. (2008). Linguistic and pragmatic language skills
in adults with autism spectrum disorder: A pilot study. Research in Autism Spectrum Disorders,
2(1), 176–187.
Loveland, K. A., & Landry, S. H. (1986). Joint attention and language in autism and developmental
language delay. Journal of Autism and Developmental Disorders, 16(3), 335–349.
Martin, I., & McDonald, S. (2004). An exploration of causes of non-literal language problems in
individuals with Asperger syndrome. Journal of Autism and Developmental Disorders, 34(3),
311–328.
Mundy, P. (1995). Joint attention and social-emotional approach behavior in children with autism.
Development and Psychopathology, 7, 63–82.
Mundy, P., & Newell, L. (2007). Attention, joint attention and social cognition. Current Directions
in Psychological Science, 16, 269–274.
Nieminen-von Wendt, T. (2004). On the origins and diagnosis of Asperger syndrome. A clinical,
neuroimaging and genetic study. [Doctoral dissertation]. University of Helsinki.
Ogden, R. (2001). Turn transition, creak and glottal stop in Finnish talk-in interaction. Journal of
the International Phonetic Association, 31, 139–152.
Rossano, F. (2013). Gaze in conversation. In J. Sidnell & T. Stivers (Eds.), The handbook of
conversation analysis (pp. 308–329). Wiley Blackwell.
Rossano, F., Brown, P., & Levinson, S. (2009). Gaze, questioning, and culture. In J. Sidnell (Ed.),
Conversation analysis: Comparative perspectives (pp. 187–249). Cambridge University Press.
Sacks, H., Schegloff, E., & Jefferson, G. (1974). A simplest systematics for the organization of
turn-taking for conversation. Language, 50(4), 696–735.
Scott, D.-W. (1985). Asperger’s syndrome and non-verbal communication: A pilot study. Psycho-
logical Medicine, 15(3), 683–687.
Solomon, O. (2004). Narrative introductions: Discourse competence of children with autistic
spectrum disorders. Discourse Studies, 6(2), 253–276.
Szatmari, P., Brenner, R., & Nagy, J. (1989). Asperger syndrome: A review of clinical features. The
Canadian Journal of Psychiatry, 34, 554–560.
Tantam, D. (1991). Asperger syndrome in adulthood. In U. Frith (Ed.), Autism and Asperger
syndrome (pp. 147–183). Cambridge University Press.
Tantam, D., Holmes, D., & Cordess, C. (1993). Non-verbal expression in autism of Asperger type.
Journal of Autism and Developmental Disorders, 23(1), 111–133.
Trevarthen, C. (1984) Emotions in infancy: Regulators of contact and relationships with persons.
In K. Scherer & P. Ekman (Eds.), Approaches to emotion (pp. 129–157). Erlbaum.
Wiklund, M. (2012). Gaze behavior of pre-adolescent children afflicted with Asperger syndrome.
Communication & Medicine, 9(2), 173–186.
Wiklund, M. (2016). Interactional challenges in conversations with autistic preadolescents: The
role of prosody and nonverbal communication in other-initiated repairs. Journal of Pragmatics,
94, 76–97.
Wiklund, M., & Stevanovic, M. (2018). Ymmärrysongelmat ja vuorovaikutustaitojen opet-
taminen lievästi autististen varhaisnuorten kuntoutuskeskusteluissa [Understanding problems
and the teaching of social interaction skills in group therapy sessions of preadolescents with
high-functioning autism]. Psykologia, 53(05–06), 421–451.
Chapter 6
Therapists’ Response Strategies
and Teaching Orientations

Abstract This chapter is directed towards therapists’ actions in group sessions

involving French- and Finnish-speaking preadolescent boys with autism spectrum
disorder (ASD). On the one hand, the focus is on how two therapists orient the group
towards achieving meaningful learning outcomes with regard to the topic of conver-
sation. The analysis concentrates on the therapists’ output or response strategies when
they orient the group discussion and assess speech produced by the children, either
validating it or parts of it or inviting them to provide more valid input. The results
show that although a specific response category may have many functions, the aim of
all of them is to maintain intersubjectivity among the participants (Wiklund & Määttä,
2021). On the other hand, the chapter discusses the therapists’ teaching orientations.
The results of the analyses suggest that prosodic skills are often overlooked in group
sessions and that the focus of the sessions tends, implicitly or explicitly, to be on
the general norms of social interaction and embodied social interaction (Wiklund &
Stevanovic, 2018).

Keywords Autism · ASD · Group therapy · Response strategies · Teaching

orientations · Interactions · Conversation analysis

6.1 Therapists’ Response Strategies in a Group Therapy

Session Involving Preadolescent Boys with ASD

This sub-chapter focuses on the therapists’ response strategies in a group session

involving four 11–13 year-old French-speaking boys with ASD. The chapter is based
on a study by Wiklund and Määttä (2021). After an introduction (Sect. 6.1.1), the
five most common response strategies will be illustrated with the help of examples
(Sect. 6.1.2). Conclusions will be presented at the end of the chapter (Sect. 6.3), after
the sub-chapter focusing on teaching orientations (Sect. 6.2).

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023 117
M. Wiklund, Speech and Interaction of Preadolescents with Autism Spectrum Disorder,
Prosody, Phonology and Phonetics, https://doi.org/10.1007/978-981-19-8117-3_6
118 6 Therapists’ Response Strategies and Teaching Orientations

6.1.1 Introduction

According to Cummings (2009: 56), atypical communication patterns and the failure
to use language appropriately or effectively constitute a prominent feature of ASD.
Specific problems concern the production and comprehension of speech acts, under-
standing non-literal language, the ability to draw upon contextual information when
interpreting the interlocutor’s speech, and the acquisition of conversational skills
such as turn-taking (Cummings, 2009: 56, 2014b: 49). The tendency of persons with
ASD to understand things literally and to miss implicit messages in interactions has
also been widely documented (Cummings, 2009; Lehtinen, 2012; Lewis et al., 2008;
Martin & McDonald, 2004; Wiklund, 2016). ASD is further characterised by the
avoidance of eye contact and other anomalies in relation to gaze and smiling (Hutt &
Ounsted, 1966; McPartland & Klin, 2006; Siegel, 1996), as well as by problems
concerning the recognition of affective prosody and non-verbal cues in the speech
of interlocutors (Korpilahti et al., 2007).
Deficits regarding narrative and conversational skills as well as skills linked to the
logical and topical continuity of discourse have been attested among adults and chil-
dren with ASD (Asberg, 2010; Capps et al., 1998; Cummings, 2014a, 2014b; Hale &
Tager-Flusberg, 2005a; Losh & Capps, 2003; Solomon, 2004). Studies focusing
specifically on minors show that children with ASD report deficits in gestural joint
attention skills, which also impairs language development (Mundy et al., 1990). They
also experience difficulties with the interpretation of messages on the basis of the
linguistic context (Loukusa et al., 2007) and with the pragmatic use of language,
perspective-taking, and shared understanding (Baron-Cohen, 1995; Baron-Cohen
et al., 1999; Eales, 1993; Happé, 1994; Kleinman et al., 2001; Tager-Flusberg, 1993,
2001). The interpretation of mental states, metaphors, and other forms of figura-
tive language is another common problem among children with ASD (Dennis et al.,
2001).
At the same time, children with ASD engage in discourse practices that are similar
to those of their neurotypical peers in many respects (Ochs & Solomon, 2004: 139;
Solomon, 2001, 2004; Wiklund, 2012). For example, Kremer-Sadlik’s (2001, 2004)
findings demonstrate that children with ASD are able to participate in question–
answer adjacency-pair sequences relatively competently, and several studies report
that they have highly developed abilities to manage various problems in interac-
tion both on their own initiative and by responding to requests for clarification
(Dindar et al., 2016; Geller, 1998; Ohtake et al., 2011; Volden, 2004). It has also
been suggested in several studies that preadolescents with ASD are able to correctly
interpret the conversational functions of combinations of dialogue particles, prosodic
features, and gestures (see Chap. 5) (Wiklund & Stevanovic, 2018; Wiklund, 2012).
Korkiakangas et al. (2012) even showed how a child with ASD was able to discrim-
inate correctly between different possible actions suggested by prosodic features in
the carer’s speech when the carer repeated the child’s utterances. This finding is
6.1 Therapists’ Response Strategies in a Group Therapy Session Involving … 119

in line with previous research results indicating that types of actions indicated by
repetitions can be inferred from prosodic features (Couper-Kuhlen, 1996).
Therapeutic care for children with ASD is based on the idea that systematic
individual and group therapy may improve their linguistic and interactional skills
(see, e.g., Wiklund & Stevanovic, 2018; Wiklund & Laakso, 2020). This idea stems
from research, several longitudinal studies having explored the evolution of these
skills in groups of children. Hale and Tager-Flusberg (2005b), for example, found
that, over the course of one year, children with ASD made significant progress in being
able to maintain a discourse topic. According to Wiklund and Stevanovic (2018), in
turn, at the same time as solving problems related to intersubjectivity, therapists also
implicitly orient preadolescents with ASD towards understanding the general norms
of everyday symmetrical interaction. Joint attention skills such as smiling can also be
taught successfully to children with ASD (Gena et al., 2005; Krstovska-Guerrero &
Jones, 2013).
There has not been extensive research on the speech patterns of neurotypical
people communicating with children with ASD, however, although the interaction
between professionals and their students, clients, or patients has been studied widely
in other contexts (see, e.g., Jones, 2015). In one of the few studies focusing on autism,
Maynard (2019) argues that when a child with ASD shows resistance or a lack of
cooperation as a reaction to an action, for example, neurotypical professionals tend
to ‘transpose’ these phenomena into features characterising the children even though
the actions were prompted by the professionals. He further suggests that behaviour in
interaction should be analysed holistically and socially rather than as emanating from
individuals. Overall, given the lack of any systematic analysis of how neurotypical
therapists respond to turns produced by children with ASD, the aim of the study by
Wiklund and Määttä (2021) presented here was to narrow this significant gap in the
literature.
The data consist of 55 min of naturally occurring interaction in a group therapy
session at a private clinic in Geneva (Switzerland) in 2016. The session involved
four 11–13 year-old boys with ASD and two female therapists. The theme of this
particular session was taunting. (For a more detailed description of the data, see
Sect. 1.2.2). Given that the goal of the session was to develop interaction and group
working skills, the initial analysis of the study (Wiklund & Määttä, 2021) focused
on the linguistic and interactional features of successful communication. On the
second level of analysis, the therapists’ orientation towards prompting examples of
taunting that are meaningful in terms of the group’s learning outcome emerged as a
salient characteristic of the data. In addition, it turned out that there was considerable
variation with regard to how the therapists responded to the boys’ input.
120 6 Therapists’ Response Strategies and Teaching Orientations

In order to analyse and categorise the responses, the data were divided into
sequences, each consisting of a boy’s turn (or several turns produced by the boys
if they occurred consecutively without the therapist’s intervention) and the ther-
apist’s response (or responses by the two of them if they occurred consecutively
without the child’s intervention). Using this method, 277 sequences were identified
(Wiklund & Määttä, 2021). Subsequently, the response strategies were categorised
and their occurrences counted. Instead of relying on previously established categories
of typical responses (e.g., Gill & Maynard, 2006; Lee, 2013), the classification was
based on phenomena that emerged from the data after the observation, in accordance
with the principles of CA (e.g., Hutchby & Wooffitt, 2008). The categories are not
absolute: response strategies belonging to the same category may, in fact, be used
in rather different ways and with different interactional consequences. Nevertheless,
this rough labelling and quantification give useful information about the proportions
of the response categories in the data. The results of this categorisation are listed as
follows (Wiklund & Määttä, 2021):
Approval (86 occurrences)
Comment on the contents of a turn/turns (57 occurrences)
Question addressed to the whole group (30 occurrences)
Question concerning the contents of a turn/turns (26 occurrences)
Rephrasing the contents of a turn/turns (24 occurrences)
New question after an answer (21 occurrences)
Ignoring a turn (19 occurrences)
Giving the floor to another participant (19 occurrences)
Repetition of a turn (17 occurrences)
Change of topic (13 occurrences)
Answer to a question (12 occurrences)
Partial repetition of a turn (10 occurrences)
Seeking approval (10 occurrences).
Most of the sequences include several response categories, such that the 19 cate-
gories listed above occur in various combinations. In the following sub-chapters,
the focus will be on the five most common categories: approval, comment on the
contents of the boys’ turn or turns, question addressed to the whole group, question
about the contents of a turn, and rephrasing the contents of a turn.
6.1 Therapists’ Response Strategies in a Group Therapy Session Involving … 121

6.1.2 Five Most Common Response Strategies

6.1.2.1 Approval

Example 6.11 :

01 Therapist 2: est-ce /que tout le monde est là René ? ((turned

is everyone is here René

02 towards her notes but gazes at René))

03 René: =euh: (0.6) on (0.4) euh::: (.) à part /Laurent
er we er except for Laurent

04 tout le monde est là. ((smiles))

everyone is here

05 Therapist 2: =oui. ((nods twice, takes notes))

yes

06 Therapist 1: =\#oui#\, ((turns her body towards a corner,

yes

07 places a writing board there))

08 (1.5)

09 Therapist 2: super. ((gazes at her notebook))

great

Example 6.1 comes from the beginning of the session (0:01:30) and illustrates
approval, a strategy that is very similar to the confirmation strategy identified by
Gill and Maynard (2006) in physician–patient encounters. Therapist 2 asks René if
everyone is present (line 01). The boy smiles and answers that everyone is present
except for Laurent (who had been present at the first session but who had left the
group) (lines 03–04). The first part of his utterance is characterised by hesitation
and a potential delivery problem, as indicated by the pauses and the lengthened
hesitation sounds (euh). Both therapists react immediately to René’s turn with the
response particle oui (‘yes’), indicating approval (lines 05–06): they have heard and
understood what the boy has said, they agree with him, and they do not expect him
to continue.
On the verbal level, the therapists act as a team here, which is a typical feature of
these data. Therapist 2 asks the original question (line 01), and she also answers first
(line 05). After Therapist 1 responds (line 06) and a 1.5 s pause (line 08), Therapist

1 Therapist 2 is turned towards her notes but gazes at René during René’s whole turn (lines 03-04)

after the word on (‘we’, line 03).

122 6 Therapists’ Response Strategies and Teaching Orientations

2 produces an evaluative response: super (‘great’, line 09). This word, produced
with a falling pitch, is another element indicating approval and positive feedback.
In addition, it functions as a conclusive element indicating the end of the opening
phase of the session and projecting the transition to the intended theme.
Significantly, the therapists do not reciprocate the smile that ends the boy’s turn.
Therapist 2 turned towards her notepad and takes notes, but she gazes at René as
she asks her question and for the main part of René’s answer; she nods twice when
she hears his response. In other words, while focusing mainly on her notepad, she
also acknowledges the child’s turn. Therapist 1 leans towards the corner of the room
holding a whiteboard, which she places in the corner. She does not look at the
other participants during the sequence. Paradoxically, whereas René’s linguistic skills
are atypical, his joint attention skills are highly developed. At the same time, the
therapists’ focus on the writing material impairs their joint attention.

6.1.2.2 Comment on the Contents of a Turn/Turns

Example 6.2:

01 René: >mais /moi je me suis couché< à minuit, ((smiles))

but me, (PRO) I went to bed at midnight

02 Therapist 2 ((touches her mouth, smiles while gazing at her notepad))

03 Therapist 1 ((gazes at René))

04 (0.6)

05 Therapist 1: wow: >tu t’es levé à quelle heure?<

wow at what time did you get up

06 (0.7)

07 René: <à huit /heures trente-et-une.>

at eight thirty-one

08 Therapist 1 (([as René finishes his turn] turns her gaze towards her

09 laptop))

10 Therapist 2 (([as René finishes his turn] smiles, nods slightly, starts

11 to write on her notepad))

12 (0.9)
6.1 Therapists’ Response Strategies in a Group Therapy Session Involving … 123

13 Therapist 1: trente-et-une [c’est précis ]((smiles))

thirty-one that’s precise

14 Alexandre: [<oh /pas de chance>]

oh, that’s too bad

15 (0.4)

16 Therapist 1: ouais (.) ça fait une /courte nuit ((smiles, turns the

yeah that’s a short night

17 laptop screen towards the keypad at 45 degrees)).

This extract includes two types of response. First, Therapist 1 asks René a question,
requesting further information regarding the contents of his turn (line 05), to which
the boy gives a very precise answer (line 07), which is a typical feature in responses
produced by children with ASD: they tend to give answers that are either short
and unnecessarily precise or long and extremely detailed (Wiklund & Stevanovic,
2018; Wiklund, 2016). Both these features are presumably attributable to the general
difficulty that people with ASD have in seeing things from the perspective of other
people (Baron-Cohen, 1995; Baron-Cohen et al., 2013). The therapist reacts to René’s
response first by turning her gaze towards the laptop and producing a comment that
the answer he gave was very precise (trente-et-une c’est précis, ‘thirty-one, that’s
precise’), accompanied by a smile (line 13); the verbal response is preceded by
a pause that lasts almost one second (line 12). After a pause, the same therapist
continues her turn with another comment, (ouais ça fait une courte nuit, ‘yeah that’s
a short night’, line 16). This comment appears to highlight the fact that the boy’s
unnecessarily precise answer does not lead to a breakdown of intersubjectivity. In fact,
it is prompted by another boy, Alexandre, who comments on René’s answer by saying
oh pas de chance (‘oh, that’s too bad’, line 14). The therapist’s turn overlaps with this
turn, potentially indicating her willingness to maintain control of the unfolding of
the conversation, which corresponds to her educational role. Overall, both comments
could be considered examples of implicitly positive feedback. At the same time, the
role of artefacts is significant in this exchange. Therapist 2 gives minimal responses
by smiling and nodding slightly, but she does not look at René at all and concentrates
on her notepad instead. Meanwhile, Therapist 1’s verbal response and smile are
preceded by a quick glance at the laptop: it is only during her final assessment
of René’s answer that she turns the screen towards the keypad, which allows her
to concentrate fully on the exchange. In other words, in this example as well, the
therapists’ joint attention is jeopardised by written materials.
124 6 Therapists’ Response Strategies and Teaching Orientations

6.1.2.3 Question Addressed to the Whole Group

Example 6.3:

01 ((Therapist 1 writes on her noteboard; therapist 2 gazes at her

02 notes))
03 Cédric: <ben ils sont ils sont plus ((Therapist 1 starts to gaze at
well they are they are

04 Cédric)) grands ((Therapist 2 glances Cédric)) que moi et ils

bigger than me and they

05 m’appellent petit.>

call me little
06 (.)
07 ((Therapist 1 starts to write on her noteboard; Therapist 2

08 glances at Cédric))

09 Therapist 1: [petit.]
little
10 Therapist 2: [↑mm::h↓] ((Redirects her gaze toward her notes))

11 (0.3)
12 Cédric: <et une /fois je lui ai /répondu tais-toi le vieux.>

and once I replied shut up old man

13 (0.6)
14 Therapist 1: ↑hmm::↓
15 (.)

16 Therapist 2: ↑d’a↓ccord.
okay

17 Therapist 1: =c’est une réaction qu’on peut avoir.

it is a reaction that one can have

18 Therapist 2: =c’est une réaction °oui°.

it is a reaction yes

19 (0.3)

20 Therapist 1: et est-ce que /petit ((raises her gaze toward the ceiling
and is little

21 from her noteboard)) c’est toujours méchant c’est toujours

is it always nasty is it always

22 de la moquerie quand on /dit il est petit.

taunting when one says he is little

23 (0.6)

24 René: euh /non,

er no
6.1 Therapists’ Response Strategies in a Group Therapy Session Involving … 125

A question addressed to the whole group indicates clearly that the child’s input
is not entirely satisfactory. Cédric says that bigger children have called him little
(lines 03–05). At the beginning of this sequence, Therapist 2 is gazing at her notes,
whereas Therapist 1 is writing on her whiteboard—overall, note-taking and writing
artefacts compromise the therapists’ joint attention in this excerpt as well. Therapist
1 starts to gaze at Cédric when he produces the word plus (‘more’), and Therapist 2
glances quickly at him as he voices the word grands (‘big’), and again at the word
petit (‘small’). At this point, Therapist 1 starts to write on her whiteboard, and she
repeats the word little (line 09). Simultaneously, Therapist 2 produces a minimal
response (mm::h, line 10), and she has redirected her gaze towards her notepad at
this point. This minimal response is lengthened and carries a marked pitch, namely
a rise followed by a fall. These features indicate that the speaker treats the preceding
turn as something surprising. At the same time, the particle indicates approval.2
Cédric then continues his story and says that he had once responded to someone by
calling him an old man (line 12). The therapists first react to this turn with response
particles that primarily indicate approval: they have heard and understood what the
boy has said (hmm and d’accord, lines 14 and 16). From a prosodic perspective, both
particles contain features that call for closer scrutiny.
In the particle hmm in line 14, an initial rise in pitch is followed by a lengthened
sound and a fall in pitch. The pitch rises 2.8 semitones (st) during the production of
the [h] sound associated with a schwa. It then falls by 8.4 st during the production
of the sequence of sounds [mm::]. The strong melodicity of the particle creates a
marked tone that appears to indicate surprise: Therapist 1 is surprised by what the
boy has just said and interprets his input as new information. In addition, the particle
is strongly lengthened, which gives time for the production of the pitch movements
and highlights the markedness. It is also interesting that here Therapist 1 repeats the
prosodic shape of the particle that has just been produced by the other therapist (line
10). Consequently, the prosodic shape, which is depicted in Fig. 6.1, also highlights
the fact that the therapists act as a team.
The particle d’accord (‘OK’, line 16), produced by Therapist 2, is characterised
by a strong initial stress and a slightly ascending pitch, followed by a descending
pitch. The pitch rises by 1.0 st during the first syllable [da] and falls by 3.4 st
R
during the second syllable [kO ]. There is a strong initial stress during the produc-
tion of the first syllable. These minor prosodic changes highlight the fact that the
turn concords with the previous turn produced by Therapist 1 (line 14): in addition
to expressing approval, the speaker treats the boy’s input as something new and
surprising. Figure 6.2 shows the prosodic properties of the particle d’accord:

2Due to the overlap of speech, it is not possible to give a more precise prosodic description of this
particle.
126 6 Therapists’ Response Strategies and Teaching Orientations

0.1879

-0.1424
0 0.6305
Time (s)

20
Pitch (semitones re 100 Hz)

12

0
-5
hmm::

0 0.6305
Time (s)

Fig. 6.1 The production of the particle ‘hmm::’ (line 14)

Subsequently, both therapists explicitly express their approval, saying that it is a

reaction that one can have (lines 17–18), and their turns follow each other without
a pause. However, given that the therapists do not comment at all on the contents
of what the boy has just said, it is not certain that they have completely understood
it. Here, their turns constitute “displays of understanding” rather than “proofs of
understanding” (Sacks, 1992).
In this case as well, the therapists clearly act as a team by reacting in a similar
manner and repeating each other’s words and even prosodic patterns. Their responses
constitute positive feedback that encourages the boy to share his experiences, showing
him that his turns were relevant and have been understood. Cédric’s speech is often
rather disfluent and he has difficulties finding words; hence, this type of positive
feedback probably means a lot to him. At the same time, the implication is that the
boy was not able to produce a satisfactory example of taunting.
Following these expressions of approval, Therapist 1 addresses a question to the
whole group (lines 20–22), asking whether calling someone little is always neces-
sarily taunting. Hence, she encourages discussion about a topic that reflects the theme
of the session and opens up the floor for the expression of divergent views, suggesting
that calling someone little does not entail taunting in all cases. Just before the onset
6.1 Therapists’ Response Strategies in a Group Therapy Session Involving … 127

0.202

-0.2288
0 0.3491
Time (s)

20
Pitch (semitones re 100 Hz)

12

-5
d’accord

0 0.3491
Time (s)

Fig. 6.2 The production of the particle ‘d’accord’ (line 16)

of this question, Therapist 1 directs her gaze slightly towards the ceiling, and Ther-
apist 2 continues to gaze at her notepad. René answers the question after a pause,
saying er no (line 24). The question has been addressed to the whole group, and
it approaches the issue on a general level (as indicated by the repetitive use of the
adverb always); therefore, it does not directly take a stance against Cédric’s turns or
question his experience. At the same time, it allows him to reconsider his example.
In truth, although he received superficially positive feedback for his first answer, the
emphasis was on the fact that he answered rather than on what he said. As a result,
the next turn proof procedure (the therapists’ responses) is evidence that his original
answer was considered unsatisfactory, and this probably led to René’s contribution
(line 24).
128 6 Therapists’ Response Strategies and Teaching Orientations

6.1.2.4 Question Concerning the Contents of a Turn/Turns

Example 6.4:

01 ((Therapist 2 writes and gazes at her notepad during the entire

02 sequence))

03 Thomas: >content et fatigué<.

happy and tired

04 (0.3)

05 Therapist 1: ((gazes at Thomas)) mm:::? ((nods)) (.) plus encore? (.)

mm a bit more
06 pourquoi?

why
07 (0.5)

08 Thomas: mm: fatigué parce que je me suis couché

I’m tired because I went to bed

09 à v(h)ingt-d(h)eux /heures, ((smiles))

at 10 p.m.
10 (0.5)

11 Therapist 1: ouais? [c'est]

yeah it’s
12 Thomas: [mm: ] (0.4) content parce que: >je suis

mm happy because I’m

13 content d'être là<
happy to be here
14 (0.5)

15 Therapist 1: ↑o↓kay. ((nods))

okay

16 ((Therapist 2 nods slightly))

This extract is from the beginning of the session and the participants are talking
about how they are feeling that day. As noted in the previous examples of this
sub-chapter, note-taking plays a major role: Therapist 2 takes notes and gazes at
her notepad throughout the entire sequence, whereas Therapist 1 gazes at Thomas
continuously. René has just been speaking, and he gives the floor to Thomas, who
says that he is happy and tired (line 03). First, Therapist 1 expresses approval and
understanding with the response particle mm (line 05), accompanied by a nod. The
particle carries a strongly rising pitch (8.0 st), which indicates that in addition to
6.1 Therapists’ Response Strategies in a Group Therapy Session Involving … 129

0.1064

-0.1112
0 0.5378
Time (s)

20
Pitch (semitones re 100 Hz)

12

0
-5
mm:::

0 0.5378
Time (s)

Fig. 6.3 The prosodic shape of the particle ‘mm’ (line 04)

expressing approval, the speaker invites the previous speaker to continue. It is also
lengthened, which gives time for the production of the pitch rise and emphasises the
above-mentioned functions. Figure 6.3 shows the prosodic form of the particle.
Following the particle mm in line 05, Therapist 1 asks Thomas why he is happy
and tired. Thomas says he is tired because he had gone to bed at 10 p.m., and he
smiles (lines 08–09). The therapist reacts to this with the response particle ouais
(‘yeah’), produced with a rising pitch (line 11), thereby indicating that she wants
to hear more. Thomas interprets the interactional meaning of this particle correctly
and continues to speak, saying that he is happy because he is present at the session
(lines 12–13). This reason for his happiness is delivered at a faster than normal speed
and is preceded by the lengthened syllable que in the compound conjunction parce
que (‘because’). These features could indicate hesitation and orientation towards the
immediate goal, namely giving answers that may be superficial but are still mean-
ingful to the therapists. If this interpretation is correct, the response also indicates an
ability to take the interlocutor’s perspective into account. The therapist reacts to this
with the response particle okay, produced with a rising–falling pitch (line 15) and
indicating approval. It further acknowledges that the previous turn has been heard
and understood and that the previous speaker is not expected to continue.
130 6 Therapists’ Response Strategies and Teaching Orientations

6.1.2.5 Rephrasing the Contents of a Turn/Turns

Example 6.5:

01 ((Therapist 1 writes on her whiteboard during the entire sequence,

02 Therapist 2 writes on her notepad, both gaze at their writing devices))

03 Alexandre: les moqueries ce sont des choses quand (0.4) peut-être pour

taunts are things that maybe for

04 nous qui: sont pas très marrantes mais pour les autres (.)

us are not very funny but for others

05 oui,

yes

06 (1.0)

07 Therapist 1: ↓ou↑ais: ((Therapist 2 raises her gaze slightly)) très

yes very

08 bien,

good

09 (0.5)

10 Therapist 2: ça peut faire ((raises her gaze slightly towards the other

it can make
11 children)) rire certains mais pas: (0.3) ((directs her

some people laugh but not

12 gaze toward Alexandre)) [pas les autres],

not others
13 Alexandre: [pas les autres],
not others
14 (0.3)

15 Therapist 2: /d’accord oui ((nods twice)) c’est une bonne idée ça.

okay yes it’s a good idea that

16 ((nods)) (0.6) okay, ((starts to write again))

okay

In this category of responses, the therapists repeat the contents of a turn produced
by one of the boys in a more elaborate form, providing one interpretation. Members
of the group discuss the concept of taunting or mockery in the excerpt. Therapist 1
has asked the boys collectively what it means to mock someone. Cédric has already
6.1 Therapists’ Response Strategies in a Group Therapy Session Involving … 131

answered, and now Alexandre takes the floor, giving a definition of mockery (lines
03–05). Therapist 1, who gazes at her whiteboard and takes notes during the entire
sequence, reacts by giving explicitly positive feedback: ouais: très bien, (‘yes very
good’, lines 07–08). In this turn, the particle ouais is lengthened and has a marked
pitch: first, the pitch falls by 6.3 st, then it rises by 5.8 st in the middle, and at the
end of the particle, it falls again by 3.5 st. This strong melodicity implies being
impressed and emphasises positive feedback regarding the contents of the previous
turn. Figure 6.4 below shows the prosodic shape of this particle.
Therapist 2 raises her head slightly upon the particle ouais, produced by Therapist
1. Subsequently, she repeats the contents of Alexandre’s turn in another, slightly more
elaborate form (lines 10–12): ça peut faire rire certains mais pas: pas les autres, (‘it
can make some people laugh but not others’). In the middle of her turn, she raises her
head slightly and directs her gaze towards the other children and towards Alexandre
at the end of her turn. There is a lengthened vowel (pas:, line 11) and a pause (line 11)
in her turn. These prosodic features indicate pre-formulation. It is noteworthy that
after the pause, Alexandre says pas les autres (‘not others’) at the same time as the
therapist is speaking (lines 12–13). This action shows Alexandre’s interactional skills:
he notices the hesitation of the therapist and comes to help. After this, Therapist 2
indicates approval and gives positive feedback, stating explicitly that his contribution

0.1863

-0.2444
0 0.6993
Time (s)

20
Pitch (semitones re 100 Hz)

12

0
-5
ouais

0 0.6993
Time (s)

Fig. 6.4 The prosodic shape of the particle ‘ouais’ (line 07)
132 6 Therapists’ Response Strategies and Teaching Orientations

is a good idea (line 15). She nods twice after having said d’accord oui (‘okay yes’),
nods again before the word okay, and starts writing again at the end of her turn, when
she also redirects her gaze towards her notepad.
The therapists in our data sometimes do quite different things in their rephrasing.
This response strategy typically aims at maintaining intersubjectivity, such that the
therapists make sure they understand the boys’ turns even if their utterances are
unclear or their speech is disfluent. Responses in this category are also used for
educational purposes: the therapists’ reformulations of the children’s turns seem to
highlight content that they consider educationally meaningful for the entire group. An
appraisal is another important function of rephrasing. The excerpt discussed above
highlights orientation towards normative language as one of the key educational
goals of the encounter, which is another function of rephrasing in this data. Thus, the
child uses the colloquial adjective marrant (‘funny’) in line 04. Although Therapist
1’s initial feedback is very positive (ouais très bien, ‘yes very good’, lines 07–08),
Therapist 2 (lines 10–12) rephrases Alexandre’s turn using the verb rire (‘laugh’), the
register of which is neutral. She pronounces it with accrued stress, thereby marking
this form as more normative than the adjective marrant, which the child used. Thus,
rephrasing is also typically used to enrich the vocabulary and teach lexical nuances.

6.2 Therapists’ Teaching Orientations in Group Therapy

Sessions Involving Preadolescent Boys with ASD

This sub-chapter focuses on therapists’ teaching orientations in group therapy

sessions involving Finnish-speaking boys with ASD. The sub-chapter is based on
a study by Wiklund and Stevanovic (2018). The sub-chapter will start with a brief
introduction (Sect. 6.2.1), followed by a presentation of implicit teaching orientation
(Sect. 6.2.2) and explicit teaching orientation (Sect. 6.2.3), as well as some thoughts
on prosodic problems from the point of view of teaching orientations (Sect. 6.2.4).

6.2.1 Introduction

The aim of the study described in this sub-chapter (Wiklund & Stevanovic, 2018)3
was to examine the comprehension problems presented in Chap. 4 from the point of
view of guidance interaction. On the one hand, the study focuses on how intersub-
jectivity (i.e., the mutual understanding between the participants) is maintained in
the sessions. On the other hand, the study discusses the therapists’ professional and
pedagogical practices and the specific challenges that the therapists encounter while
trying to achieve the goals of the sessions.

3 The study was published in Finnish.

6.2 Therapists’ Teaching Orientations in Group Therapy Sessions Involving … 133

Studies on guidance and counselling interaction have shown that there are
several different goals that the instructors/trainers/teachers/therapists can orient to
when facing their students/clients during a guidance or counselling interaction.
Vehviläinen (2014), for example, distinguishes four different guidance orientations:
(1) a ‘supportive orientation’ that focuses on showing that the client is being emphati-
cally heard and accepted; (2) an ‘examining orientation’ that focuses on the examina-
tion of different dimensions of the experience, the situation, or the topic that is being
addressed; (3) a ‘problem-solving orientation’ that uses different pieces of advice
and turns which include recommendations or information that provide solutions to
problems or things that are being worked on; and (4) a ‘teaching orientation’ that
aims at structuring the guidance process as a learning process that proceeds according
to a pre-established plan. The first two of these orientations—the supportive orien-
tation and the examining orientation—are typical approaches in therapeutic interac-
tion. Service encounters, in turn, often use the third orientation, the problem-solving
orientation. The study described in this sub-chapter (Wiklund & Stevanovic, 2018)
focuses on the last orientation, the teaching orientation. According to Vehviläinen
(2014), in the teaching orientation, the instructor often uses various proactive means,
such as exercises, models, and other means that support learning and understanding.
In the data of this study, consisting of the Finnish data of the project involving 11–
13 year-old boys with ASD (see Sect. 1.2.1), the teaching orientation is more implicit.
In these data, the therapists clearly aim to teach the boys interactional skills with the
help of various verbal and non-verbal cues. As the following sub-chapters will show,
on the one hand, these cues help to proactively maintain intersubjectivity or to solve
comprehension problems that have already occurred. On the other hand, the thera-
pists’ cues also communicate an understanding of mundane interactional norms that
the boys with ASD are expected to learn to manage little by little as the therapeutic
programme proceeds.
Conversation analytic research on interaction focuses on discovering in detail
how persons with ASD construct their actions in collaboration with other participants.
Dindar et al. (2016), for example, have studied how children with ASD manage inter-
actional challenges. The study shows that children with ASD are sometimes able to
take an active role in solving interactional problems and in maintaining intersubjec-
tivity in conversations. When an interactional problem occurs, the children with ASD
use both verbal and non-verbal means to solve it (Dindar et al., 2016). Stevanovic et al.
(2017), in turn, have studied dyadic conversations between adult men with Asperger
Syndrome (AS) and neurotypical men. Where the neurotypical participants of the
study felt that expressing emotions and even overlapping speech made the conver-
sation a more positive experience, the persons with AS thought that precisely these
features made the conversation a more negative experience (Stevanovic et al., 2017).
Koskinen (2015) compared in her case study the minimal responses and nods of
a participant with AS and his neurotypical interlocutor during narrative sequences.
Where the neurotypical interlocutor used minimal responses and nods as reactions
to the narration of the participant with AS, these reactions were absent in the conver-
sational action of the participant with AS. The responses of the person with AS also
134 6 Therapists’ Response Strategies and Teaching Orientations

often occurred with delays and ignored what was most emotionally important in the
narration of the neurotypical participant (Koskinen, 2015).
Studying the interaction of persons with ASD is important for many reasons: it
helps neurotypical persons to understand autism, and it provides interactional tools
for persons with ASD, their family members, and people working with them. A
more profound understanding of the interactional features characterising ASD also
provides new directions and tools for developing the teaching of social skills targeted
at persons with ASD (Ke et al., 2018). There are positive experiences of applying
the conversation analytic approach pedagogically (Rendle-Short, 2017) as well as of
various movement and dance therapies (Koch et al., 2015).
The examples presented in the following sub-chapters will show how the interac-
tional problems of persons with ASD are manifested in conversations, and how the
therapists react to these problems. On one hand, the analyses describe the causes of
comprehension problems (see also Sect. 4.2) as well as the therapists’ ways of trying
to maintain intersubjectivity despite the problems. On the other hand, the analyses
will show how the therapists use the problems that have occurred to serve peda-
gogical purposes. In these cases, paradoxically, comprehension problems and how
they are typically managed in interaction act as resources for fulfilling the therapists’
institutional functions (Wiklund & Stevanovic, 2018: 426).

6.2.2 Implicit Teaching Orientation

Many previous studies have shown that a literal understanding of metaphors and
proverbs is typical of persons with ASD (Cummings, 2009; Lewis et al.. 2008;
Martin & McDonald, 2004). As has already been demonstrated (see Sect. 4.2.1),
studies carried out within the framework of this project also show that the tendency
to understand things literally is the most common cause of comprehension problems
in the interactions of preadolescents with ASD (Lehtinen, 2012; Wiklund, 2016). In
practice, this tendency may appear, for example, as the concrete understanding of a
word used in an abstract sense, or as difficulty noticing an implicit cue or message in
what has been said. In these data, the informants often reply only to what has been
asked literally and do not understand, for example, an implicit invitation to speak
more about the topic that is being addressed.
The following example illustrates comprehension problems caused by a literal
understanding of speech and the difficulties of a child with ASD in understanding
implicit cues. Indeed, in this extract, a boy with ASD, Jaakko, has obvious problems
in interpreting questions asked by the therapists, and often he replies only to what has
been literally asked. At the beginning of the extract, the male therapist asks him about
a situation of bullying where Jaakko had been pushed by another pupil at school and
he had injured his knee (lines 01–02). Jaakko misunderstands the therapist’s question,
and his answer is slightly off-topic (lines 03–06).
6.2 Therapists’ Teaching Orientations in Group Therapy Sessions Involving … 135

Example 6.64 :

01 MT: .hhhh mites ↑se käsiteltiin se (0.7) #ö#- tönimishomma tai

how was it dealt with that er pushing thing or

02 [#ö °olikos°]

er was
03 Jaakko: [mm ]
04 Jaakko: öö tönimishomma ↓#well#- (0.4) >#jos sitä tehää ↑tarkotuksel se

er the pushing thing well if it is done on purpose it

05 on kiu#saamista<# (.) .hhh silleen (0.3) ↑leikkimieli#sesti ni

is bullying like that playfully so

06 se ei o[o ]#.

it is not
07 MT: [mm] .hhh mut tää (.) ku sun (0.4) polvi venähti

mm but this when your knee got strained

08 nii kuinka se (0.7) [sekin sekin on ]

so how it it also it also is

09 Jaakko: [ihan selkee tönäts] #m# ihan selkee (.)

clearly a push m clearly
10 ei ollu #sel- oli ihan selkee tönäisy#?

it was not cle- it was clearly a push

11 (0.3)

12 MT: m[m ]

13 FT: [mt. ] miten se hoidettiin, ((moves her hands on the table))

how was it taken care of

14 ?: .hh[hh ]

15 Jaakko: [#öö#] (0.4) suo↑raan käsivoimalla, (0.3) ((pushes forwards

er directly by force using hands

16 with his hands))

17 MT: ää

eh
18 FT: .hhh nii mut sen jälkeen. (0.6) hmm
yes but after it hmm

4The same extract has previously been analysed in Finnish in Lehtinen (2012) and in Wiklund
and Stevanovic (2018).
136 6 Therapists’ Response Strategies and Teaching Orientations

19 (1.3)

20 MT: selvit[tely

settling
21 FT: [nii [°muistat (-)°

yes do you remem-

22 Jaakko: [<#ee: selvittelyä#>

eh settling

23 (0.3) #ää# (0.3) #sitä ei tullu >suoraa sen jälkee koska mun

eh it didn’t follow immediately after because I

24 piti< vaan (0.5) pitää se kipu poissa#, ((FT nods))

just needed to keep the pain away

25 (0.5)

26 MT: mm
27 (0.4)

28 Jaakko: #jjä#

(no meaning)

29 (1.1)

30 MT: mt. .hh mut puhutteliko opettaja sitä tönäsijää? tiesiks se

but did the teacher talk to the pupil who pushed you did he know

31 Jaakko: =@ee joo@? (0.4) °#se on meidän ↑luokalta#°,

er yeah he is from our class

32 (0.3)

33 MT: nii-i?

yes

34 (0.9)

35 MT: mitä sitten? ((turns his gaze away from Jaakko for a moment))

then what

36 (0.4)

37 Jaakko: >↑no sitten vaan että #kyl puhut-#< (0.4) #hänelle puhut↑tiin

well then only that they tal- they talked to him yeah

38 (0.2)

39 FT: m[m

40 MT: [mm

41 (2.5)
6.2 Therapists’ Teaching Orientations in Group Therapy Sessions Involving … 137

The comprehension problem emerges when Jaakko misunderstands the male ther-
apist’s question ‘how was it dealt with that er pushing thing’ (line 01). Jaakko does
not understand that the question refers to a specific event in which he was himself
pushed by another pupil at school. Instead, he starts to analyse the nature of ‘pushing’
on a general level (lines 03–06). After Jaakko has answered in a slightly off-topic
manner, the therapist immediately makes a direct repair initiation and backtracks
in the conversation to specify his question (lines 07–08). The therapist uses the
question word kuinka (‘how’) in his repair initiation (line 08). The problem is that
Jaakko understands the question word literally: he starts to explain how the pushing
concretely happened (lines 09–10). The female therapist joins the conversation and
tries to get Jaakko to tell the group about how the bullying case was dealt with at
school (line 13). She uses the verb hoitaa (‘take care of’) in her question (‘how
was it taken care of’) (line 13), and Jaakko interprets it as referring to the way the
pushing was concretely executed by the bully (lines 15–16). This can be inferred
both from the contents of his answer (‘directly by force using hands’) (line 15)
and from the pushing movement of his hands (lines 15–16). The female therapist
reacts to this by formulating a new repair initiation, ‘yes but after it’ (line 18). The
chain of particles (nii mut, ‘yes but’) occurring at the beginning of the turn indicates
that the speaker has understood the preceding turn but would like something to be
added to it from another point of view (Hakulinen et al., 2004: § 799, § 801, § 1034).
After a silence (line 19) and after new requests by the therapists (lines 20–21), Jaakko
answers (lines 22–24). He does not, however, seem to treat the expression sen jälkeen
(‘after it’) as an abstract cue indicating a causal connection but rather as a concrete,
temporal expression: he replies literally to what has been asked, in other words, what
happened immediately after the act of bullying. The male therapist first produces a
turn consisting of the particle mm (line 26) with which he indicates that he has heard
and understood the preceding turn (Hakulinen et al., 2004: § 798). After that, he
produces a direct repair initiation using two interrogative clitics (-kO) (line 30). The
particle mut (‘but’) indicates again that the speaker is looking for another point of
view (Hakulinen et al., 2004: § 1034). In this repair initiation, the therapist expresses
clearly which questions he would like Jaakko to answer. The answer that Jaakko
gives does not, however, resolve the problem of comprehension. First, Jaakko gives
a slightly unclear answer to the polar questions that he has been asked (ee joo, ‘er
yeah’) (line 31), and then he explains his reply with a brief comment (‘he is from our
class’) (line 31). Nevertheless, the turn is topically inconsistent. Firstly, it is not clear
which question Jaakko replies to with the chain on particles ee joo (‘er yeah’) (line
31). Secondly, it is unclear how the second part of the turn (‘he is from our class’)
(line 31) is related to what has been said earlier. It is possible that this longer part is a
justification for the preceding particle joo (‘yeah’) and that the utterance is a reply to
the second question asked by the male therapist: ‘did he know’ (line 30) → ‘yeah he
is from our class’ (line 31). The relationship between the different parts of the turn as
well as the relationship between the whole answer and the questions that have been
asked are, however, slightly unclear and make the unambigious interpretation of the
turn difficult. The male therapist reacts to Jaakko’s turn by producing a particle that
ends in a rising pitch: nii-i (‘yes’) (line 33). With this minimal response, the therapist
138 6 Therapists’ Response Strategies and Teaching Orientations

indicates that he is taking the role of the recipient and in this way gives space for
his interlocutor to continue speaking (Routarinne, 2003). Jaakko does not, however,
continue speaking (line 34). Consequently, the therapist produces an outspoken ques-
tion concerning the continuation of the conversation: mitä sitten (‘then what’) (line
35). Jaakko replies precisely to this question, and the comprehension problem gets
resolved (line 37) as the therapists react to Jaakko’s answer with minimal responses
indicating approval (lines 39–40). Nevertheless, it is possible that Jaakko understands
the therapist’s question literally. In other words, it is possible that he interprets the
word sitten (‘then’) (line 35) literally as an adverb of time and not as an invitation
to justify how his previous comment (‘he is from our class’) (line 31) is related to
the topic. On the one hand, the facts that the boy detaches the word ‘then’ (sitten)
from the established expression ‘then what’ (mitä sitten) (line 35) and still does not
justify his previous comment but only replies to what has been literally asked from
him support the hypothesis that he may still understand the question of the therapist
literally. On the other hand, it is noteworthy that the sensitive nature of the event
may lead to unwillingness to talk about the topic and may therefore influence how
the boy replies to the questions he is being asked.
Thus, the example illustrates (1) how persons with ASD have the tendency to
interpret what is being said literally, and (2) implicit guidance after having problems
with interpreting the speech of the interlocutor. As one can see from the action of
the person with ASD in the example, the tendency to interpret words concretely can
also make it difficult for these persons to understand abstract causal connections (for
example, ‘after it’ or ‘then what’). When this is the case, even outspoken requests
to continue talking about the topic that is being discussed are not necessarily under-
stood by the person with ASD. But even if the problems of comprehension were
rather significant in this example, it is noteworthy that from the point of view of
the therapy, the teaching of interactional skills took place implicitly. Resolving the
interactional challenges was carried out by the therapist using repair mechanisms
of mundane, symmetrical interaction that are typically used when looking for an
adequate answer that is missing or when the answer of the interlocutor seems incon-
sistent (Pomerantz, 1984). Even if one can expect persons with ASD to make progress
in their interactional skills when comprehension problems are resolved by the ther-
apists in this manner, the teaching orientation of the therapist is not emphasised in
such situations. In the example presented above, the teaching orientation becomes
apparent mostly through the fact that—despite the comprehension problems—the
therapists did not let the topic that they started to change according to the responses
of the boy, which would be typical of mundane, symmetrical interaction (Kurhila &
Lindholm, 2016). Instead, the therapists tried repeatedly to get the interaction back
onto its original track.
6.2 Therapists’ Teaching Orientations in Group Therapy Sessions Involving … 139

6.2.3 Explicit Teaching Orientation

As already shown in Chap. 5, persons with ASD often use atypical non-verbal features
in their interactions. These features include, for example, atypical eye contact, facial
expressions, gestures, and postures (McPartland & Klin, 2006; Szatmari et al., 1989;
Tantam et al., 1993; Wiklund, 2012). The direction of gaze is one prominent feature
in the interaction of persons with ASD. Indeed, as already mentioned (see Chap. 5), a
lack of eye contact is typically one of the first symptoms to draw adult attention when
ASD is suspected in a child (Hutt & Ounsted, 1966; Kylliäinen, 2007). If eye contact
occurs, it can be atypical: either staring or evasive (Nieminen-von Wendt, 2004, etc.).
Atypical features of embodied interaction and comprehension problems created by
these features also occur in the data of this project (see Sect. 1.2). The preadolescents
of the data avoid direct eye contact especially when they produce speech, whereas
when they listen to others, maintaining the eye contact seems to be easier for them
(Chap. 5; Wiklund & Stevanovic, 2018; Wiklund, 2012). When producing trouble-
source turns (i.e., speaking turns that cause problems of comprehension), there is no
eye contact between the participants in 84.1% of cases (see Sect. 4.2.3; Wiklund,
2016). Typically, the boys with ASD avoid eye contact by (1) fixing their gaze straight
ahead, (2) letting their gaze wander around, or (3) looking at their own hands (see
Chap. 5; Wiklund, 2012).
Example 6.7, presented earlier (see Sect. 4.2.1), illustrates the explicit teaching
orientation. In this extract, the male therapist tries to resolve a problem of non-verbal
interaction exhibited by Heikki: during the entire session, Heikki has been staring
at a mirror on the opposite side of the room instead of looking at his interlocutors
(Wiklund, 2012) (see Sect. 5.3). In this example, Heikki wants to take the floor (lines
02–03), and the therapist tries to make him look at his interlocutor with an implicit
cue included in his turn: ‘who do you ask’ (line 04). The verbal cue is associated
with pointing hand gestures (lines 05–06).

Example 6.7:

01 MT: onks ↑pojilla mitäs sitte ky↑syisitte?

do the boys have what would you ask then

02 ((Heikki raises his hand and stares at the mirror straight in 03

front of him; FT looks at MT))

04 MT: öö #k- m-# (0.4) ↑tääl↓lä on ↑kukas (0.4) >keneltä sä kysyt<,

er k- m- here is who who do you ask

05 ((FT looks at Heikki; MT glances at Rikhard and points at him 06

with his hand; Heikki glances at MT))

140 6 Therapists’ Response Strategies and Teaching Orientations

07 (.)

08 Heikki: öö[:,]

er

09 MK: [He]ikki ↑keneltä sä kysyt. ((Heikki and Mikael look

Heikki who do you ask

10 straight forwards, the others look at Heikki))

11 (.)

12 Heikki: öö ↑Rikhardilt?

er Rikhard

13 MT: >kat- ja #a-#< missä päin R[ikhard o]n. ((tries to make

loo- and a- in which direction Rikhard is

18 eye contact with Heikki, makes gestures towards Rikhard))

19 Rikhard: [khhhhh,]

20 (0.4)

21 Rikhard: ↓°tä[äl°],

here

18 Heikki: [ni]in tai ↑siis että:,

yes or well I mean

19 MT: ↑missä ↓päin, (.) £heh He(h)ikki [Rikhard on,£]

in which direction heh He(h)ikki Rikhard is

20 ((makes a gesture with his hand again, laughs))

21 Rikhard: [↑lännessä ] vai idässä:]

in the West or in the East

22 FT: [missäs on R]ikhard ↑istuu]

where is Rikhard sitting

25 tässä pöy#dän ääressä#. ((looks at Heikki and leans towards him))

at this table

24 MT: katsotko, (.) ↑niin (.) ↑hänen #suuntaansa#;

would you look in his direction

25 ((Heikki glances in the direction of Rikhard and MT; MT makes a

6.2 Therapists’ Teaching Orientations in Group Therapy Sessions Involving … 141

26 gesture in the direction of Rikhard))

27 Heikki: ai [↑niis ] siis;

oh yes well

28 FT: [£↑mmh,£]

29 MT: £NII, (.) °nii,° katso ↑hänen£ suun#taansa ku sä# [ky]syt,

yes yes look at his direction when you ask

30 ((again makes a gesture towards Rikhard, puts a hand on his

31 hips, moves his legs))

32 Heikki: [ai,]

oh

33 MT: ni,

yeah

34 Heikki: >nii tota:<,? (.) .hhhh ö[ö:, ] ((keeps glancing at

yeah well er

35 Rikhard and the mirror in front of him))

36 MT: [mhh,]

37 (1.0)

38 Heikki: mi↑käs se, (.) mikäs se [paikan n]imi oli? ((looks straight ahead

at the

what was what was the name of the place

39 mirror))

40 MT: [° Heikki.°] ((tries to make eye

41 contact with Heikki, points at Rikhard, smiles; FT looks at

42 Heikki))
43 (.)

44 Rikhard: ↑TEuva. ((leans forwards))

45 (0.4)

46 Heikki: mil↑lasta ↓siel on.

what is is like there

47 (0.4)

48 Rikhard: no se=om ↑Pohjam↓maalla sinne:, (.) ö kestää ajamiseen, (.)

142 6 Therapists’ Response Strategies and Teaching Orientations

well it is in Ostrobothnia to drive there it takes

49 joku:, (.) >s' o'< vähän niinkuj Jyväskylä ↑kaks,

something like it is a bit like Jyväskylä two

50 ((MT looks at Rikhard and sighs; FT looks at MT))

51 (0.4) ö puol ↓tuntii tai s'tten, (.) .hhh (0.6) ↑kaks tuntii ja

er half an hour or then two hours and

52 puol ↓tuntii, (.) j- se, h (.) ttai s'te kolme tun°tia°; (0.4)

a half an hour a- it h or then three hours

53 sama ku on Helsinki Jyv°äskylä°, ((leans backwards)) (2.0)

the same as Helsinki Jyväskylä

A detailed analysis of the extract can be found in Sect. 4.2.1. However, even
without going into detail, the example clearly illustrates the fact that persons with
ASD tend to have difficulties understanding implicit messages in interaction. The
fact that the implicit message is related to a feature of non-verbal interaction in the
self-monitoring of which the boy has particular challenges may make it more diffi-
cult for him to understand the hint. The mirror on the opposite side of the room
also clearly disturbs the boy’s concentration. In addition, the group situation may
make it more challenging for a participant who has problems with interaction to
concentrate on such things as implicit messages (Lindholm, 2012). Nevertheless,
implicit messages aiming at repairing problems of embodied interaction are partic-
ularly difficult for persons with ASD to understand (Wiklund & Stevanovic, 2018).
Where repair practices of mundane, symmetrical interaction have been developed to
resolve problems of comprehension created by linguistic features in particular, there
are no similar resources—that would be routine and not face-threatening for the
interlocutor—for repairing problems of embodied, non-verbal interaction. Even if
the therapist’s first question, ‘who do you ask’ (line 04), could have, in the best case,
acted here as a “neutral” repair mechanism, understanding such a hint would have
required a rather developed capacity to perceive contextual features that are relevant
to the interpretation of the turn. Most often, one cannot expect persons with ASD
to be able to perceive such features (Stevanovic & Lindholm, 2016). Thus, contrary
to the previous example (see Sect. 6.2.2), in this extract, the therapists finally had to
comment explicitly on the boy’s problems with interaction.

6.2.4 Prosodic Problems: Beyond Teaching Orientations?

This sub-chapter discusses the interactional problems created by prosody and the
therapists’ opportunities to orient to them pedagogically (Wiklund & Stevanovic,
2018). It is often said that the manner in which we speak conveys as much meaning
as the words we use (Couper-Kuhlen, 2000: 2). It is also well known that prosodic
6.2 Therapists’ Teaching Orientations in Group Therapy Sessions Involving … 143

features constitute a large part of this manner. As already mentioned (see Chap. 2),
in phonetics, prosody encompasses phenomena related to speech rhythm, melody,
stress, phrasing, and voice quality. These are related to duration (timing), pitch (voice
fundamental frequency), loudness (intensity), and speech rate (Crystal, 1969, 1980).
Prosodic phenomena are suprasegmental; that is, they typically involve units of at
least one syllable in length. In this context, prosody is understood in a broad sense,
and it also encompasses such features as breathing, creaky voice, nasal voice, and
whispering (Couper-Kuhlen, 2000).
As already mentioned in Chap. 2, it is well established that people with ASD tend
to have deviant prosodic features in their speech. Even though the speech of persons
with ASD is often described as monotonous (see, e.g., Nakai et al., 2014), the pitch
range of people with ASD may also be exceptionally wide (see, e.g., Baltaxe, 1984;
Fosnot & Jun, 1999; Diehl & Paul, 2013; Nadig & Shaw, 2012). The speech of persons
with ASD may also be overly fast, “jerky”, or loud, or it may be characterised by
large pitch excursions, a quiet voice, an inconsistent pause structure, deviant word
stress, and/or a creaky or nasal voice (Asperger, 1944; Baltaxe, 1984; Baltaxe &
Simmons, 1985, 1992; Fay & Schuler, 1980; Lehtinen 2010; Paul 1987; Paul et al.,
2005a, 2005b; Shriberg et al., 2001; Tager-Flusberg, 2000; Thorson et al., 2016).
Generally speaking, prosodic extremes are typical of persons with ASD (Wiklund &
Vainio, 2019). However, as already mentioned (see Sect. 2.3), prosodic features rarely
constitute the main causes of comprehension problems—even if trouble-source turns
most often carry certain atypical prosodic features (Wiklund, 2016). In the Finnish
data for this project (see Sect. 1.2.1), the most frequent prosodic feature occurring in
trouble-source turns is a creaky voice (35.3%). A quiet voice is also relatively frequent
(31.4%). Large pitch excursions concern almost one-fourth of the cases (23.5%).
Other features—that concern less than one-fifth of the trouble-source turns—consist
of stretched syllables (17.6%), jerky speech rhythm (15.7%), bouncing pitch (11.8%),
and accelerated speech (11.8%) (see Sect. 2.3).
As managing the prosodic features of speech is an essential part of interactional
skills, one could expect the group therapy of children with ASD to take these features
into account at least from time to time. This does not, however, seem to be the
case in the data of this project. In the light of understanding repair mechanisms of
mundane, symmetric interaction, this is not surprising at all (Wiklund & Stevanovic,
2018). Indeed, as already mentioned, our interactional system does not really provide
routine, “neutral” means for repairing interactional problems caused by atypical gaze
behaviour—even if, as the example showed, certain pointing hand gestures can be
used to direct the gaze of the interlocutor to a direction that would be consistent with
interactional norms. Concerning speech prosody, this type of subtle interactional
means does not really exist. Addressing the topic would require bringing it up on
an explicit, metalinguistic level. In addition, this type of topic would probably be
face-threatening for the person who has used atypical prosody in his/her speech.
It is also noteworthy that, from the point of view of the teaching orientation, there
is another, even more fundamental problem: people do not generally have means
(concepts and terms) to describe prosodic parameters (such as pitch, intensity, and
rhythm) very accurately. People are rarely aware of in what combinations these
144 6 Therapists’ Response Strategies and Teaching Orientations

parameters can be expected to occur in various sequential positions and phases of

speaking turns (Stevanovic & Kahri, 2011). Thus, prosodic problems can be also
more generally speaking—and not only in the group therapy sessions of children with
ASD—something that is beyond the teaching orientation (Wiklund & Stevanovic,
2018).

6.3 Conclusion

The first part of this chapter addressed therapists’ response strategies in the French
data of the project (see Sect. 1.2.2). The theme of the group session, involving four
French-speaking boys with ASD, was taunting. The five most frequent response
categories occurring in the session were analysed: approval, commenting on the
contents of a turn, addressing a question to the whole group, asking a question about
the contents of a turn, and rephrasing (Wiklund & Määttä, 2021). From a linguistic
perspective, some of the boys in the data show clear traces of disfluency, whereas
others speak quite fluently. Prominent phenomena in the therapists’ responses include
their working as a team and providing cues about the interactional goal of their
responses in the prosodic forms of discourse particles. Apart from the linguistic
issues, the children’s interaction does not present features that are typically asso-
ciated with ASD, such as a lack of smiling and eye contact, as well as difficulty
in maintaining the discourse topic. At the same time, note-taking and attention to
writing devices prevented the therapists from maintaining constant eye contact and
reciprocating smiles (Wiklund & Määttä, 2021).
Only two examples presented in this chapter (see Sect. 6.1)—namely the examples
concerning rephrasing and questions addressed to all—were related to definitions or
valid examples of the topic of the encounter. Overall, the large number of response
sequences (n = 277) in a 55 min encounter (i.e., almost five responses per minute)
indicates that responses must have many other functions in addition to prompting
and assessing functional definitions and providing examples of taunting. The high
number of approvals, the most frequent category (n = 57), is particularly telling in
this respect (Wiklund & Määttä, 2021).
The therapists form a solid team that progresses towards the learning outcomes
resolutely. They give the children either implicit or explicit feedback in their
responses, and most of the strategies that feature in the data seem to aim at encour-
aging them to participate in the conversation, to share their own experiences, to
elaborate on their turns, to see the situation from the perspective of other people, and
to express themselves clearly. Indeed, although only the category ‘approval’ indi-
cates that the therapists are entirely happy with the child’s input and do not want
more precise or more adequate definitions and examples, their responses contain no
examples of solely negative feedback. This indicates that the way in which responses
are formulated has a strong impact: the aim is to encourage the children to participate
actively in order to develop their interactional skills, rather than to discourage them.
The educational agenda is clearly present throughout the session and regulates the
6.3 Conclusion 145

therapists’ responses. Hence, all the interaction and all the examples and definitions
are induced, filtered, and eventually validated by the therapists: the children can only
speak through the therapists, who retain the vocalic power in the situation (Connor,
2000: 23; Wiklund & Määttä, 2021).
The large number of sequences in which the children could elaborate their own
previous turns when prompted to do so confirms previous findings (Kremer-Sadlik,
2001, 2004; Wiklund, 2012, 2016) indicating that children with ASD are able to
participate in question–answer adjacency-pair sequences in relatively competent
ways (see Chap. 5). The results of the prosodic analyses are also in line with Wiklund
(2012), according to which children with ASD are able to interpret implicit interac-
tional meanings of combinations of dialogue particles and prosodic features correctly
(see Chap. 5). Contrary to Maynard’s (2019) findings, these children do not display
overtly uncooperative or resistant attitudes (Wiklund & Määttä, 2021).
Concerning teaching orientations (Wiklund & Stevanovic, 2018), the observations
presented in this chapter were focused on the repair sequences that aimed at resolving
comprehension problems between the participants as well as on the professional
activity of the therapists and the specific challenges that they encountered while
striving to fulfil the institutional agenda in the Finnish data of the project, where two
groups of preadolescent boys with ASD (three + four participants) spoke with each
other and with their two therapists (see Sect. 1.2.1. The analyses showed how the
therapists strove to maintain intersubjectivity and to resolve comprehension problems
with the help of interactional cues. They also communicated understanding of general
norms of interaction that the children with ASD are expected to learn to manage
during the group therapy sessions. For example, the boys should learn to express
themselves clearly (Wiklund & Stevanovic, 2018).
The teaching orientation in relation to interactional skills is generally implicit,
and it is based on repair mechanisms of normal mundane, symmetric interaction
(Wiklund & Stevanovic, 2018). The advantage of the implicit teaching orientation
is that the boys have the opportunity to work on their interactional skills in prac-
tice (about learning by doing, see Dewey, 1963[1938]) without their interactional
problems being emphasised. It is however noteworthy that especially when the prob-
lems concern embodied interaction, the therapists’ teaching orientation is sometimes
explicit, and—in these data—teaching problems of speech prosody seem to elude
even the explicit teaching orientation. However, as the skills of embodied interaction
and self-monitoring of speech prosody are essential in interaction and in the manage-
ment of social relationships (e.g., Stevanovic & Kahri, 2011), it would be important
to develop means for teaching these skills in a way that is not face-threatening for
the participants. Various technology-assisted teaching methods could be particularly
fruitful in this regard (Ke et al., 2018).
The fact that the informants with ASD often understand what is being said liter-
ally and do not understand all implicit cues impairs the fluency of interaction. This
feature has often been seen as proof of the impairment of the ‘theory of mind’ of
persons with ASD (Baron-Cohen, 1995; Baron-Cohen et al., 2013; Flavell, 1999). In
addition, at least some of the problems of self-expression can be interpreted through
the same framework: in these cases, the essential feature is that persons with ASD
146 6 Therapists’ Response Strategies and Teaching Orientations

often have difficulties understanding what way of expressing oneself is suitable and
understandable for the interlocutor in the situation at hand. On the other hand, the
data also include extracts where the informants show that they can interpret at least
some implicit interactional cues and that they have a reflexive understanding of the
possible deficits of their self-expression. Thus, the manifestation of the impairment
of the ‘theory of mind’ is not systematic in these data, but some of the informants are
clearly able to make conclusions about the states of the mind of the other participants
at times. Each informant is an individual who has individual qualities and capacities,
and this should obviously be taken into account in the therapy of persons with ASD
(Wiklund & Stevanovic, 2018).

References

Asberg, J. (2010). Patterns of language and discourse comprehension skills in school-aged children
with autism spectrum disorders. Scandinavian Journal of Psychology, 51, 534–539.
Asperger, H. (1944). Die “Autistischen Psychopathen“ im Kindesalter. Archiv Für Psychiatrie Und
Nervenkrankheiten, 117, 76–136.
Baltaxe, C. (1984). The use of contrastive stress in normal, aphasic and autistic children. Journal
of Speech and Hearing Research, 27, 97–105.
Baltaxe, C., & Simmons, J. (1985). Prosodic development in normal and autistic children. In E.
Schopler & G. Mesibov (Eds.), Communication problems in autism (pp. 95–125). Plenum Press.
Baltaxe, C., & Simmons, J. (1992). A comparison of language issues in high-functioning autism and
related disorders with onset in children and adolescence. In: E. Schopler & G. Mesibov (Eds.),
High-functioning individuals with autism, (pp. 210–225). New York: Plenum Press.
Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind. MIT Press.
Baron-Cohen, S., O’Riordan, M., Stone, V., Jones, R., & Plaisted, K. (1999). Recognition of faux
pas by normally developing children and children with Asperger syndrome or high-functioning
autism. Journal of Autism and Developmental Disorders, 29(5), 407–418.
Baron-Cohen, S., Lombardo, M., & Tager-Flusberg, H. (Eds.). (2013). Understanding other minds:
Perspectives from developmental social neuroscience (3rd ed.). Oxford University Press.
Capps, L., Kehres, J., & Sigman, M. (1998). Conversational abilities among children with autism
and children with developmental delays. Autism, 2(4), 325–344.
Connor, S. (2000). Dumbstruck: A cultural history of ventriloquism. Oxford University Press.
Couper-Kuhlen, E. (1996). The prosody of repetition: On quoting and mimicry. In E. Couper-
Kuhlen & M. Selting (Eds.), Prosody in conversation: Interactional studies (pp. 366–405).
Cambridge University Press.
Couper-Kuhlen, E. (2000). Prosody. In J. Verschueren, J.-O. Östman, & C. Bulcaen (Eds.),
Handbook of pragmatics (pp. 1–19). John Benjamins Publishing Company.
Crystal, D. (1969). Prosodic systems and intonation in English. Cambridge University Press.
Crystal, D. (1980). A first dictionary of linguistics and phonetics. Deutsch.
Cummings, L. (2009). Clinical pragmatics. Cambridge University Press.
Cummings, L. (2014a). Pragmatic disorders. Springer.
Cummings, L. (2014b). Communication disorders. Palgrave Macmillan.
Dennis, M., Lazenby, A. L., & Lockyer, L. (2001). Inferential language in high-function children
with autism. Journal of Autism and Developmental Disorders, 31, 47–54.
Dewey, J. (1963[1938]). Experience and education. Collier-Macmillan Books.
Diehl, J., & Paul, R. (2013). Acoustic and perceptual measurements of prosody production on the
profiling elements of prosodic systems in children by children with autism spectrum disorders.
Applied Psycholinguistics, 34(1), 135–161.
References 147

Dindar, K., Korkiakangas, T., Laitila, A., & Kärnä, E. (2016). Building mutual understanding: How
Children with autism spectrum disorder manage interactional trouble. Journal of Interactional
Research in Communication Disorders, 7(1), 49–77.
Eales, M. J. (1993). Pragmatic impairments in adults with childhood diagnosis of autism or devel-
opmental receptive language disorders. Journal of Autism and Developmental Disorders, 23(4),
598–617.
Fay, W., & Schuler, A. L. (1980). Emerging language in autistic children. University Park Press.
Flavell, J. H. (1999). Cognitive development: Children’s knowledge about the mind. Annual Review
of Psychology, 50, 21–45.
Fosnot, S. M., & Jun, S.-A. (1999). Prosodic characteristics in children with stut-
teringor autism during reading and imitation. In J.J. Ohala, Y. Hasegawa, M. Ohala, D.
Granville & A.C. Bailey (Eds.), Proceedings of the 14th International Congress of Phonetic
Sciences (ICPhS), https://www.internationalphoneticassociation.org/icphs-proceedings/ICPhS1
999/papers/p14_1925.pdf.
Geller, E. (1998). An investigation of communication breakdowns and repairs in verbal autistic
children. The British Journal of Development Disabilities, 44(87), 71–85.
Gena, A., Couloura, S., & Kymissis, E. (2005). Modifying the affective behavior of preschoolers
with autism using in-vivo or video modeling and reinforcement contingencies. Journal of Autism
and Developmental Disorders, 35(5), 545–556.
Gill, V., & Maynard, D. W. (2006). Explaining illness: Patients’ proposals and physicians’ responses.
In J. Heritage & D. W. Maynard (Eds.), Communication in medical care: Interaction between
primary care physicians and patients (pp. 115–150). Cambridge University Press.
Hakulinen, A., Vilkuna, M., Korhonen, R., Koivisto, V., Heinonen, T. R., & Alho, I. (2004). Iso
suomen kielioppi. SKS.
Hale, C. M., & Tager-Flusberg, H. (2005a). Brief report: The relationship between discourse deficits
and autism symptomatology. Journal of Autism and Developmental Disorders, 35, 519–524.
Hale, C. M., & Tager-Flusberg, H. (2005b). Social communication in children with autism: The
relationship between theory of mind and discourse development. Autism, 9(2), 157–178.
Happé, F. G. E. (1994). Autism: An introduction to psychological theory. Harvard University Press.
Hutchby, I., & Wooffitt, R. (2008). Conversation analysis (2nd ed.). Polity Press.
Hutt, C., & Ounsted, C. (1966). The biological significance of gaze aversion with particular reference
to the syndrome of infantile autism. Behavioral Science, 11(5), 346–356.
Jones, R. H. (2015). Discourse and health communication. In D. Tannen, H. E. Hamilton, & D.
Schiffrin (Eds.), The handbook of discourse analysis (pp. 841–857). Wiley Blackwell.
Ke, F., Whalon, K., & Yun, J. (2018). Social skill interventions for youth and adults with autism
spectrum disorder: A systematic review. Review of Educational Research, 88(1), 3–42.
Kleinman, J., Marciano, P. L., & Ault, R. L. (2001). Advanced theory of mind in high-functioning
adults with Autism. Journal of Autism and Developmental Disorders, 31(1), 29–36.
Koch, S. C., Mehl, L., Sobanski, E., Sieber, M., & Fuchs, T. (2015). Fixing the mirrors: A feasibility
study of the effects of dance movement therapy on young adults with autism spectrum disorder.
Autism, 19(3), 338–350.
Korkiakangas, T. K., Rae, J. P., & Dickerson, P. (2012). The interactional work of repeated talk
between a teacher and a child with autism. Journal of Interaction Research in Communication
Disorders, 3(1), 1–25.
Korpilahti, P., Jansson-Verkasalo, E., Mattila, M.-L., Kuusikko, S., Suominen, K., Rytky, S., Pauls,
D. L., & Moilanen, I. (2007). Processing of affective speech prosody is impaired in Asperger
syndrome. Journal of Autism and Developmental Disorders, 37(8), 1539–1549.
Koskinen, E. (2015). Mielen luenta keskustelussa. [Master’s thesis]. University of Helsinki.
Kremer-Sadlik, T. (2001). How children with autism and Asperger syndrome respond to questions:
An ethnographic study. [Doctoral dissertation]. University of California.
Kremer-Sadlik, T. (2004). How children with Autism and Asperger Syndrome respond to questions:
A ‘naturalistic’ theory of mind task. Discourse Studies, 6(2), 185–205.
148 6 Therapists’ Response Strategies and Teaching Orientations

Krstovska-Guerrero, I., & Jones, E. A. (2013). Joint attention in autism: Teaching smiling coordi-
nated with gaze to respond to joint attention bids. Research in Autism Spectrum Disorder, 7(1),
93–108.
Kurhila, S., & Lindholm, C. (2016). Ymmärtämisen haasteet. In: M. Stevanovic & C. Lindholm
(toim.), Keskustelunanalyysi. kuinka tutkia sosiaalista toimintaa ja vuorovaikutusta (pp. 259–
275). Tampere: Vastapaino.
Kylliäinen, A. (2007). Face and gaze processing in children with autism. [Doctoral dissertation].
University of Tampere.
Lee, S.-H. (2013). Response design in conversation. In J. Sidnell & T. Stivers (Eds.), The handbook
of conversation analysis (pp. 415–432). Wiley Blackwell.
Lehtinen, M. (2010). Observations on the prosodic characteristics of Finnish-speaking youngsters
with Asperger syndrome. In A. Botinis (Ed.), Proceedings of ‘ExLing 2010—ISCA Tutorial and
Research Workshop on Experimental Linguistics’ (pp. 93–96). ISCA/University of Athens.
Lehtinen, M. (2012). “Aiheesta toiseen”—Ymmärrysongelmatilanteita Asperger-nuorten kuntou-
tuskeskusteluissa. In L. Leskelä & C. Lindholm (Eds.), Haavoittuva keskustelu. Keskustelu-
nanalyyttisia tutkimuksia kielellisesti epäsymmetrisestä vuorovaikutuksesta (pp. 235–277).
Kehitysvammaliitto.
Lewis, F. M., Woodyatt, G. C., & Murdoch, B. E. (2008). Linguistic and pragmatic language skills
in adults with autism spectrum disorder: A pilot study. Research in Autism Spectrum Disorders,
2(1), 176–187.
Lindholm, C. (2012). Vuorovaikutuksen haasteita ja mahdollisuuksia—Tapaustutkimus muis-
tisairaiden päivätoiminnasta. In L. Leskelä & C. Lindholm (Eds.), Haavoittuva keskustelu.
Keskustelunanalyyttisiä tutkimuksia kielellisesti epäsymmetrisestä vuorovaikutuksesta (pp. 57–
98). Kehitysvammaliitto.
Losh, M., & Capps, L. (2003). Narrative ability in high-functioning children with autism or
Asperger’s syndrome. Journal of Autism and Developmental Disorders, 33, 239–251.
Loukusa, S., Leinonen, E., Kuusikko, S., Jussila, K., Mattila, M.-L., Ryder, N., Ebeling, H., &
Moilanen, I. (2007). Use of context in pragmatic language comprehension by children with
Asperger syndrome or high-functioning autism. Journal of Autism and Developmental Disorders,
37(6), 1049–1059.
Martin, I., & McDonald, S. (2004). An exploration of causes of non-literal language problems in
individuals with Asperger syndrome. Journal of Autism and Developmental Disorders, 34(3),
311–328.
Maynard, D. W. (2019). Why social psychology needs autism and why autism needs social
psychology: Forensic and clinical considerations. Social Psychology Quarterly, 82(1), 5–30.
McPartland, J., & Klin, A. (2006). Asperger’s syndrome. Adolescent Medicine Clinics, 17(3),
771–788.
Mundy, P., Sigman, M., & Kasari, C. (1990). A longitudinal study of joint attention and language
development in autistic children. Journal of Autism and Developmental Disorders, 20, 115–128.
Nadig, A., & Shaw, H. (2012). Acoustic and perceptual measurement of expressive prosody in
high-functioning autism: Increased pitch range and what it means to listeners. Journal of Autism
and Developmental Disorders, 42(4), 499–511.
Nakai, Y., Takashima, R., Takiguchi, T., & Takada, S. (2014). Speech intonation in children with
autism spectrum disorder. Brain and Development, 36(6), 516–522.
Nieminen-von Wendt, T. (2004). On the origins and diagnosis of Asperger syndrome. A clinical,
neuroimaging and genetic study. [Doctoral dissertation]. University of Helsinki.
Ochs, E., & Solomon, O. (2004). Introduction: Discourse and autism. Discourse Studies, 6(2),
139–146.
Ohtake, Y., Wehmeyer, M., Nakaya, A., Takahashi, S., & Yanagihara, M. (2011). Repair strategies
used by verbal students with autism during free play. Focus on Autism and Other Developmental
Disabilities, 26(1), 3–14.
Paul, R. (1987). Communication. In: D. Cohen & Anne Donnellan (Eds.), Handbook of autism and
pervasive developmental disorders, (pp. 61–84). Hoboken: Wiley.
References 149

Paul, R., Augustyn, A., Klin, A., & Volkmar, F. R. (2005a). Perception and Production of Prosody
by Speakers with Autism Spectrum Disorders. Journal of Autism and Developmental Disorders,
35(2), 205–220.
Paul, R., Shriberg, L., Mcsweeney, J., Cicchetti, D., Klin, A., & Volkmar, F. R. (2005b). Brief
report: relations between prosodic performance and communication and socialization ratings in
high functioning speakers with autism spectrum disorders. Journal of Autism and Developmental
Disorders, 35(6), 861–869.
Pomerantz, A. (1984). Pursuing a response. In J. M. Atkinson & J. Heritage (Eds.), Structures of
social action: Studies in conversation analysis (pp. 152–164). Cambridge University Press.
Rendle-Short, J. (2017). Conversation analysis: A tool for analysing interactional difficulties faced
by children with Asperger’s syndrome. In M. O’Reilly, J. Lester, & T. Muskett (Eds.), A prac-
tical guide to social interaction research in autism spectrum disorders (pp. 297–323). Palgrave
Macmillan.
Routarinne, S. (2003). Tytöt äänessä. Parenteesit ja nouseva sävelkulku kertojan vuorovaiku-
tuskeinoina. SKS.
Sacks, H. (1992). Lectures on conversation (Vol. 1 & 2). Blackwell.
Shriberg, L., Paul, R., McSweeney, J., Klin, A., Cohen, D., & Volkmar, F. R. (2001). Speech and
prosody characteristics of adolescents and adults with high functioning autism and Asperger
syndrome. Journal of Speech, Language and Hearing Research, 44, 1097–1115.
Siegel, B. (1996). The world of the autistic child: Understanding and treating autistic spectrum
disorders. Oxford University Press.
Solomon, O. (2001). Narrative Introduction Practices of Children with Autism and Asperger
Syndrome. [PhD dissertation]. Los Angeles: University of California.
Solomon, O. (2004). Narrative introductions: Discourse competence of children with Autistic
spectrum Disorders. Discourse Studies, 6(2), 253–276.
Stevanovic, M., Henttonen, P., Koski, S., Kahri, M., Voutilainen, L., Koskinen, E., Nieminen- von
Wendt, T., Tani, P., Sihvola, E., & Peräkylä, A. (2017). On the Asperger experience of interaction:
Interpersonal dynamics in dyadic conversations. Journal of Autism, 4(2).
Stevanovic, M., & Kahri, M. (2011). Puheäänen musiikilliset piirteet ja sosiaalinen toiminta: Kohti
prosodian sosiologiaa. Sosiologia, 48(4), 1–24.
Stevanovic, M., & Lindholm, C. (2016). Sosiaalisen havaitsemisen haasteet. In M. Stevanovic &
C. Lindholm (Eds.), Keskustelunanalyysi. Kuinka tutkia sosiaalista toimintaa ja vuorovaikutusta
(pp. 276–294). Vastapaino.
Szatmari, P., Brenner, R., & Nagy, J. (1989). Asperger syndrome: A review of clinical features. The
Canadian Journal of Psychiatry, 34, 554–560.
Tager-Flusberg, H. (1993). What language reveals about the understanding of minds in children
with Autism. In S. Baron-Cohen, H. Tager-Flusberg, & D. J. Cohen (Eds.), Understanding Other
Minds: Perspectives from Autism (pp. 138–157). Oxford University Press.
Tager-Flusberg, H. (2000). Understanding the language and communicative impairments in autism.
International Review of Research on Mental Retardation, 23, 185–205.
Tager-Flusberg, H. (2001). Understanding the language and communication impairments in autism.
International Review of Research in Mental Retardation, 23, 185–205.
Tantam, D., Holmes, D., & Cordess, C. (1993). Non-verbal expression in autism of Asperger type.
Journal of Autism and Developmental Disorders, 23(1), 111–133.
Thorson, J. C., Usher, N., Patel, R., & Tager-Flusberg, H. (2016). Acoustic analysis of prosody in
spontaneous productions of minimally verbal children and adolescents with autism. In Barnes, J.,
Brugos, A., Shattuck-Hufnagel, S., & Veilleux, N. (Eds.), Proceedings of Speech Prosody (Vol.
8).
Vehviläinen, S. (2014). Ohjaustyön opas: Yhteistyössä kohti toimijuutta. Gaudeamus.
Volden, J. (2004). Conversational repair in speakers with autism spectrum disorder. International
Journal of Language & Communication Disorders, 39(2), 171–189.
Wiklund, M. (2012). Gaze behavior of pre-adolescent children afflicted with Asperger syndrome.
Communication & Medicine, 9(2), 173–186.
150 6 Therapists’ Response Strategies and Teaching Orientations

Wiklund, M. (2016). Interactional challenges in conversations with autistic preadolescents: The

role of prosody and nonverbal communication in other-initiated repairs. Journal of Pragmatics,
94, 76–97.
Wiklund, M., & Laakso, M. (2020). Comparison of disfluent and ungrammatical speech of pread-
olescents with and without ASD. Journal of Autism and Developmental Disorders, 51(8),
2773–2789.
Wiklund, M., & Määttä, S. (2021). Therapists’ response strategies in a group session involving
French-speaking children with autism spectrum disorder. Journal of Interdisciplinary Voice
Studies, 6(1), 109–130.
Wiklund, M., & Stevanovic, M. (2018). Ymmärrysongelmat ja vuorovaikutustaitojen opet-
taminen lievästi autististen varhaisnuorten kuntoutuskeskusteluissa [Understanding problems
and the teaching of social interaction skills in group therapy sessions of preadolescents with
high-functioning autism]. Psykologia, 53(05–06), 421–451.
Wiklund, M., & Vainio, M. (2019). Pitch-related features in the speech of Finnish- and French-
speaking boys with autism in data coming from group therapy sessions. In H. E. H. Lenk, J.
Härmä, B. Sanromán Vilas, & E. Suomela-Härmä (Eds.), Studies in Comparative Pragmatics
(pp. 45–63). Cambridge Scholars Publishing.
List of Transcription Conventions

CLI Clitic
PER Person
PL Plural
PRT Particle
SG Singular
. Strongly falling pitch at the end of a prosodic unit
; Slightly falling pitch at the end of a prosodic unit
, Flat pitch at the end of a prosodic unit
,? Slightly rising pitch at the end of a prosodic unit
? Strongly rising pitch at the end of a prosodic unit
↓ Segment produced on a lower pitch level than the surrounding speech
↑ Segment produced on a higher pitch level than the surrounding speech
sika Prominent stress
> tosi < Accelerated speech rate
< paitsi > Slowed speech rate
joo: Lengthened vowel
MITÄ Increased level of loudness
.hhh Clearly audible inhalation (one ‘h’ corresponds to 0.1 s)
hhh. Clearly audible exhalation (one ‘h’ corresponds to 0.1 s)
.joo Word produced with an inhalation
@just@ Marked voice
k(h)iva Word produced laughingly
£niimpä£ Word produced with a smile
°nii° Word produced more quietly than the surrounding speech
[ Overlap of speech begins
] Overlap of speech ends
(.) Micropause (duration of less than 0.2 s)
(0.6) Pause (duration measured in seconds)
(lapset) Unclear speech
= Turn starts immediately after the end of the previous turn

© The Editor(s) (if applicable) and The Author(s), under exclusive license 151
to Springer Nature Singapore Pte Ltd. 2023
M. Wiklund, Speech and Interaction of Preadolescents with Autism Spectrum Disorder,
Prosody, Phonology and Phonetics, https://doi.org/10.1007/978-981-19-8117-3
Conclusion

This book has explored the speech and the interaction of preadolescents with ASD
from various points of view. Firstly, a central aspect of this book has been speech
prosody (Chap. 2). Concerning prosody, the studies presented in the book focused,
on the one hand, on the perception of the atypicality of speech prosody by neurotyp-
ical (i.e., non-autistic) persons. It was shown, for example, that neurotypical adults
find the prosody of preadolescent boys with ASD more atypical than the prosody
of age- and gender-matched neurotypical controls. The difference is statistically
significant (Wiklund et al., forthcoming). On the other hand, the analyses showed
which prosodic features are more specifically considered atypical by neurotypical
persons. These include, for example, such features as sing-song or bouncing pitch,
disconnected speech rhythm, large pitch excursions, or flatness of pitch (Wiklund &
Vainio, 2019; Wiklund et al., forthcoming). Interestingly, the speech of the infor-
mants with ASD is also frequently perceived as having been produced by a non-
native speaker of the language (Wiklund et al., forthcoming). Creaky voice is also
more common in the ASD group than in the control group (Lohi, 2020). Generally
speaking, prosodic extremes seem to characterise the speech of preadolescent boys
with ASD (Wiklund & Vainio, 2019). The role of prosodic features in the creation
of comprehension problems was also discussed. The analyses showed that the most
frequent prosodic feature occurring in trouble-source turns (i.e., speaking turns that
cause problems of comprehension) is a creaky voice (Wiklund, 2016). However,
prosodic features are rarely the main cause of the comprehension problems that
occur in spontaneous interaction between preadolescents with ASD and their ther-
apists (Wiklund & Stevanovic, 2018; Wiklund, 2016). At the end of Chap. 2, the
use of prosody by preadolescent boys with ASD in questions and in statements was
addressed (Wiklund et al., 2021). The analyses suggested that preadolescent boys
with ASD are able to use the utterance-final rising pitch as an interactional resource.
They are also able to produce and correctly interpret prosodic features that indicate
finality in spontaneous interaction as well as to emphasis words with the help of
prosodic features, call for other participants’ reactions, and indicate that they take

© The Editor(s) (if applicable) and The Author(s), under exclusive license 153
to Springer Nature Singapore Pte Ltd. 2023
M. Wiklund, Speech and Interaction of Preadolescents with Autism Spectrum Disorder,
Prosody, Phonology and Phonetics, https://doi.org/10.1007/978-981-19-8117-3
154 Conclusion

other participants into account with the help of utterance-final pitch rises (Wiklund
et al., 2021). Thus, preadolescent boys with ASD seem to be able to use intonation as
an interactional resource, even though reciprocal interaction is, generally speaking,
difficult for people with ASD (APA, 2013).
Secondly, this book focused on disfluencies of speech (Chap. 3). According to
the studies presented here, disfluencies are more frequent in the speech of pread-
olescent boys with ASD than in the speech of age- and gender-matched controls
(Wiklund & Laakso, 2020). Furthermore, a qualitative difference can be noted: the
ASD group produces long, complex disfluent turns with word searches, self-repairs,
false starts, fillers, prolongations, inconsistent syntactic structures, and grammat-
ical errors, whereas in the control group, the disfluencies are mainly fillers and
sound prolongations (Wiklund & Laakso, 2020). The disfluencies and ungrammatical
expressions that occur in the ASD participants’ interactions also cause comprehen-
sion problems, whereas comprehension problems do not occur in the control group
(Wiklund & Laakso, 2019).
Thirdly, this book addressed comprehension problems occurring in spontaneous
interaction between preadolescent boys with ASD and their neurotypical therapists
(Chap. 4). The studies presented in this book showed that certain prosodic and non-
verbal features are often associated with trouble-source turns that cause comprehen-
sion problems between preadolescent boys with ASD and their neurotypical ther-
apists (Wiklund, 2016). For example, in 84% of the cases, there is no eye contact
between the speaker producing a trouble-source turn and the person who initiates
the conversational repair sequence. Sometimes the lack of eye contact is associated
with overlapping speech. Concerning prosody, the most frequent feature occurring
in trouble-source turns produced by the boys with ASD is a creaky voice; it occurs
in 35% of the trouble-source turns of the data. A quiet voice, large pitch excursions,
stretched syllables, and jerky speech rhythms are examples of other prosodic features
that appeared in the trouble-source turns of the data (Wiklund, 2016). The results
of the studies presented in this book demonstrate that the tendency of persons with
ASD to avoid direct eye contact as well as the occurrences of certain deviant prosodic
features in their speech are factors that affect the fluidity of interaction and are related
to the creation of comprehension problems. However, only in a very few cases do
non-verbal and prosodic features seem to be the main cause of the comprehension
problem (Wiklund, 2016). The two most common causes of comprehension prob-
lems in these data are overly literal interpretation of speech and topical discontinuities
(Lehtinen, 2012; Wiklund & Stevanovic, 2018; Wiklund, 2016).
One chapter of the book focused on the gaze behaviour of preadolescent boys with
ASD in naturally occurring group therapy sessions (Chap. 5). The analyses showed
that preadolescent boys with ASD are able to interpret the conversational func-
tions of combinations of dialogue particles, prosodic features, and gestures correctly
(Wiklund, 2012; Wiklund & Määttä, 2021; Wiklund et al., 2021). Three main patterns
are used to avoid eye contact when speaking: (1) fixing one’s gaze straight ahead,
(2) letting one’s gaze wander around, and (3) looking at one’s own hands (Wiklund,
2012). According to the results, the boys with ASD generally do not look at the inter-
locutors at all at the beginning and in the middle of their speaking turns. Interestingly,
Conclusion 155

sometimes the boys turn to look at the interlocutors at the end of their speaking turn,
which proves that these children are able to use gaze as a source of feedback. It is
also noteworthy that looking at the speaker when listening seems to be easier for
the preadolescents with ASD than looking at the listeners when speaking (Wiklund,
2012).
The book also dealt with therapists’ teaching orientations and response strategies
in naturally occurring group therapy sessions involving preadolescent boys with ASD
(Chap. 6). The analyses showed that on the one hand, the therapists solve problems
of intersubjectivity, but on the other hand, they also simultaneously communicate an
understanding of the general norms of social interaction that the boys are expected
to learn to manage in the sessions (Wiklund & Stevanovic, 2018). Thus, the ther-
apists’ teaching orientation is mainly implicit, and it is based on the normal repair
practices associated with mundane symmetrical interaction. However, when teaching
embodied social interaction skills, the explication of the teaching orientation is some-
times necessary. Interestingly, the teaching of prosodic skills tends to elude even
the explicit teaching orientation in group therapy sessions (Wiklund & Stevanovic,
2018). When the therapists’ response strategies are analysed, different strategies can
be found (Wiklund & Määttä, 2021). Although a specific response category may
have various functions depending on the context, the analyses show that the aim of
all the strategies is clearly to maintain intersubjectivity among the participants (the
boys with ASD and their therapists). This aim is visible, for example, in the absence
of overtly negative feedback in the data. In addition, prosodic features give strong
clues concerning the interactional meanings carried by the response particles used
by the therapists (Wiklund & Määttä, 2021).
References

American Psychiatric Association (APA). (2013). Diagnostic and statistical manual of mental
disorders, (DSM-V) (5th edn.). American Psychiatric Association.
Lehtinen, M. (2012). “Aiheesta toiseen”—Ymmärrysongelmatilanteita Asperger-nuorten kuntou-
tuskeskusteluissa. In L. Leskelä & C. Lindholm (Eds.), Haavoittuva keskustelu. Keskustelu-
nanalyyttisia tutkimuksia kielellisesti epäsymmetrisestä vuorovaikutuksesta (pp. 235–277).
Kehitysvammaliitto.
Lohi, N. (2020). Prosody of adolescent Finnish boys diagnosed with ASD: Variation within the
group and comparison with neurotypically developed adolescents. [Master’s thesis]. University
of Helsinki.
Wiklund, M. (2012). Gaze behavior of pre-adolescent children afflicted with Asperger syndrome.
Communication & Medicine, 9(2), 173–186.
Wiklund, M. (2016). Interactional challenges in conversations with autistic preadolescents: The
role of prosody and nonverbal communication in other-initiated repairs. Journal of Pragmatics,
94, 76–97.
Wiklund, M., Ihaksinen, K., & Vainio, M. (2021). Autismikirjon poikien intonaation käyttö
kysymys-ja kerrontavuoroissa spontaanissa vuorovaikutuksessa [The Use of Intonation in Ques-
tions and Narrative Turns in the Spontaneous Interaction of Boys Afflicted with Autism Spectrum
Disorder]. Puhe Ja Kieli, 41(1), 43–70.
Wiklund, M., & Laakso, M. (2019). Ungrammatical utterances and disfluent speech as causes
of comprehension problems in interactions of preadolescents with highly functioning autism.
Clinical Linguistics & Phonetics, 33(7), 654–676.
Wiklund, M., & Laakso, M. (2020). Comparison of disfluent and ungrammatical speech of pread-
olescents with and without ASD. Journal of Autism and Developmental Disorders, 51(8),
2773–2789.
Wiklund, M., & Määttä, S. (2021). Therapists’ response strategies in a group session involving
French-speaking children with autism spectrum disorder. Journal of Interdisciplinary Voice
Studies, 6(1), 109–130.
Wiklund, M., & Stevanovic, M. (2018). Ymmärrysongelmat ja vuorovaikutustaitojen opet-
taminen lievästi autististen varhaisnuorten kuntoutuskeskusteluissa [Understanding problems
and the teaching of social interaction skills in group therapy sessions of preadolescents with
high-functioning autism]. Psykologia, 53(05–06), 421–451.
Wiklund, M., & Vainio, M. (2019). Pitch-related features in the speech of Finnish- and French-
speaking boys with autism in data coming from group therapy sessions. In H. E. H. Lenk, J.
Härmä, B. Sanromán Vilas, & E. Suomela-Härmä (Eds.), Studies in comparative pragmatics
(pp. 45–63). Cambridge Scholars Publishing.

© The Editor(s) (if applicable) and The Author(s), under exclusive license 157
to Springer Nature Singapore Pte Ltd. 2023
M. Wiklund, Speech and Interaction of Preadolescents with Autism Spectrum Disorder,
Prosody, Phonology and Phonetics, https://doi.org/10.1007/978-981-19-8117-3
158 References

Wiklund, M., Vainio, L., Saalasti, S., & Vainio, M. (forthcoming). Puheen prosodian havaittu
epätyypillisyys suomenkielisillä autismikirjon varhaisnuorilla [Perceived Atypicality of Speech
Prosody of Finnish-Speaking Preadolescents with Autism Spectrum Disorder]. Puhe ja kieli.