Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide 1635500273, 9781635500271

Citation preview

Phonological Treatment of Speech Sound Disorders in Children A Practical Guide

Phonological Treatment of Speech Sound Disorders in Children A Practical Guide

Jacqueline Bauman-Waengler, PhD, CCC-SLP Diane Garcia, MS, CCC-SLP

5521 Ruffin Road San Diego, CA 92123 e-mail: [email protected] Website: http://www.pluralpublishing.com

Copyright © 2020 by Plural Publishing, Inc. Typeset in 11/14 Stone Informal by Flanagan’s Publishing Services, Inc. Printed in the United States of America by McNaughton & Gunn, Inc. All rights, including that of translation, reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, recording, or otherwise, including photocopying, recording, taping, Web distribution, or information storage and retrieval systems without the prior written consent of the publisher. For permission to use material from this text, contact us by Telephone:  (866) 758-7251 Fax:  (888) 758-7255 e-mail: [email protected] Every attempt has been made to contact the copyright holders for material originally printed in another source. If any have been inadvertently overlooked, the publishers will gladly make the necessary arrangements at the first opportunity. Disclaimer: Please note that ancillary content (such as documents, audio, and video, etc.) may not be included as published in the original print version of this book. Library of Congress Cataloging-in-Publication Data Names: Bauman-Wèangler, Jacqueline Ann, author. | Garcia, Diane, 1967- author. Title: Phonological treatment of speech sound disorders in children : a practical guide / Jacqueline Bauman-Waengler, Diane Garcia. Description: San Diego, CA : Plural Publishing, [2018] | Includes bibliographical references and index. Identifiers: LCCN 2018005895| ISBN 9781635500271 (alk. paper) | ISBN 1635500273 (alk. paper) Subjects: | MESH: Speech Sound Disorder — therapy | Speech Therapy — methods | Child | Adolescent | Case Reports Classification: LCC RJ496.S7 | NLM WL 340.3 | DDC 618.92/85506 — dc23 LC record available at https://lccn.loc.gov/2018005895

Contents

Preface ix Acknowledgments xi Reviewers xiii



1 Introduction to Phonological Treatment of

1

Speech Sound Disorders

Jacqueline Bauman-Waengler and Diane Garcia Importance of Phonological Treatment Methods 1 How Is Speech Sound Disorder Defined? 2 Prevalence and Impact of Speech Sound Disorders 4 Principles of Phonological Treatment 5 Workbook Content and Layout 7 Summary of Treatment Approaches 7 References 11



2 Ways to Collect and Analyze Data

13

3 Minimal Pair Therapy

65

Jacqueline Bauman-Waengler Overview 13 Phonetic Transcription and Diacritics 13 Independent Versus Relational Analyses 17 Using a Standardized Speech Assessment for Your Data Collection 35 Analyzing Errors 36 Measuring Intelligibility 47 References 50



Diane Garcia Overview 65 Research Supporting Minimal Pair Therapy 70 Initial Data Collection and Selecting Targets 73 Establishing Treatment Goals 77 Beginning Therapy 79 Group Therapy 82 Case Study 84 References 86

v

vi

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

4 Multiple Oppositions Intervention

89

Diane Garcia Overview 89 Research Supporting Multiple Oppositions Intervention 93 Initial Data Collection and Selecting Targets 96 107 Beginning Therapy Group Therapy 111 114 Case Study References 122



5 Maximal Opposition Therapy

125

6 Complexity Approaches

159

7 Phonotactic Therapy

189

8 Core Vocabulary Approach

213

Diane Garcia Overview 125 Research Supporting Maximal Opposition Therapy 131 Initial Data Collection and Selecting Targets 135 Beginning Therapy 141 Group Therapy 144 Case Study 145 References 148



Jacqueline Bauman-Waengler Overview 159 Research Supporting Speech Sound Disorders and the 165 165 Complexity Approach Initial Data Collection and Selecting Targets 169 Beginning Therapy:  Targeting Singletons 175 Beginning Therapy:  Consonant Clusters 177 Establishing Treatment Goals 178 Group Therapy 180 Case Study 181 References 184



Diane Garcia Overview 189 Research Supporting Phonotactic Therapy 194 Initial Data Collection and Selecting Targets 196 Beginning Therapy 203 Case Study 205 References 211



Jacqueline Bauman-Waengler Overview 213

Contents

vii

Research Supporting Speech Sound Disorders and Core Vocabulary 214 Intervention Initial Data Collection and Selecting Targets 216 Beginning Therapy:  Establishing and Working on Core Vocabulary 217 223 Group Therapy Establishing Treatment Goals 223 Case Study 224 References 227



9 Cycles Phonological Remediation Approach

231

Jacqueline Bauman-Waengler Overview 231 233 Research Supporting the Cycles Approach Initial Data Collection and Selecting Targets 235 Beginning Therapy:  Overview and Establishing Cycles 241 Establishing Treatment Goals 248 Group Therapy 250 Case Study 251 References 254

10 Phonological/Phonemic Awareness and the Treatment of Children with Speech Sound Disorders

257

Jacqueline Bauman-Waengler Overview 257 Research Supporting Speech Sound Disorders and 259 Phonological/Phonemic Awareness Therapy Initial Data Collection and Selecting Targets 265 Beginning Therapy:  Combining Work on Phonological Awareness 270 with Treatment of Speech Sound Disorders Establishing Treatment Goals 280 Case Study 282 References 284

11 Case Studies

287

Jacqueline Bauman-Waengler and Diane Garcia Overview 287 Case Study #1:  Elias 287 Case Study #2:  Cameron 292 Case Study #3:  Maddie 299 Case Study #4:  Kelly 304 References 311

Index 313

Preface

The conception of this workbook grew out of a need to provide practicing clinicians with information about phonological treatment possibilities that could be easily implemented in their day-to-day workplace. It appeared that clinicians were still largely treating children with speech sound disorders within the framework of the traditional sound-by-sound approach. Although this served speech-language therapists well for decades, more recent developments have documented that several treatment protocols may trigger changes to a child’s entire phonological system, bypassing the treatment of individual sounds. Some of these phonological therapies are not new, and some had evolved from one type of target selection process to another as more research was accumulated. In the end, there did not seem to be a resource that compiled very practical information on how to implement these various procedures in a usable format. Our goal is to provide a workbook that will be useful for speech-language therapists who work with children with speech sound disorders. Specifically, this workbook targets selected treatment methods that could be suitable for the many children who have difficulties with the form (phone) and function (phoneme) of speech sounds. There are many workbooks and textbooks that give information about the sound-by-sound approach, the so-called traditional approach. However, this book attempts to demonstrate treatment principles, which are broadly labelled phonological, in a systematic manner that clinicians can easily follow. It is unique in that it provides easy-to-follow guidelines that allow clinicians to systematically move from assessment results to treatment planning and implementation. Although there are dozens of methods available, our selection was based on those approaches that were frequently noted in publications and, in most cases, those which have research results supporting them. Two practical principles guided our efforts: the time constraints that working clinicians are faced with, and the necessity to demonstrate the treatment guidelines in an easy manner. With large caseloads, and even larger amounts of paperwork, working clinicians are constantly faced with their limitations. And although “newer” phonological treatment protocols are available, there is often not enough time to see how they could be easily and efficiently implemented. This workbook is an attempt to streamline diagnostic and therapeutic possibilities. To that end, data collection forms are provided and case studies accompany all of the chapters. Clinicians should be able to easily organize the information from an articulation test or a speech sample, while demonstrations via case studies guide the clinicians through the necessary steps to apply their data directly to therapy. To summarize, this workbook provides methods that have proven successful in treating children with speech sound disorders. It is very practically oriented so that readers

ix

x

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

can easily see the progression from data that has been collected, to treatment goals and outcome measures. In addition, diagnostic and therapy materials that can be directly used are provided within each chapter. Finally, it demonstrates the use of these techniques when applicable within a group therapy setting, which is the main mode of delivery for most clinicians and the least discussed in typical workbooks.

Ac k n o w l e d g m e n t s

It takes a village and this project was supported by many people. First a special thank you to Valerie Johns, Executive Editor, and Kalie Koscielak, Project Editor at Plural Publishing. Valerie believed in this project from the beginning while Kalie was so patient and supportive through all its stages of development. Thank you. And to Jessica Bristow, Production Assistant, for the great front cover of this book and Linda Shapiro who was willing to invest her time on a Sunday to help us out. Our deepest appreciation to the children who participated in the videos, and their families. You are all movie stars! And thank you to Nicholas Sousa, the videographer, for his professionalism and attention to detail. Diane would like to acknowledge her husband Steve and her daughters, Emma and Kate, without whose encouragement and support this project would never have been started or finished. Thank you to Allison Adams, Kate Garcia, and Gina Rocque for reading draft manuscripts of specific chapters. And a special note of thanks to Kathleen Newcombe for providing after-school snacks and a quiet place to write, as well as invaluable suggestions and editing notes. Jackie would like to thank her friends who have been a great support network throughout this project. And a special thanks to Jennifer Taps Richard, San Diego Unified School District, co-owner and director of SLPath. com. She has patiently answered questions and provided great resources that helped me develop many of the chapters.

xi

Reviewers

Plural Publishing, Inc. and the authors would like to thank the following reviewers for taking the time to provide their valuable feedback during the development process:

Iris Johnson Arnold, PhD, CCC-SLP Associate Professor Department of Speech Pathology and Audiology Tennessee State University Nashville, Tennessee

Verna Mae Chinen, MS, CCC-SLP Instructor Communication Sciences and Disorders Department Hawaii Department of Education and University of Hawaii at Manoa Honolulu, Hawaii

Christine Gerber, MS, CCC-SLP Clinical Supervisor Communication Disorders and Sciences State University of New York at Fredonia Fredonia, New York

Joy Good, PhD, CCC-SLP Assistant Professor Department of Communications Disorders Arkansas State University Jonesboro, Arkansas

xiii

To all the speech-language clinicians who are working hard to make a difference in the lives of children.

1 Introduction to Phonological Treatment of Speech Sound Disorders

Importance of Phonological Treatment Methods Speech sound disorders play a central role in the lives of most Speech-Language Pathologists (SLPs) and the children they serve. In a recent ASHA survey on SLP caseloads in the schools, over 90% of clinicians reported that they work with children with articulation/ phonological disorders (American Speech-Language-Hearing Association, 2014). Every day, therapists are faced with questions such as: What options are available for evidencebased treatment of speech sound disorders? How do I implement the various methods? What types of analyses should I conduct prior to choosing a therapy approach? How do I write appropriate intervention goals and gather necessary data to support clinical decision making? This workbook will attempt to answer these questions and more. The majority of the therapy approaches presented in these pages focus on the improvement of children’s phonological skills — that is, knowledge of the organization and function of phonemes within our language system. Regardless of the type or etiology of a speech sound disorder, phonological treatment methods may be an important component of an effective intervention plan. For children who present with a phonological disorder as their primary impairment, one or more of these methods may form the core of their therapy program. For others, particularly those with complex needs, phonological treatment may be one piece of a much larger intervention puzzle. In recent decades, our profession has experienced exciting developments regarding the treatment of phonological deficits. The advancement of innovative theories and the ensuing research has resulted in new therapeutic protocols that are more efficient and effective. Yet many clinicians continue to use less optimal strategies due to limitations in time or lack of knowledge. This workbook is designed to help bridge the gap between research and application. Our goal is to provide therapists with a road map for implementing new treatment methods that will have a significant, positive impact on children’s speech intelligibility and communicative effectiveness.

1

2

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

How Is Speech Sound Disorder Defined? As reflected in the title of this workbook, we have chosen to use the term “speech sound disorder” to refer to all types of sound production difficulties, including, articulation disorders, phonological disorders, and motor speech disorders (e.g., apraxia and dysarthria). This label has been adopted by the American Speech-Language-Hearing Association (ASHA) to describe all disorders related to the “perception, motor production, and/or phonological representation of speech sounds” (n.d.). See Figure 1–1 for ASHA’s definition. The two most common subtypes of speech sound disorders are articulation disorders and phonological disorders. The term articulation disorder has been used for decades to refer to an inability to physically produce speech sounds. In his classic text Speech Correction, Charles Van Riper (1939) listed articulation disorders as one of the subcategories of speech disorders. It was characterized by substitutions, omissions, additions and/or distortions of speech sounds. Articulation disorders originally reflected a client’s inability to perceive and/or discriminate specific sounds, and to produce them motorically. They were considered a motor production problem in which the individual had to relearn the positioning and movements of the articulators. The suggested therapy (phonetic or traditional motor approach) consisted of a clinician training the client in a sound-by-sound method to correctly produce individual speech sounds (Van Riper, 1939). Articulation disorder is a term still widely used by clinicians and the traditional motor approach, now almost 80 years later, is still a method utilized among the majority of SLP clinicians (e.g., Brumbaugh & Smit, 2013). A conceptual shift occurred in the 1970s, when the field of linguistics began to influence how speech-language clinicians viewed their profession. The resulting nomenclature “phonology, morphology, semantics, syntax and pragmatics” was born from this linguistic orientation. Phonology is a subcategory of language, and the way speech sounds function within the linguistic system became a focus of intervention. Thus, the phoneme (i.e., the function of a speech sound), rather than the phone (i.e., the form of a speech sound), was now analyzed. Phonological disorders are language-based, with the core difficulty being the language-specific function of phonemes. According to ASHA (n.d.), they “result from impairments in the phonological representation of speech sounds and speech segments — the system that generates and uses phonemes and phoneme rules and patterns within the context of spoken language.” Therefore, phonological interventions, such as those presented in this book, focus on the phoneme as a meaning differentiating entity. That is, they specifically target the use of sounds to create meaning in the context of spoken words. Children who demonstrate phonological disorders will obviously benefit from phonological interventions. However, we should not overlook the possibility that children who present with articulation needs or motor speech difficulties may also experience deficits in the area of phonological learning. In fact, Strand and McCauley (2008) recommend that we do not view the various types of speech sound disorders as mutually exclusive entities, but rather consider the relative contribution that each may make to a particular child’s unique profile. Children often demonstrate difficulty with more than one aspect

3

(e.g., hearing impairment) or idiopathic (i.e., unknown etiology)!

syndrome or condition (e.g., Down syndrome) sensory deficit

Caused by: motor-speech disorder (apraxia and dysarthria) structural condition (cleft palate/craniofacial anomalies)

the function of speech sounds (Phonological Disorders)

(Articulation Disorders) and/or

Impacting: the form of speech sounds

Figure 1–1.  Speech Sound Disorders as defined by the American Speech-Language-Hearing Association. Source: Adapted from ASHA (n.d.) Practice Portal http://www.asha.org/Practice-Portal/clinical-Topics/Articulation-and-Phonology/

prosody that impacts speech intelligibility

syllable shapes stress patterns and/or

of speech sounds/segments

motor production and/or phonological representation

A disorder involving: perception

Speech Sound Disorders

4

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

of speech production. The same child may have trouble physically producing sounds (i.e., articulation), using those sounds contrastively to create words (i.e., phonology), and sometimes even achieving the needed motor planning for successful speech (i.e., apraxia). As clinicians, we must identify each child’s areas of deficit and select interventions that will teach them the skills they need to be effective communicators.

Prevalence and Impact of Speech Sound Disorders How many children actually have speech sound disorders? Speech/language specialists who work with children know that those with speech sound disorders are a large percentage of their caseload. In addition, children with speech sound disorders encompass a wide range of difficulties. These children could have been diagnosed with childhood apraxia of speech or have specific speech errors as a result of a cleft palate. They could have relatively isolated speech sound problems, such as a lateral s-sound, or they could have multiple errors and be partially or totally unintelligible. Although it would seem easy to count the number of children with speech sound disorders, if you look at the literature you will find significant variation in prevalence figures. Figure 1–2 is an overview of the percentages of children who have reportedly been diagnosed with speech sound disorders. As can be seen from Figure 1–2, reports of the number of children diagnosed with speech sound disorders vary widely. According to a systematic review conducted by Law, Boyle, Harris, Harkness, and Nye (2000), estimates of prevalence for children ages 5 to 7 years ranged from 2% to 25%. These statistics highlight the scope of the problem before us. Speech sound disorders are a common and widespread phenomena that have a negative impact on the lives of many children and their families. Recent research indicates that this impact may be even greater than previously known. Although once viewed as an isolated impairment, we now know that difficulties with speech sound production are often correlated with literacy and academic challenges (Peterson, Pennington, Shriberg, & Boada, 2009). A recent study conducted by Farquharson (2015) revealed that even children who no longer exhibit a speech disorder and have been dismissed from therapy, demonstrate lower skills in some areas of language and literacy than their typically developing peers. These areas of deficit include expressive and receptive vocabulary and word reading. Early identification and remediation of speech sound disorders is particularly important, given studies indicating that children who do not achieve typical speech production by 6 years of age are more likely to have difficulty acquiring literacy skills (Nathan, Stackhouse, Goulandris, & Snowling, 2004; Raitano, Pennington, Tunick, Boada, & Shriberg, 2004). The good news is that we can make a difference in these children’s lives. If we identify speech sound disorders early and provide appropriate intervention, we can reduce the length and severity of phonological disorders (Baker & McLeod, 2011a; Harrison & McLeod, 2010). The treatment methods presented in this book offer effective and efficient ways to achieve this outcome.

1 Introduction to Phonological Treatment of Speech Sound Disorders

4-Year-Old Children

• 4% of 1,494 children in Australia. • Eadie et al., 2015.

5-Year-Old Children

• 6.4% of 1,655 children in Canada. • Beitchman, Nair, Clegg, & Patel, 1986.

6-Year-Old Children

• 3.8% of 1,328 children in the U.S. • Shriberg, Tomblin, & McSweeney, 1999.

7-Year-Old Children

• 10–13% of 15,496 children in the UK had some degree of “speech impairment,” 2.5% had been seen by a speech therapist. • Peckham, 1973.

8-Year-Old Children

• 18% of 741 children in the UK had a “speech impairment.” • Roulstone, Miller, Wren, & Peters, 2009.

5

Figure 1–2.  Selected prevalence figures for children with speech sound disorders.

Principles of Phonological Treatment There are several principles of treatment that are utilized throughout this workbook. The following underlying concepts reflect these: 1. The focus is on improving phonological systems rather than just the articulation of individual speech sounds (Bowen, 2011). Thus, a linguistically relevant conceptual framework (phonological system), rather than a motoric one (articulatory system), is emphasized. Phonological intervention targets the linguistic function of sounds; that is, the use of phonemes to create meaningful words. Traditional articulation therapy, on the other hand, centers on the physical production of sounds, including the place, manner, and voicing needed for their correct realization. However, although the concentration consists of primarily establishing the child’s phonological system, the articulatory features of a target sound cannot be ignored. Often articulatory principles will first be used to establish the error sound(s), and then the correct production will later be used in contrasting phonemes. 2. Selection of target behaviors is usually based on patterns reflecting the sound errors (Bernthal, Bankson, & Flipsen, 2017). Traditional articulation

6

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

therapy advises us to target each sound error in turn until all sounds have been remediated. Phonological approaches invite us instead to look for patterns of errors. Individual targets may become the vehicle for remediation of whole groups of phonemes or structures. For example, if a child learns to contrastively produce the fricative /s/ in minimal pair words, this new phonological knowledge may generalize to production of other fricatives. Or, if a child learns to use final /t/ to differentiate Consonant-Vowel-Consonant (CVC) from Consonant-Vowel (CV) productions, the use of final phonemes in all CVC structures may improve. 3. Phonemic contrasts that were previously neutralized are established (Fey, 1992; Williams, 2010). Phonemic/phonological contrast is the ability to successfully differentiate between phonemes and structures, so that a variety of words may be produced. To this end, these approaches involve analysis of a child’s entire phonology, including both phonemes and word shapes, in order to determine what possibilities the child possesses for creating these contrasts. For example, if a child uses the phonemes /s/ and /t/ contrastively, he or she will produce words such as “sea” and “tea” appropriately. However, if the child collapses the contrast between the two phonemes and produces both as /t/, he or she will produce these two words as “tea”— that is, as homonyms. Homonyms are two words having the same pronunciation but different meanings. In this case (“tea” used for “tea” and “sea”) should not be homonyms. Production as such could lead to confusion for the listener. Word shapes also create contrast, as seen in the word pair “sea” and “seat.” If a child is not able to contrast between CV and CVC word shapes, he or she might produce these two words both as “sea,” and thus again create homonyms. Children with phonological disorders often have difficulty achieving differentiation between phonemes and word shapes, and, therefore, have an abundance of homonymy in their lexicons. The increase of contrast and resulting decrease in homonymy are, therefore, important goals of many phonological methodologies. 4. Intervention targets are carefully selected to facilitate generalization, or broader change, in the phonological system (Gierut, 2007). Going sound-by-sound through the errors a child produces has not demonstrated generalization to other sounds or patterns (Hodson, 2011). However, phonological approaches have demonstrated this generalization ability. For example, targeting complex sound patterns such as consonant clusters, has generalized to singleton productions and to other clusters. This ability to create broader changes is an important aspect of phonological approaches. 5. Intervention should be meaning-based (Lowe, 1994). Phonological intervention addresses the use of phonemes to create meaningful words. Although several of the approaches use traditional motor strategies or other types of cueing to obtain a particular sound, the target is then immediately used in meaningful words and not practiced in isolation. Thus, these therapy approaches focus on the linguistic function of segments, rather than

1 Introduction to Phonological Treatment of Speech Sound Disorders

7

their physical form. The final objective is improving children’s ability to communicate in daily interactions and, therefore, specific functional activities in natural contexts are a component of some phonological methodologies.

Workbook Content and Layout In this workbook, you will find practical guidelines for the implementation of phonological treatment methods in your daily work as a clinician. To this end, Chapter 2 includes information regarding assessment and analysis, including easy-to-use data sheets that are demonstrated in the chapters that follow. Each of the next eight sections (Chapters 3 to 10) provides a clear explanation of a phonological intervention method, using case studies to illustrate its application in a step-by-step progression. The approaches included in this book are: Chapter 3: Minimal Pair Therapy, Chapter 4: Multiple Oppositions, Chapter 5: Maximal Oppositions, Chapter 6: Complexity Approaches, Chapter 7: Phonotactic Therapy, Chapter 8: Core Vocabulary Approach, Chapter 9: Cycles Approach, and Chapter 10: Phonological/Phonemic Awareness. The chapters devoted to a specific intervention method contain: 1. A brief overview of the therapy method; 2. Examples of supporting research; 3. Target selection procedures; 4. Sample goals and data collection strategies; 5. Treatment guidelines; 6. Group therapy ideas. The final chapter of this book includes case studies that illustrate the application of the intervention methods presented. Short video clips demonstrating the therapy methods selected for each case study are available online at http://www.pluralpublishing.com/ publication/ptssd. Below you will find a brief description of each method included in this workbook.

Summary of Treatment Approaches There are many therapies to choose from that effectively address phonological deficits. In fact, Baker and McLeod (2011a) identified 46 different intervention approaches and seven target selection options for treatment of phonological disorders in a narrative review of 134 studies published between 1979 and 2009. In this workbook, we have selected eight of these to review, based upon factors including supporting evidence and available resources for implementation. Some of the approaches included in this book represent comprehensive therapeutic protocols (e.g., cycles approach, multiple oppositions), while others primarily describe a specific target selection strategy (e.g., complexity

8

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

approach, phonotactic therapy). All are designed to remediate phonological difficulties, yet do not necessarily exclude the principles which govern a traditional sound-by-sound approach. Many of these therapies, as are noted in later chapters, first use aspects of traditional articulation therapy. There is not a dichotomy between articulation and phonological treatment, but rather both work together to optimally aid children with speech sound disorders. Here are brief summaries of the eight intervention methods. Minimal pair therapy is the original contrastive therapy using pairs of words that differ by only one phoneme (a minimal word pair). Targets are typically based on the specific error pattern demonstrated by the child. Thus, the intended sound and the error production are set in opposition to each other. For example, if the child produced /f/ as /p/ then minimal pairs such as “fan” and “pan” could be used. If the child produces “fan” and “pan” as /pæn/ then the semantic consequences of misproductions are highlighted. The goal is an increase in the use of contrastive phonemes and word shapes for the purpose of functional communication (Barlow & Gierut, 2002). Multiple oppositions therapy is a variation of minimal pair therapy that incorporates larger treatment sets. Rather than targeting one contrast pair at a time, a whole family of homonyms is targeted simultaneously. Thus, if the child uses /t/ for /k, ʃ, tS, and st/, minimal word pairs contrasting each of these phonemes would be created. An example would be using the words “top,” “cop,” “shop,” “chop,” and “stop” as a contrastive word set during intervention activities. Each child’s unique phonological system is first analyzed to identify phonemic collapses, and then targets are developed to systematically increase contrasts and reduce homonymy across the child’s phonological system (Williams, 2010). Maximal oppositions therapy uses pairs of words with two phonemes that represent maximal distinctions in their production. Optimal targets incorporate contrasts across many distinctive or descriptive features (i.e., place, manner, voicing), and at least one major class feature (i.e., sonorant, consonantal, vocalic). A variation, known as Treatment of the Empty Set, creates word pairs representing two unknown sounds (i.e., not able to be produced by the child), instead of one known sound (i.e., produced by the child), and one unknown sound. These adjustments to target selection have been shown to produce broader, more system-wide change in children’s phonological systems (Gierut, 1989, 1992). The complexity approach is actually a target-based approach. Therefore, the foundation consists of how targets are selected, and not a specific treatment protocol. It is unique in that it targets sounds that are more “complex.” More complex sounds are those (1) that are excluded from the child’s inventory, (2) that are not stimulable, (3) imply production of other classes of sounds, and (4) that are later-developing. These categories create the priority for target selection. In addition, specific consonant clusters could alternatively be chosen as targets. Consonant clusters are chosen based on sonority (specific loudness differences between the elements of the cluster), the lack of stimulability of the individual cluster elements, and the productional difficulty of the specific elements. It appears that by targeting word pairs that contain complex phonemic elements, more generalization occurs in the child’s phonological system (Gierut, 2001). Phonotactic therapy emphasizes word shapes as important structures for creating contrast in a child’s developing phonological system. This approach to target selec-

1 Introduction to Phonological Treatment of Speech Sound Disorders

9

tion allows clinicians to simultaneously expand a child’s phonotactic inventory, sound inventory, and vocabulary. Phonotactics looks at the rules governing how syllables and words are constructed. It includes rules about how consonants may combine to form clusters, how consonants and vowels may combine to form syllables, and how syllables may join together in words, including their stress patterns. Phonotactic intervention expands both the syllable and word structure possibilities of the child while also introducing new phonemes within these structures. It represents a practical option for children who demonstrate restricted word shape inventories, or who exhibit whole word error patterns (Velleman, 2002, 2016). Core vocabulary is a whole-word approach for children who demonstrate an inconsistent speech disorder. Inconsistency is characterized by the unpredictable use of different phonemes when saying the same word repeatedly. Children with an inconsistent speech disorder may indeed produce the same word differently each time they say it. These children appear to need a unique type of treatment procedure (Dodd, Holm, Crosbie, & McIntosh, 2010). The core vocabulary chapter is structured to first give specific guidelines for diagnosing an inconsistent speech disorder. Treatment is based on stabilizing the productions of a core set of vocabulary items that are very functional for the specific child. If consistency of productions can be attained, these children are able to transition to other types of goal-directed phonological treatment protocols. The cycles phonological remediation approach targets a broad age-range of children with severe-to-profound expressive speech sound disorders. It was developed to facilitate the acquisition of intelligible speech in “cycles,” periods of time where one specific pattern is targeted. The cycles phonological remediation approach has a detailed way to select targets, structure therapy, and move from cycle-to-cycle of patterns. In contrast to other mentioned approaches, cycles training targets specific preselected words in such a way that 100% accuracy is obtained in production-practice drills. Thus, it is critical that the child be “stimulable” and capable of producing the pattern with “assists,” such as tactile cues or amplification. According to the authors, the concept that sets this approach apart from others is that it appears that children make progress in a relatively short time (Hodson & Paden, 1991). The last treatment chapter describes using phonological/phonemic awareness strategies as a portion of treatment for speech sound disorders. First, this chapter briefly summarizes the very complex interdependency between phonological/phonemic awareness, speech sound disorders, and learning to read and write. It defines key terms, such as phonological versus phonemic awareness, and gives examples of the types of tasks that are considered within each area (Gillon, 2018). These tasks are also defined according to broad-based developmental parameters. For example, certain skills are considered very early, such as segmenting the individual words in a sentence (how many words are in “I am at home”), while others typically occur later developmentally, such as phoneme manipulation (“What would the word “team” be if you reversed the “t” and the “m”?). The second portion of the chapter attempts to show how phonological/phonemic awareness activities could be directly connected to treatment for speech sound disorders. Table 1–1 provides a brief description of the characteristics of each approach. You may use this chart as a quick guide for identification of therapies that may be appropriate for a particular child. References for this table are extensively provided within each of

10

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 1–1.  Summary of Phonological Treatment Options Type of Phonological Treatment

Approximate Age Range, Severity

Characteristics of Disorder

Minimal Pair Therapy e.g., Barlow & Gierut, 2002

3+ years of age. Mild to mild-moderate phonological disorder.

Does the child have one or only a few speech sound errors that involve a lack of contrast?

Multiple Oppositions e.g., Williams, 2010

Research supports 3 to 6½ years of age, most with moderate or severe speech sound disorders.

Does the child demonstrate a high number of phonemic collapses? A high degree of homonymy?

Maximal Oppositions e.g., Gierut, 1989, 1992

Research supports 3;6 to 5+ years of age. At least six sounds excluded from inventory and extensive “gaps” in their phonemic inventory.

Does the child demonstrate multiple speech sound errors across a variety of sound classes? Could maximal oppositions be found that would reflect the child’s phonological system needs?

Complexity Approaches e.g., Gierut, 2001

3 to 7 years of age with a moderate to severe speech sound disorder.

Does the child’s phonological system need overall change? Does targeting later sounds to achieve earlier ones seem to be warranted?

Phonotactic Therapy e.g., Velleman, 2001, 2016

Age not specified. Moderate, severe, or profound speech sound disorder.

Does the child have difficulties with early syllable or word shapes?

Core Vocabulary e.g., Dodd, Holm, Crosbie, & McIntosh, 2010

3 to 11 years of age, however, the most progress was made with children around 3 years old.

Does the child demonstrate an inconsistent speech sound disorder?

Cycles Phonological Remediation Approach e.g., Hodson & Paden, 1991

Research supports 2;9 to 7 years of age. Children should demonstrate a severe phonological disorder.

Is the child unintelligible?

Phonological/Phonemic Awareness Training e.g. Gillon, 2018

From 3 to 9 years of age. Children should have skills that test below the norm range.

Are skills low?

the chapters. We encourage clinicians to use the information provided in this workbook within the context of an evidence-based framework for clinical decision making, which includes consideration of current research and client values, as well as your own clinical expertise (please see Baker & McLeod, 2011b for an excellent tutorial on evidence-based practice (EBP) and speech sound disorders). We hope this workbook will inspire you to implement new strategies to achieve optimal outcomes with the children you serve.

1 Introduction to Phonological Treatment of Speech Sound Disorders

11

References American Speech-Language-Hearing Association (n.d.) Speech sound disorders: Articulation and phonology [Practice portal]. Retrieved from www.asha.org/Practice-Portal/Clinical-Topics/Artic ulation-and-Phonology. American Speech-Language-Hearing Association (2014). ASHA 2014 school survey: SLP caseload characteristics report. Rockville, MD: Author. Baker, E., & McLeod, S. (2011a). Evidence-based practice for children with speech sound disorders: Part 1 narrative review. Language, Speech, and Hearing Services in Schools, 42, 102–139. Baker, E., & McLeod, S. (2011b). Evidence-based practice for children with speech sound disorders: Part 2 application to clinical practice. Language, Speech, and Hearing Services in Schools, 42, 140–151. Barlow, J. & Gierut, J. (2002). Minimal pair approaches to phonological remediation. Seminars in Speech and Language, 23(1), 57–67. Beitchman, J. H., Nair, R., Clegg, M., & Patel, P. G. (1986). Prevalence of speech and language disorders in 5-year-old kindergarten children in the Ottawa-Carleton region. Journal of Speech and Hearing Disorders, 51, 98–110. Bernthal, J. E., Bankson, N. W., & Flipsen Jr., P. (2017). Articulation and phonological disorders: Speech sound disorders in children (8th ed.). Boston, MA: Pearson. Bowen, C. (2011). Children’s speech sound disorders. Retrieved from http://www.speech-languagetherapy.com/ Brumbaugh, K. M., & Smit, A. B. (2013). Treating children ages 3–6 who have a speech sound disorder: A survey. Language, Speech, and Hearing Services in Schools, 44, 306–319. Dodd, B., Holm, A., Crosbie, S., & McIntosh, B. (2010). Core vocabulary intervention. In A. L. Williams, S. McLeod, & R. J. McCauley (Eds.), Interventions for speech sound disorders, (pp. 117– 136). Baltimore, MD: Brookes. Eadie, P., Morgan, A., Ukoumunne, O. C., Ttofari Eecen, K., Wake, M., & Reilly, S. (2015). Speech sound disorder at 4 years: Prevalence, comorbidities, and predictors in a community cohort of children. Developmental Medicine & Child Neurology, 57, 578–584. Farquharson, K. (2015). After dismissal: Examining the language, literacy, and cognitive skills of children with remediated speech sound disorders. Perspectives on School-Based Issues, 16(2), 50–59. Fey, M. E. (1992) Clinical forum: Phonological assessment and treatment. Articulation and phonology: Inextricable constructs in speech pathology. Language, Speech, and Hearing Services in Schools, 23, 225–232. (Reprinted from Human Communication Canada, 1983, 9, 7–16). Gierut, J.A. (1989). Maximal opposition approach to phonological treatment. Journal of Speech and Hearing Disorders, 54, 9–19. Gierut, J.A. (1992). The conditions and course of clinically induced phonological change. Journal of Speech and Hearing Research, 35, 1049–1063. Gierut, J. A. (2001). Complexity in phonological treatment: Clinical factors. Language, Speech, and Hearing Services in Schools, 32, 229–241. Gierut, J. A. (2007). Phonological complexity and language learnability. American Journal of Speech-Language Pathology, 16, 6–17. Gillon, G. T. (2018). Phonological awareness: From research to practice (2nd ed.). New York, NY: Guilford Press. Harrison, L. J., & McLeod, S. (2010). Risk and protective factors associated with speech and language impairment in a nationally representative sample of 4- to 5-year-old children. Journal of Speech, Language, and Hearing Research, 53, 508–529.

12

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Hodson, B. W. (2011). Enhancing phonological patterns of young children with highly unintelligible speech. The ASHA Leader, 16, 16–19. Hodson, B. W., & Paden, E. P. (1991). Targeting intelligible speech: A phonological approach to remediation (2nd ed.). Austin, TX: Pro-Ed. Law, J., Boyle, J., Harris, F., Harkness, A., & Nye, C. (2000). Prevalence and natural history of primary speech and language delay: Findings from a systematic review of the literature. International Journal of Language and Communication Disorders, 35, 165–188. Lowe, R. J. (1994). Phonology: Assessment and intervention application is speech pathology. Baltimore, MD: Williams & Wilkins. Nathan, L., Stackhouse, J., Goulandris, N., & Snowling, M. (2004). The development of early literacy skills among children with speech difficulties: A test of the “critical age hypothesis.” Journal of Speech, Language, and Hearing Research, 47(2), 377–391. Peckham, C. S., (1973). Speech defects in a national sample of children aged seven years. British Journal of Disorders of Communication, 8, 2–8. Peterson, R. L., Pennington, B. F., Shriberg, L. D., & Boada, R. (2009). What influences literacy outcome in children with speech sound disorder? Journal of Speech, Language, and Hearing Research, 52, 1175–1188. Raitano, N. A., Pennington, B. F., Tunick, R. A., Boada, R., & Shriberg, L. D. (2004). Pre-literacy skills of subgroups of children with speech sound disorders. Journal of Child Psychology and Psychiatry, 45(4), 821–835. Roulstone, S., Miller, L. L., Wren, Y., & Peters, T. J. (2009). The natural history of speech impairment of 8-year-old children in the Avon Longitudinal Study of Parents and Children: Error rates at 2 and 5 years. International Journal of Speech-Language Pathology, 11, 381–391. Shriberg, L. D., Tomblin, J. B., & McSweeny, J. L. (1999). Prevalence of speech delay in 6-year-old children and comorbidity with language impairment. Journal of Speech, Language, and Hearing Research, 42, 1461–1481. Strand, E. A., & McCauley, R. J. (2008). Differential diagnosis of severe speech impairment in young children. The ASHA Leader, 13, 10–13. Van Riper, C. (1939). Speech correction: Principles and methods. Englewood Cliffs, NJ: Prentice-Hall. Velleman, S. L. (2002). Phonotactic therapy. Seminars in Speech and Language, 23, 43–55. Velleman, S. L. (2016). Speech sound disorders. Philadelphia, PA: Wolters Kluwer. Williams, A. L. (2010). Multiple oppositions intervention. In A. L. Williams, S. McLeod, & R. J. McCauley (Eds.), Interventions for speech sound disorders in children. Baltimore, MD: Brookes.

2 Ways to Collect and Analyze Data

Overview This chapter is an overview of the different ways that data can be collected for establishing treatment targets and therapy goals described within this workbook. This chapter will show how to streamline the data collected from a standardized speech assessment and a spontaneous speech sample. Other tools will be offered as a means to supplement your original results. As this workbook contains phonetic transcription extensively, the first goal is a brief review of phonetic transcription and the more common diacritics. In addition, independent, as well as relational analyses will be defined. Depending on the age and level of the child, both of these options provide worthwhile information. Finally, intelligibility measures will be outlined. For certain children, a goal relating to increasing intelligibility or overall accuracy might be a better choice, as opposed to focusing on individual sounds. If this is the case, then measuring intelligibility and/or accuracy will become a necessity. This chapter will provide the clinician with practical strategies to increase diagnostic efficiency with the limited time that is available.

Phonetic Transcription and Diacritics Phonetic transcription is one of those skills that everyone learned at some point in their academic program. But as with most skills, if transcription is not used routinely, it will soon become forgotten and a chore to use. However, it is an excellent way to document the present levels and progress of our children. Some basic guidelines that are applied throughout this workbook include the use of brackets [  ] and virgules /  /. Although many clinicians use these interchangeably, there are some important differences. Brackets [ ], are used when we refer to the actual spoken sound, in other words, the phone. If a child says a particular speech sound, then brackets should be used. On the other hand, virgules / / are used when referencing the phoneme. Diacritics are marks that can be added to transcription symbols to give them a specific phonetic value. There are many different diacritic marks; however, based on the authors’ experience, only a handful are utilized on a regular basis. Later in this section,

13

14

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

these diacritics will be presented. This review is not comprehensive but references will be given if the reader requires other specific diacritics to exemplify a particular sound. The International Phonetic Alphabet is the standard that is used in transcription. One website that offers an audio link to the vowels, consonants, and diacritics is https://web​ .uvic.ca/ling/resources/ipa/charts/IPAlab/IPAlab.htm The audio link uses Quick Time Player for the various sound productions. This is probably on your computer, but if not, it will need to be installed before you can access the sounds. As you click on the various symbols to listen to them, the phonetic description is noted at the bottom of your screen.

Vowels There are many vowels of American English. For vowels, the tongue placement is used to characterize their production features: high versus low (the tongue close to the palate versus farther away), and front versus back vowels (if the highest point of the tongue is forward in the mouth or more toward the back of the mouth). Although there are other production features used to describe vowels, the only one that might be relevant for the purpose at hand is whether the vowel shows some degree of lip rounding (rounded vowels), or more lip spreading (unrounded vowels). Front vowels are basically unrounded, produced with lip spreading, whereas back vowels are rounded vowels, with lip rounding. The degree of spreading and rounding is the most for high vowels, and decreases somewhat as the tongue placement moves lower in the mouth. Thus, low front vowels show relatively little lip spreading, and low back vowels demonstrate little lip rounding. The lowest back vowel [ɑ] is considered unrounded. The central vowels with r-coloring [] and [2] are also produced with lip rounding, but the degree varies from speaker to speaker. The vowel quadrilateral is frequently used to demonstrate the features of vowels. Figure 2–1 is a diagram of the vowel quadrilateral with word examples. The rather strange shape of the vowel quadrilateral is an attempt to exemplify the oral cavity. The top line indicates the roof of the mouth, while the bottom line represents the floor of the mouth. The front boundary is more slanted due to the fact that the tongue at the front of the mouth has a greater range of motion; it can go up and down but also move forward and backward. On the other hand, the back of the tongue is relatively restricted to a rather up and down movement. In addition to the vowels represented in this vowel quadrilateral, there are also diphthongs in American English. A diphthong is characterized by a gliding motion from the first vowel (the onglide) to the second vowel (the offglide). Table 2–1 is a list of diphthongs with their phonetic transcription and words exemplifying the vowel. It should be noted that the reader may have learned a different way to transcribe the diphthongs. Specific textbooks vary somewhat in the transcription of diphthongs. In addition, the two vowel symbols can have a bar or line above or below the symbols indicating that the vowels are one single unit and not two separate vowels or the second vowel symbol may be elevated ([oU] or [eI]), demonstrating less intensity in its production.

15

2  Ways to Collect and Analyze Data

FRONT

CENTRAL

BACK

[i] eat, see

[u] loop, soon

[ɪ] in, big

HIGH

[ʊ] wood, could [ə] away, [ɝ] turn, shirt ago

[e]1 gate, cake [ɛ] neck, shed

[ɚ] river, gather

[o]1 no, boat

[ʌ] cup, love

[ɔ]2 caught, law

MID

[æ] bad, ask LOW [a]2 shop, talk

[ɑ]2 hot, mop

1.

These vowels can be transcribed as [eɪ] and [oʊ] as they are typically diphthongs in most stressed syllables.

2.

The three vowels [a], [ɑ], and [ɔ] vary from speaker to speaker and word situation. They are allophonic variations in American English.

Figure 2–1.  The vowel quadrilateral of American English with word examples.

Table 2–1.  Diphthongs of American English with Word Examples [eI] or [e]

hay, plate, take, apron, major

[oU] or [o]

oats, phone, soap, open, yoyo

[ɔI]

boy, toy, coin, noisy, ahoy

[aI] or [AI]*

my, kite, like, biker, shiny

[aU] or [AU]*

out, house, town, cowboy, chowder

*The differences in the last two transcriptions, using the [a] or [ɑ], reflect slightly different productions. However, the two productions are not phonemically relevant, as they do not signal differences in meaning.

Consonants Table 2–2 contains the phonetic transcription of each of the consonants of American English with word examples. In this text, the consonantal “r” that occurs, for example, at the beginning of a word or syllable, is transcribed as [ɹ]. You may have learned to

16

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 2–2.  Phonetic Transcription of Consonants of American English [p]

pig, pea, open, carpet, top, nap

[z]

zoo, zero, closing, dozen, tease, nose

[b]

bee, bow, rabbit, cabin, tub, robe

[ ʃ ]

show, shoe, wishing, cushion, rush, dish

[t]

two, ten, guitar, eighteen, hot, night

[Z]

Jacques, measure, pleasure, beige, rouge

[d]

dime, dish, roadway, Tuesday, mud, hide

[θ]

think, thick, nothing, without, bath, south

[k]

cat, cow, acorn, bacon, tack, luck

[ð]

the, that, clothing, brother, bathe, breathe

[g]

go, girl, wagon, bigger, wag, log

[h]

house, hot, ahoy, ahead [h] does not occur in the word-final position in American English

[m]

mat, mop, hammer, famous, tame, room

[tʃ]

chew, chin, teacher, reaching, beach, witch

[n]

no, not, tiny, pony, run, soon

[dZ]

jet, jam, magic, pigeon, cage, bridge

[ŋ]

singer, longer, ring, wrong [ŋ] does not occur in word-initial position in American English

[w]

win, well, away, freeway

five, fell, offer, coffee, off, cough

[ j ]

[f]

[w] does not occur in the word-final position in American English use, yarn, kayak, yo-yo [j] does not occur in the word-final position in American English

[v]

van, vase, favor, seven, move, give

[ɹ]

rat, red, arrow, around, mother*, fur*

[s]

so, sock, icy, kissing, less, choice

[ l ]

leg, left, yellow, color, well, call

*Words with “r” in the final position could be transcribed with [2] or [] or with the rhotic diphthongs. For example, “mother” [mð2] or “fur” [f].

transcribe the “r” in “red” as [r]. The transcription [r] is commonly used for r-sounds in words regardless of the place they occur (word-initial, -medial, or -final). However, according to the International Phonetic Alphabet, the accurate symbol is [ɹ]. This symbol will be used throughout this workbook. In addition, in word-final position “r” can be transcribed as a rhotic vowel (for example “car” is [kɑ2]) or as a vowel + [r], such as [kɑr]. For the purposes of this workbook, the rhotic vowel will be used in word-final position. In addition, there are a few phonetic symbols that may prove to be handy, and are commonly used when transcribing children’s speech. These symbols do not belong just to the transcription of disordered speech as many are used to transcribe normal variations in productions. These include the glottal stop, the flap or tap, syllabics, and the notation for syllable boundaries. Symbols with explanations are contained in Appendix 2–1.

2  Ways to Collect and Analyze Data

17

Allophones and Allophonic Variation Allophones or allophonic variations are differences in production that do not change the meaning of a word. Allophonic variations include articulatory changes a speaker might make when pronouncing phones within a specific context. Some changes are predictable. For example, in American English the stops /p, t, k/ are often aspirated. Aspiration is the slight puff of air heard between the end of the plosive and the following vowel transcribed by an elevated h after the stop. The initial stops in “pie,” “talk,” and “keep” are all aspirated, [phaI], [thɑk], khip]. However, when /p, t, k/ are preceded by /s/, they are unaspirated, for example, [spaI], [stɑp]. Unlike other languages, such as Chinese or Korean, the aspiration or lack thereof does not change the meaning in American English. Allophonic variations are not considered aberrant productions. They occur in specific, context-dependent circumstances. Table 2–3 contains several allophonic variations noted in American English. During the administration of a standardized measure, the child may produce these variations. It is a good idea to check the guidelines of the test as many of these might not possibly be considered errors.

Diacritics Diacritics are specific marks added to phonetic transcription symbols to give them a particular phonetic value. There are literally dozens of diacritical markers that can be added to vowel and consonant symbols. In addition, the extIPA system was created in 1990, and most recently revised in 2015, to address the needs of clinicians when transcribing disordered speech. The extensions of the International Phonetic Alphabet for disordered speech are contained at the following website: https://www.international phoneticassociation.org/sites/default/files/extIPA_2016.pdf With a specific child, it may become necessary to use many of the symbols within the extIPA, or even devise new symbols to describe your needs. However, the authors have found that a handful of diacritics are typically used frequently. Table 2–4 contains a list of commonly used diacritics. The next section will examine ways that data can be organized and collected for a specific child. Each child, based on his/her age and level, may need a somewhat different approach. The following are two ways that data can be summarized.

Independent Versus Relational Analyses An independent analysis is one way to collect speech sound/phonemic data from a child. Independent in this case means that the child’s system is analyzed independently from what is considered “correct” or a typical pronunciation. In this type of analysis, only the child’s productions are collected, there is no comparison, such as the child substituted [w] for [r]. An independent analysis becomes a valuable tool, for example, if the child is very young and cannot name pictures for a standardized speech assessment. It also may become a necessity if the child’s phonology is so deviant that the clinician would like to find patterns that may govern this system.

Table 2–3.  Examples of Allophonic Variations in American English Allophonic Variation Aspiration: p, t, k

Changes Aspirated in initial stressed position

Examples [phɑt] and [thɑp] but [spɑt] and [stɑp]

Unaspirated when preceded by [s]

Notes for Standardized Speech Assessment The aspiration or lack of aspiration will not be noticed except if it is excessive.

Devoicing: b, d, g

When in final position and followed by silence or at the end of an utterance, these phonemes become devoiced.

The dog ran. [ðə dɑg ɹæn] but: Pointing to a picture the child says “dog” [dɑk]

Standardized speech assessments are single words and the elicited words are followed by a pause. Therefore, devoicing of final [b, d, g] in which they are produced as [p, t, k] could be quite normal and possibly not considered an error. Check the guidelines of the test.

Unreleased Stops: p, b, t, d, k, g

The first plosive is unreleased when (1) followed by another stop or by an affricate,

cooked [kUk t]

Unreleased stops may sound like an omission. Check the context.

(2) [p, b] are followed by /m/ or (3) [t, d] are followed by /n/.

Help me [hElp mi]

Glottal Stop Replacement: t

The glottal stop [ʔ] replaces [t] when followed by a syllabic [n].

button [bʔn]

Tap Replacement: t, d

The tap [ɾ] replaces [t, d] when they are between vowels.

ladder [læɾ2]

18

that chair [ðæt tSE2]

He had nine [hi hæd naIn] kitten [kIʔn] cotton [kɑʔn]

lady bug [leɾi bg] Betty [bEɾi]

May sound like a sound omission. Words with these structures are often on standardized speech assessments. The glottal stop replacement would possibly not be considered an error. Check the guidelines of the test. Words with these structures are often on speech tests. The tap replacement would possibly not be considered an error. Check the guidelines of the test.

Table 2–3.  continued Allophonic Variation

Changes

Examples

Devoicing Fricatives and Affricate: v, z, ð, Z, dZ

Devoicing occurs in final position when it is followed by a pause or at the end of an utterance.

He goes. [hi gos]

Light and Dark /l/

There are two variants of /l/. The “light” production is alveolar and occurs at the beginning of words, followed by a vowel. The “dark” /l/ is produced in any other position or when followed by a consonant or pause.

“lion” – light /l/

[l] is devoiced when following a voiceless stop in an initial stressed position.

“clown” – voiceless [l]

Devoicing: /l/

Pointing to “cheese” [tSis] He has a grudge [hi hæz ə gɹtS]

“blue” – dark /l/ “feel” – dark /l/ “glasses” – dark /l/

“plate” – voiceless [l] “slide” – voiced [l]

Devoicing: /ɹ, j, w/

/ɹ, j, w/ are devoiced following a voiceless stop in initial stressed position.

“train” – voiceless /ɹ/ “cute” – voiceless /j/ “twin” – voiceless /w/

Notes for Standardized Speech Assessment Standardized speech assessments are typically single words and the elicited words are followed by a pause. Therefore, devoicing of final voiced fricatives and affricates could be quite normal. Check the guidelines of the test. Although these [l] variants sound somewhat different, they will be perceived as /l/. Children often have more difficulty with dark [l] productions, thus /l/ at the beginning of words is easier than at the end.

There could be several words on standardized assessments that represent a voiceless stop + [l]. This may not be counted as an error. These clusters are often noted on standardized assessments. The devoicing could possibly not be considered an error. Check the guidelines of the test.

Aspiration:  Superscript h following stop [ph] Unreleased Stops:   following the stop [lIp], the stop does not have the release phase. Glottal Stop:  [ʔ], a stoppage of air at the vocal folds followed by a sudden release. Tap: [ɾ], articulated with a single tap of the tongue tip against the alveolar ridge. Source:  Adapted from Facultad de Filologia, Universidad de Sevilla (2011).

19

Table 2–4.  Commonly Used Diacritics Description of Production

Description of Symbol

Dentalized [s] and [z]

The tongue tip is too far forward when producing [s] and [z]. It does not sound like [θ], [ð] but sounds somewhere in between.

A small three-corned square without the bottom horizontal line of the square.

Dentalized [s] = [ s ]

Palatalized [s] and [z]

The tongue tip is too far back when producing [s] and [z]. It does not sound like [ ʃ ] or [Z] but approaches [ ʃ ] or [Z] perceptually.

An elevated j (  ) placed directly after the [s] or [z] symbol.

Palatalized [s] = [s]

Lateralized [s] and [z]

The lateralized “s” is produced with the tongue tip actually in contact with the alveolar ridge while the lateral edges, which are normally raised for [s], are either lowered on one side (a unilateral lisp) or on both sides (bilateral lisp).

The symbol for the lateralized [s] is not a diacritic but rather a completely different symbol. This lateral production is considered a primary articulation, thus enough of the features are changed that the resulting production is a different phoneme.

Lateral [s] = [  ]

Children often produce [ ʃ ] and [Z] without their characteristic lip rounding. Although the phonemic value of these sounds is preserved, as a clinician we perceive these sounds as being different than a typical [ ʃ ] or [Z]. This unrounding may also carry over to the affricates [t ʃ ] and [dZ].

The diacritic used for the loss of lip rounding is a twoended arrow placed below the sound in question.

Unrounded production of [ ʃ ] = [ ʃ  ]

Diacritic

Symbol

Diacritics For Deviant S-Productions

Unrounded Productions

20

Dentalized [z] = [ z ].

Palatalized [z] = [z]

Lateral [z] = [].

͍

Unrounded production of [Z] = [Z].

͍

Table 2–4.  continued

Diacritic Unreleased Productions

Description of Production

Description of Symbol

At the end of a word or utterance or before a pause, a child may produce the closure between the articulators for the stop, but not the release phase.

The symbol for unreleased is a small corner [  ] appearing directly after the symbol.

If the word “cat” is articulated with an unreleased [t], the transcription would be [kæt].

The symbol that was noted for the dentalized [s], [ s ], is placed above and below the symbol.

Interdentalized  [t],  [d], [n], and  [l] = [t], [d], [n ], and [ l ].

Symbol

This lack of release may sound as if the child has deleted the stop. Interdentalized Alveolar Sounds

Younger children may go through a phase where specific sounds are produced with the tongue tip too far forward. This occurs primarily on the alveolar sounds such as [t], [d], [n], and [l]. It is clearly noticeable as the tongue tip is visible on the production of these sounds.

21

22

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

A relational analysis compares the child’s sound productions to the adult target. Therefore, it is a measure of the correctness of a child’s speech sound production. When a clinician gives a child an articulation or phonology test, a relational analysis is typically the outcome, that is, the result is an overview of correct and incorrect sounds.

Organizing Your Analysis There are many different ways to organize your analysis. For both independent and relational analyses these parameters could be used: Phonetic:  The speech sounds that are evident in the child’s system Phonemic:  The phonemes that are used contrastively by the child Phonotactic:  The structure of words (word shapes) and syllables (syllable shapes) that are demonstrated. Prosodic:  Suprasegmental patterns noted (Velleman, 2016) We will analyze data collected from a small speech sample produced by a child, Andy, age 3 years 4 months. Andy was considered unintelligible and did not respond to attempts using a standardized picture naming test. Therefore, each of these four organizational methods will be demonstrated using Andy as an example.

Phonetic Inventory This is the inventory of the speech sounds, the phones that the child uses regardless if they are accurate or not. It is the phonetic transcription and any diacritics that might be necessary to document the child’s speech. It may also include the distribution of these sounds, i.e., where each of the sounds occurs in relationship to the word or utterance. We are analyzing the distribution of speech sounds when we state that a child produced a specific sound at the beginning of a word, for example, but not at the end. The sounds at the beginning of the word are labelled initial, word-initial or initiating sounds, while those at the end of a syllable are final, word-final or terminating sounds. For the purpose at hand, any sounds that occur somewhere within the word are considered medial. Diacritics used to characterize a specific transcription will also be included in this inventory. For example, a child might produce [s] as well as [s ], a dentalized [s] production and [] a lateralized s-production. All three of these would be included in the phonetic inventory. Table 2–5 is a form that could be used to list the phonetic inventory. For each of the sounds, a + mark could indicate that the sound is present in the child’s inventory. Sound substitutions are not listed, only those sounds that the child used in the utterances. Note that on Table 2–5, the phone [Z] is not present. Because it is so infrequent in American English, it has been deleted from the inventory form. Table 2–6 is a speech sample from Andy. Table 2–7 is the inventory form filled out for him. According to Table 2–7, Andy’s phonetic inventory contains a large percentage of stops, nasals, the glide [w], and the fricative [h]. All other fricatives and affricates are not in his inventory and no consonant clusters are evidenced. His vowel inventory seems to be fairly complete although he shows no central vowels with r-coloring or rhotic diphthongs.

Table 2–5.  Inventory and Distribution of Phones Put a + in each box corresponding to the phone that is used in the sample. Substitutions are not noted, only those phones noted in the speech sample. Early Sounds

I

M

Later Sounds

F

Stops

Fricatives

p

f

b

v

t

s

d

z

k

θ

g

ð

Nasals

ʃ

m

Liquids

n

ɹ

ŋ

------

I

M

F

------1

l

Glides

Affricates

w

------

tʃ

j

------

dZ

Fricative h

------

1

“r” in word-final position will be transcribed as a rhotic vowel. See vowel sounds.

List any consonant clusters that are seen in the speech sample. Consonant Clusters: Initial Medial Final







Vowels:  Put a + below the vowel sound if present in the inventory: i

/@

I

eI

/2

E

aI



aU

u

U

oU

ɑ/ɔ/a

ɔI

23

24

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 2–6.  Andy, Age 3;4, Speech Sample Results fish

[It]

go

[doU]

fork

[oU@t]

ball

[bɑ]

spoon

[pun]

soap

[toUp]

up

[]

red

[wEd]

thumb

[tm]

mouse

[maUt]

leaf

[wip]

yellow

[E.woU]

one

[wn]

nose

[noUt]

house

[haUt]

two

[tu]

chin

[In]

cow

[taU]

three

[ti]

love

[wp]

vacuum

[.tu]

wagon

[.d@n]

shoe

[tu]

that

[t]

kiss

[kIt]

jeep

[dip]

throw

[toU]

doggie

[ɑ.ti]

little

[It@]

zoo

[tu]

toe

[toU]

baby

[eI.bi]

cat

[t]

jumping

[.pIŋ]

mommy

[ɑ.mi]

monkey

[.ti]

going

[oU.Iŋ]

daddy

[.di]

my

[maI]

Phonemic Inventory This is the inventory of phonemes, those units that the child uses to contrast meaning. For the phonemic inventory the virgules / / are used. Under normal circumstances, the phonemic inventory does not include the diacritics that are used to give detail to a specific production. For example, the [sj], a palatalized production of /s/ would be listed as /s/ not as /sj/. As clinicians, you may have learned the dichotomy between broad transcription and narrow transcription. Although narrow transcription reflects the information for a phonetic inventory, broad is used for a phonemic inventory. One can usually assume that the phonemic inventory can be derived from the data using a speech sample or some form of naming task. If the child substitutes [t] for [k] in all given situations and never uses [k], then we would probably conclude that /t/ is in the child’s phonemic inventory but not /k/. This is typically how clinicians determine the phonemic inventory of a child. They look at the sound substitutions and omissions and use that data to determine a phonemic inventory. For example, a child substitutes [t] for [k], [s], [ ʃ ], [θ], and [tʃ ]. If these target sounds are not noted in other contexts, then it could be concluded that the child has /t/ in her/his phonemic inventory, but not /k/, /s/, /ʃ/, /θ/, or /tʃ/. This would indicate a collapse of phonemic contrasts, that is, specific contrasts that typically occur in the phonemic system of American English are not realized. The collapse of phonemic contrasts is a central concept for specific types of phonological treatment protocols, for example, Multiple Oppositions (see Chapter 4) uses this information to establish treatment targets. However, substitutions alone are not enough to determine the collapse of phonemic contrasts. The original definition states: the phoneme is the smallest linguistic unit, which when combined with other such units, contrastively distinguishes meaning. With

Table 2–7.  Phonetic Inventory for Andy Early Sounds

I

M

Later Sounds

F

Stops

M

F

Fricatives

p

+

+

b

+

+

t

++++ ++++ +++

++++

++++ ++

s

d

++

++

+

z

k

+

++++

f v

θ

g

ð

Nasals

ʃ +

Liquids

+

++++

ɹ

------

++

l

m

++

n ŋ

+

Glides w

I

------1

Affricates ++++

+

j

------

tʃ

------

dZ

Fricative h

+

------

1

See vowels.

Consonant Clusters: Initial Medial Final none noted











Vowels:  Put a + below the vowel sound if present in the inventory: i

I

eI

E



u

++++++ ++

++++++

+

++

+++++

+++++

/@ ++++++++ +

/2

aI

aU

+

+++

U

oU

ɑ/ɔ/a

++++++ ++

+++

ɔI

25

26

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

this definition in mind, we might need to use minimal pairs to see if the child is able to produce specific phonemes contrastively. To this end, Appendix 2–2 is a list of minimal pair words which are ordered according to vowels. The words are all CV or CVC words which the clinician can choose from based on the needs of a specific child. One could find pictures to match the words or just use imitation to see if the child does produce phonemic contrasts. If the child does use two phonemes contrastively, those would be a portion of the child’s phonemic inventory. Appendix 2–3 includes minimal pair words that are ordered according to frequently occurring substitutions and phonological processes. Table 2–8 is a short cut method to narrow down the possibilities for phonemically contrastive units. It looks at homonymy and substitution patterns to establish which phonemic contrasts to further assess. The data from Andy is in Table 2–9. Table 2–8.  Phonemic Inventory:  Contrasts Used 1. Look through your sample. Do you see two (or more) utterances that are pronounced the same? 2. If yes, list those utterances. Utterance

Words that it is used for

Phonemes not Contrasted





















Summary:  Phonemes that do not appear to be used contrastively

3. List all other targets and substitutions. Target → Substitution →

Target → Substitution →

Target → Substitution →

→



→



→

→



→



→

→



→



→

4. Summarize targets and substitutions. Check with minimal pairs.

2  Ways to Collect and Analyze Data

27

Table 2–9.  Phonemic Inventory:  Contrasts for Andy 1. Look through your sample. Do you see two (or more) utterances that are pronounced the same?  Yes 2. If yes, list those utterances. Utterance

Words that it is used for

Phonemes not Contrasted

[tu]



shoe, two, zoo



/t, ʃ, z/

[toU]



toe, throw



/t, θ/

[tt]



that, cat



/ð, k/









Summary:  Phonemes that do not appear to be used contrastively /t/ vs. /ʃ, z, θ, k, ð/

3. List all other targets and substitutions. Target → Substitution

Target → Substitution

Target → Substitution

f

→

p



v

→

p



ɹ

→

w

s

→

t



g

→

d



l

→

w

dZ

→

d



→



→



→



→

→

4. Summarize targets and substitutions. Check with minimal pairs. /ʃ, θ, z, k, s, ð/ → /t/ /ɹ, l/ → /w/ /dZ, g/ → /d/ /f, v/ → p

According to Table 2–9, Andy has several phonemic contrasts that are not observed. For example, /t/ is used for six different phonemes / ʃ, θ, z, k, s, ð /, while /w/, in addition to being used correctly, replaces both /ɹ/ and /l/. Finally, /p/ is used for /f/ and /v/ while /d/ replaces /g/ and /dZ/. These oppositions could be checked with minimal pairs. Andy seems to use a variety of vowels accurately.

Phonotactic Inventory The types of syllables produced by the child is the basis of the phonotactic inventory. An overview of the components of a syllable is contained in Figure 2–2.

28

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Nucleus = typically a vowel, peak of syllable Onset = all units before the nucleus, Coda = all units after the nucleus, Rime1 = Nucleus + Coda One-Syllable Word CV Shape

One-Syllable Word CVC Shape

(onset + nucleus)

(onset + nucleus + coda)

bye [baɪ] Open Syllable

Onset

[ b]

cat [kӕt] Closed Syllable

Rime

Onset

Nucleus

[aɪ]

Nucleus [k]

Coda

[ӕ]

[t]

Two-Syllable Word Shape CCVCVC (consonant cluster as onset + nucleus/ consonant as onset + nucleus + consonant as coda) skipping [skɪ.pɪŋ]

Rime

Onset

Nucleus

[sk]

[ɪ]

Onset Nucleus

[p]

[ ɪ]

Coda

[ŋ]

1 Rime

is a linguistic term which refers to the nucleus together with the coda (Loos, Anderson, Day, Jordan, & Wingate (2013).

Figure 2–2.  Examples of syllable structures in American English: Consonants (C) and Vowels (V).

The phonotactic inventory is especially valuable for younger children who demonstrate a restricted phonemic inventory and/or for the clinician who notes that there is a restricted range of differing syllable types. Typically, the phonotactic inventory includes: (1) the number of syllables in a word, (2) whether a vowel nucleus is noted, (3) the child’s

2  Ways to Collect and Analyze Data

29

production of primarily open versus closed syllables, and (4) the presence, or absence, of consonant clusters (Velleman, 2016). It also notes any constraints that you might see in the child’s productions. For example, the child might delete all final consonants or demonstrate use of a consonant only in specific contexts. The following explanation is seen as a guideline to Table 2–10, which could be used to chart the phonotactics of a child. The number of syllables examines how many syllables the child uses in words and also whether there are certain syllable counts that are more frequent. For example, a child with a limited speech sound inventory might produce fewer syllables in a word, or they may produce multisyllabic words but only very infrequently. The syllable shape of the individual utterances would also be a portion of the phonotactics. Syllables can vary from a single vowel (V), or consonant (C), to multisyllabic words, such as “umbrella” pronounced [@.bwE.w@], (V1C1C2V2C2V1). A second variable is the presence of a vowel as the nucleus of a syllable. Most syllables contain a vowel as their center or nucleus. There are exceptions; the syllabic consonants can serve as the nucleus of the syllable in words such as the casual pronunciation of “bottle” [bɑtl͎ ] or “little” [lItl͎ ]. Children with limited phonotactic skills may produce a consonant as the nucleus of a syllable. For example, the child says [m] for “mom,” or [m.m] for “mama.” Open versus closed syllables denotes a syllable structure. Open is a syllable with a vowel (V) in the word-final position, whereas closed syllables contain a consonant (C) as the final element. Children with very limited speech sound inventories have a tendency to use open syllables. They might have two-syllable words but delete the consonants at the end of the syllables. In this respect “basket” may be [bæ.k@] or, “vacuum” [væ.u]. The use of consonant clusters may also be very limited in children with restricted phonemic/phonetic inventories. As clinicians, we typically do not expect consonant clusters until a later age. A different picture is seen if we separate occurrence versus mastery. Many clusters are indeed not mastered until very late. McLeod and Arciuli (2009) state that specific clusters are still being mastered beyond 9 years of age. However, in normally developing children, clusters are present in 2-year-olds, although they may show substitutions such as [pw] for [pl] or [pr] (McLeod, van Doorn, & Reed, 2001). It is important to note whether a child uses consonant clusters in the word-initial or -final position. Although early clusters frequently occur initially, other clusters, such as –nt or –nd, are noted early in word-final positions. The use of constraints is also a portion of our phonotactic analysis. In this case, constraints are any patterns noted that seem to limit or restrict the productional possibilities of our clients (Blache, 2000). Children with limited inventories are already demonstrating specific speech sound/phonemic constraints. Other constraints could be noted in the number of syllables; for example, the child who possibly uses only one-syllable words. Here the child’s constraint would be when she/he says any word, it must be one syllable. Or there could be a constraint on the syllable shape; for example, the child might use CVCV for two-syllable words regardless of the original syllable shape. In this instance, “flowers” would be uttered as [fɑ.w@] or “wagon” as [wæ.d@]. Now the constraint is that the child can produce two-syllables but the syllable shape must be CVCV. This is a restriction to the child’s system. Also certain vowels or consonants might be used only in

Table 2–10.  Phonotactic Inventory Form Client’s Name: ___________________________ CA: ______________ Utterance

30

Number of Syllables

Word Shape

Vowel as Nucleus

Open vs. Closed

Consonant Cluster

2  Ways to Collect and Analyze Data

31

Table 2–10.  continued Number of Syllables

Utterance

1= 2= 3+=

Word Shape

Most frequent type?

Vowel as Nucleus

Open vs. Closed

Yes percentage =

Open percentage, Closed percentage

Consonant Cluster

List:

Examples of Word Shapes: C1 [m]

V1 [ɑ] (could also include glottal stop + vowel [ʔɑ] or vowel + glottal stop [Vʔ]

C1V1 [mɑ]

V1C1 [p]

C1V1C1V2 [mɑ.mi]

C1V1C1V1 [mɑ.mɑ]

C1V1C1 [pɑp]

C1V1C2V2 [dɑ.gi]

C1V1C2 [kt]

V1C1V2 [oUkeI]

C1C2V1 [pweI]

C1V1C2C3 [wEnt]

C1C2V1C2V2 [bwE.w@] for umbrella

C1C2V1C2V2C3 [bwE.w@s] for umbrellas

Noted Constraints: __________________________________________________________________ __________________________________________________________________ __________________________________________________________________

certain contexts. Thus, [k] could be used at the beginning of a word but not at the end: “cat” would be [kæt] but “cake” [keI]. Positional limitations when producing consonant clusters could also be a constraint. For example, consonant clusters are used occasionally by a child at the end of a word, but never at the beginning: “clown” would be [kaUn] but “trains” [teInt]. Constraints can vary, therefore, the clinician will have to look at the transcription and see if any patterns of restrictions or limitations are specifically noted. Table 2–11 is a Phonotactic Inventory Form filled out for Andy. The Phonotactic Inventory Form (see Table 2–10) might be a good way to collect data if you are planning on using Phonotactic Therapy such as that noted in Chapter 7.

Table 2–11.  Phonotactic Inventory Form for Andy Client’s Name: Andy Utterance

CA: 3;4

Number of Syllables

Syllable Shape

Vowel as Nucleus

Open vs. Closed

Consonant Cluster

fish [It]

1

VC

Yes

Closed

No

ball [bɑ]

1

CV

Yes

Open

No

up []

1

V

Yes

Open

No

mouse [maUt]

1

CVC

Yes

Closed

No

one [wn]

1

CVC

Yes

Closed

No

two [tu]

1

CV

Yes

Open

No

three [ti]

1

CV

Yes

Open

No

wagon [.d@n]

2

VCVC

Yes

Open + Closed

No

kiss [kIt]

1

CVC

Yes

Closed

No

doggie [ɑ.ti]

2

VCV

Yes

Open + Open

No

toe [toU]

1

CV

Yes

Open

No

jumping [.pIŋ]

2

VCVC

Yes

Open + Closed

No

going [oU.Iŋ]

2

VVC

Yes

Open + Closed

No

go [doU]

1

CV

Yes

Open

No

spoon [pun]

1

CVC

Yes

Closed

No

red [wEd]

1

CVC

Yes

Closed

No

leaf [wip]

1

CVC

Yes

Closed

No

nose [noUt]

1

CVC

Yes

Closed

No

chin [In]

1

VC

Yes

Closed

No

love [wp]

1

CVC

Yes

Closed

No

shoe [tu]

1

CV

Yes

Open

No

jeep [dip]

1

CVC

Yes

Closed

No

little [I.t@]

2

VCV

Yes

Open + Open

No

baby [eI.bi]

2

VCV

Yes

Open + Open

No

mommy [ɑ.mi]

2

VCV

Yes

Open + Open

No

daddy [.di]

2

VCV

Yes

Open + Open

No

fork [oUt]

1

CVC

Yes

Closed

No

soap [toUp]

1

CVC

Yes

Closed

No

thumb [tm]

1

CVC

Yes

Closed

No

yellow [E.woU]

2

VCV

Yes

+ Open

No

32

2  Ways to Collect and Analyze Data

33

Table 2–11.  continued

Utterance

Number of Syllables

Syllable Shape

Vowel as Nucleus

Open vs. Closed

Consonant Cluster

house [haUt]

1

CVC

Yes

Closed

No

cow [taU]

1

CV

Yes

Open

No

vacuum [.tu]

2

VCV

Yes

Open + Open

No

that [t]

1

CV

Yes

Open

No

throw [toU]

1

CV

Yes

Open

No

zoo [tu]

1

CV

Yes

Open

No

cat [t]

1

CV

Yes

Open

No

monkey [.ti]

2

VCV

Yes

Open + Open

No

my [maI]

1

CV

Yes

Open

No

Most frequent types? CV, CVC, VCV

Yes % = 100

Open percentage = Approx. 2/3 were open Closed percentage =

List: None

1 = 28 2 = 11 3+ = 0 Average = 1.28 Noted Constraints:

Number of syllables.  Primarily one-syllable words. No three-syllable words. Syllable shape.  Primarily C1V1, C1V1C1, and V1C1V2. Syllable shape and number of syllables interact, two-syllable shapes are reduced from C1V1(C2)C3V2 ([dZmpIŋ] “jumping” to V1C1V2C2 ([.pIŋ]. Positional constraint on [k].  [k] appears at the beginning of one-syllable words but is deleted, initiating the second syllable of two-syllable words. At the end of one-syllable words, a substitution for [k] is used. Positional constraint of [f].  [f] is deleted at the beginning of a word. At the end of a word, Andy uses a substitution.

A summary of Andy’s phonotactic skills: Number of Syllables:  They are primarily one-syllable words. Only one-third of the utterances were two-syllable in length. There were no three-syllable words. Syllable shape:  They are primarily CV, CVC, and VCV. Two-syllable shapes are reduced from CV(C)CV to VCV or VCVC. Thus, “ba-by” becomes [eI.bi] and “jumping” becomes [.pIŋ]. Vowel as Nucleus:  Always present Open vs. Closed:  Approximately 2/3 of the utterances are open syllables.

34

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Consonant Clusters:  No clusters noted Constraints:  (1) Number of syllables, only one or two; (2) Reduction of syllable structure in two-syllable utterances; (3). Positional constraint [k], can appear at the beginning of a one-syllable word, however, it is produced as [t] at the end of one-syllable words. The [k] is deleted at the beginning of the second syllable of two-syllable words. Positional constraint [f], deleted word-initially, [p] is used as a substitution word-finally. According to Flipsen (2006), the average number of syllables per word is 1.26. at 3 years of age. Andy appears to be at or slightly above that level. However, by 3;0 years of age, children typically have a range of consonant clusters both word-initially and -finally (McLeod et al., 2001). Two-syllable words with consonant clusters are also present in 3- to 4-year-olds (Shriberg, 1993).

Prosodic Patterns Prosodic, or suprasegmental, features include stress (loudness), pitch (intonation), and durational (rate) variations that occur across segments. Another term used within the realm of prosody includes rhythm. Rhythm refers to how stressed and unstressed syllables are distributed over time. Occasionally, we notice that children with speech sound difficulties also evidence prosody that is noticeably off, or perhaps, stereotypical; for example, a child who stresses words only on the first syllable, or one who demonstrates equal and even stress on every syllable. Therefore, it is important that we are able to document the prosodic patterns that are noticed. Prosodic development begins at a very early age and includes both perceptual and productional aspects. Before their first words, children are able to perceive stress differences in words. At the beginning of the jargon stage of babbling (around 10 months of age), strings of babbles modulated by intonation, rhythm, and pausing are being produced. An important aspect of the first 50-word stage (approximately 12 to 18 months) is prosodic variation. Children use pitch variation to indicate differences in meaning between utterances. For example, a rising pitch at the end of [dædæ] may indicate something quite different contextually than a falling pitch [dædæ]. It appears that intonation develops before stress, and the production of contrastive stress is noted around the time the child begins two-word utterances. Now the stress of the individual units signals differences in meaning. Thus, [dædæ goU] may indicate something quite different than [dædæ goU] (Crystal, 1986). Thus, adult-like intonation patterns are noted prior to the appearance of the first words, whereas the onset of stress patterns occurs around age 2. However, the prosodic system takes many years to fully develop. Some aspects are still developing until children are at least 12 years old. There are very few standardized tests available for examining receptive and expressive prosodic abilities in children. Box 2–1 is a composite of those tests that are available for children. Although none of them has robust normative data, if you would like to examine a child’s prosody in more detail, one of these could be used. Please note that there are other tests available that indirectly examine prosody, but these are more directed at examining the emotional aspects of prosody, for example, as conveyed through facial expressions.

2  Ways to Collect and Analyze Data

35

Box 2–1.  Instruments for Assessing Prosody in Children Author(s), Year

Age Range

PROP Prosodic Profile

Crystal, 1982

Children and adults

Obtains information on expressive prosodic patterns

No normative data.

PVSP ProsodyVoice Screening Profile

Shriberg, Kwiatkowski, Rasmussen, Lof, & Miller, 1990

3 to 19

Assesses prosody and voice in conversational speech

Normed on 252 individuals, ages 3 to 19.

Prosody-Voice Profile

Shriberg, 1993

3 to 6

A perceptually based assessment of voice and prosody based on a sample of conversational speech.

Validated on 62 children with delayed speech and 13 with suspected apraxia of speech.

PEPS-C Profiling Elements of Prosody in Speech Communication

Peppé & McCann, 2003, Peppé, 2015

4 to14

Assesses receptive and expressive prosodic skills

Normed on 120 children, ages 5 to 14.

PPAT Perception of Prosody Assessment

Klieve, 1998

7 to12

Evaluates prosodic perception in children

Normed on 6 children, ages 7 to 12.

Name of Test

Purpose

Normative Data

Informally, prosodic features can be noted based on a speech sample or by using probes. Appendix 2–4 offers one informal way to screen prosodic features in children. It is based on production of intonation, stress, and duration, and not on the comprehension of these features. If you notice that a particular child is having difficulty producing one or more of these probes, then further assessment with a formalized test would be advised.

Using a Standardized Speech Assessment for Your Data Collection Standardized speech tests answer questions about the phonetic inventory. As noted in the previous section, based on the error patterns we can assume certain things about the phonemic inventory, however, once again minimal pairs could be used to be certain that the child is using specific phonemes contrastively. This relational analysis usually

36

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

contains information about the distribution of speech sounds as well. Thus the assessment may test specific sounds in initial-, medial- and final-word positions. For additional information, probes could be used to further test certain sounds. Probes for word-initial and -final single syllable words are contained in Appendix 2–5. One way of maximizing the results of a speech assessment is to examine all the words produced by the child that demonstrate an error on the test. Thus, those words that demonstrate an error are transcribed, the rest are left blank indicating that a correct production was heard. This will possibly increase the number of words with common sounds such as [s] or [ɹ], [2] or []. Also word-final consonant clusters, which are infrequently pinpointed on a standardized assessment, can be noted. A matrix for recording errors is usually provided as a portion of the speech assessment. Table 2–12 is a slightly modified form of the one seen in the previous section (see Table 2–5, page 23), Table 2–12 could also be used if you are working from a speech sample. The form allows for word-initial, -medial, and -final phones and consonant clusters. Table 2–13 demonstrates this form filled out for Andy. Andy demonstrates a widespread substitution of [t] for several sounds, [k, g, s, z, θ, ð, ʃ]; [d] is used inconsistently for [g] and [dZ], [p] for both [f] and [v], and finally [w] for [ɹ] and [l]. Andy deletes initial and final consonants. Consonant clusters are always reduced to one element and often this is a substitution, for example [t] for [k]. The substitutions in this matrix are similar as those noted in his phonemic inventory (page 27). However, this matrix does not show the influence the number of syllables has on the production of initial consonants, as noted in the Phonotactic Inventory. Therefore, for Andy the Phonotactic Inventory gives us more information about the patterns he demonstrates.

Analyzing Errors Place-Manner-Voice Analysis This way of analyzing errors compares the target and the error sounds according to a broad-based phonetic categorization: articulatory features of place, manner and voicing. Each sound of American English can be classified according to these three articulatory features, which then allows for a comparison of productions. The end result is to see if specific patterns emerge. Patterns would be defined by a frequent use of one or more of these three articulatory parameters when target and error sounds are compared. This method of analysis can only be used for a child’s substitutions. It does not account for distortions, deletions, or possible additions. Figure 2–3 is one way the sounds of American English can be categorized according to these three parameters. Let’s look at a couple of examples. If we analyze [s] according to place, manner, and voicing parameters, we find that the place is alveolar, manner is fricative, and it is a voiceless sound. We have noted that our child substitutes a [t] for [s], not an uncommon error. This substitution would have the same place, alveolar, however, the manner is a stop (as compared to the fricative manner for [s]). The [t] is comparable in voicing,

Table 2–12.  Form for Recording Relational Analysis:  Inventory and Distribution of Phones Correct production: 3    Substitution: Write in the sound    Deletion:  Early Sounds

I

M

Later Sounds

F

Stops

Fricatives

p

f

b

v

t

s

d

z

k

θ

g

ð

Nasals

ʃ

m

Liquids

n

ɹ

ŋ

------

I

M

F

l

Glides

Affricates

w

------

tʃ

j

------

dZ

Fricative h

------

List any consonant clusters that are noted, including substitutions or deletions of any sounds within the cluster. Consonant Clusters: Initial Medial Final











37

Table 2–13.  Relational Analysis for Andy, age 3;4 Correct production: 3    Substitution: Write in the sound    Deletion:  Early Sounds

I

M

Later Sounds

F

Stops

I

M

F

Fricatives f

, 

p

v



p

, 

s

t

t, t, t

3

3

z

t

t

3, t, t

t, t

t

θ

t, t, t

, d

d, t

ð

t

ʃ

t

p

3

3

b

3, 

3

t

3, 3

3

d

, 

k g

, 3, 3

Nasals 3, 

Liquids

3

3333

ɹ

w

-----

33

l

w, w, 

m

3, , 

n ŋ

3

Glides

t

w

, 

Affricates

w

3, 

-----

tʃ



j



-----

dZ

, d

3

-----

Fricative h

List any consonant clusters that are noted, including substitutions or deletions. Consonant Clusters: Initial Medial Final θɹ → t (2×) mp → p sp → p



kj → t





ŋk → t







For the consonant clusters, the resulting articulations are also counted within Table 2–13. For example, production of [p] for [mp] and [sp] are both noted under [p] and the substitutions of [t] for [θ] and [k], ([θɹ], [kj], and [ŋk]) are noted under the respective sounds.

38

39

2  Ways to Collect and Analyze Data

           

Sonorant

 

PLACE Labial Coronal Dorsal MANNER Bilabial Labiodental Interdental Alveolar Postalveolar Palatal Velar Glottal b t Stop p         d           k   ɡ       Fricative f v s z θ   ð   ʃ   ʒ           h*   Affricate ʧ   ʤ                        

Nasal*

m

Liquid* Glide*

w

 

 

 

 

n

 

 

 

 

 

 

 

 

 

l

 

 

 

 

 

 

 

 

 

 

 

ɹ  

  *Sonorant    

j  

 

ŋ  

 

 

 

 

 

 

 

 

Shaded area indicates a voiced sound.

Figure 2–3.  Place-manner-voicing descriptive features: American English consonants.  

it is also a voiceless sound. This same child says [d] for [z], a comparable example as [t]/[s], except both [d] and [z] are voiced sounds. This same child also substitutes [d] for [n] and [l]. Now a pattern begins to emerge. [d] = alveolar, stop, voiced [z] = alveolar, fricative, voiced [n] = alveolar, nasal, voiced [l] = alveolar, liquid, voiced This child substitutes a voiced alveolar stop for voiced alveolar fricatives, nasals and liquids. Thus, this child seems to have difficulty with manner features. These placemanner-voicing features are used to select targets in specific types of therapies. Their role will be specifically demonstrated in several of the later chapters. A summary of Andy’s substitutions using Place-Manner-Voicing Analysis demonstrates: [t]/[k, g, s, z, θ, ð, ʃ] Patterns:  Place, manner and voicing are affected. Both place and manner are changed five times each. However, to a lesser extent, voicing changes are also present (from voiced to voiceless). [d]/[g, dZ] Patterns:  Place and manner are changed. Voicing remains constant with all substitutions. [p]/[f, v] Patterns:  Place, manner, and voicing are changed. [w]/[ɹ, l] Patterns:  Place and manner substitutions could be noted. Voicing remains relatively constant. Andy demonstrates changes along all three parameters:  Place → alveolar, Manner → stops, Voicing → voiceless sounds.

40

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Phonological Processes Several standardized tests presently describe error patterns with phonological processes. Many clinicians feel that it gives them a good overview of the specific child’s patterns of errors and difficulties. And as phonological processes are used to determine targets in some of the phonological treatment protocols, a brief review seems in order. There are typically three categories of phonological processes: syllable structure, substitution, and assimilation processes. Syllable structure processes describe those sound changes that affect the structure of the syllable. Substitution processes describe those sound changes in which one sound is replaced by another. It may affect just one sound, as in [wIt] for “with” or a whole class of sounds, for example, all fricatives are replaced by stops. Assimilation processes, or sometimes called harmony processes, describe changes in which a sound becomes similar to, or is influenced by, a neighboring sound within an utterance. Many assimilation processes are quite normal and occur frequently in the speech of children and adults. Table 2–14 provides an overview of each process with examples. There are assimilations that may occur on assessment measures. If it is an inconsistent production, which seems to occur only on specific words, then it is might not be a sound error. Box 2–2 gives a few of these examples. If you are using phonological processes to establish goals, then it is important that you note which processes occur most frequently. For specific types of therapy it might also be helpful to know the sounds affected by the process. Table 2–15 is an example that could be used for this purpose. Table 2–16 is a summary table that has been filled out for Andy. Andy demonstrates a high frequency of initial consonant deletion, fronting, and stopping. It is interesting to note that Andy’s phonotactic summary gives us a good explanation for many of his occurrences of initial consonant deletion. In two syllable words, he consistently deletes the initial consonant.

Table 2–14.  Overview of Selected Phonological Processes Phonological Process

Overview

Example

Syllable Structure Processe Syllable Reduction/ Unstressed Syllable Deletion/ Weak Syllable Deletion

Deletion of typically the unstressed syllable

“pajamas” → [dZm@z], “banana” → [nn@]

Cluster Reduction/ Simplification

Simplification of clusters to one- or two-element clusters

Three-element clusters: “street” → [trit] or [tit] Two-element clusters: “blue” → [bu]

Deletion of Initial Consonant

Deletion of the first consonant of a word

“cow” → [aU]

Deletion of Medial Consonant

Deletion of the middle consonant of a word

“rabbit → [ɹ.It]

Deletion of Final Consonant

Deletion of the final consonant of a word

“dog” → [dɑ]

Reduplication

Second syllable becomes a repetition of the first, thus, simplifying the structure of the word.

“water” → [wɑwɑ] Partial reduplication: “water” → [wɑwi]

Fronting: Velar

Substitution is more anteriorly located than the intended sound. Velar refers to velar sounds [k, g] being produced more anteriorly as [t, d].

“walk” → [wɑt] or “key” → [ti]

Fronting: Palatal

Substitution is more anteriorly located than for palatal sounds. Typically [ʃ, Z] are produced as [s, z].

“shoe” → [su]

Fronting: Other

Not common but an example would be [t] is replaced by [p].

“toe” → [poU]

Stopping (Fricatives and Affricates)

Substitution of stops for fricatives or the omission of the fricative portion of affricates.

“zoo” → [du] and “chin” → [tIn]

Other Substitution Processes

continues

41

Table 2–14.  continued Phonological Process

Overview

Example

Substitution Processes continued Stopping: Other, e.g., /m/ → /b/ Deaffrication

Production of affricates as homorganic (homorganic = the same place) fricatives, thus, the fricative is the only portion remaining of the affricate.

“cheese” → [ʃiz]

Affrication

Replacement of fricatives by homorganic affricates

“sheep” → [tSip]

Stridency Deletion

Strident sounds [f, v, s, z, ʃ, Z, tS, dZ] are deleted or substituted with nonstrident consonants.

“see” → [ti] or [i]

Denasalization

Replacement of nasals by homorganic stops

“moon” → [bun]

Gliding of Liquids

Replacement of a liquid [l, ɹ] by a glide sound [w, j]

“rabbit” → [wb@t]

Gliding of Fricatives

Replacement of a fricative by a glide [w, j] sound

“van” → [jn], [v] → [ j ]

Vowelization/ Vocalization also includes Derhotacization

Replacement of liquids [l, ɹ] , rhotic vowels [, 2] or nasals [m, n, ŋ] by vowels

“ladder” → [ld@] or “apple” → [poU]

Backing

Substitution is more posteriorly located than for the intended sound

“soup” → [hup]

Voicing

Replacement of a voiced for a voiceless sound

“soup” → [zup]

Devoicing

Replacement of a voiceless for a voiced sound

“big” → [pIg]

Glottal Replacement

Use of a glottal stop [ʔ] to replace a standard consonant

“hat” → [hʔ]

Liquidization

Replacement of a non-liquid sound by a liquid [l, ɹ]

“this” → [lIs], [ð] → [l]

Other

42

Table 2–14.  continued Phonological Process

Overview

Example

Assimilation Processes Labial Assimilation

Change of a nonlabial into a labial sound due to the influence of a neighboring labial sound

“top” → [bɑp], The final labial [p] influences the beginning [t].

Velar Assimilation

Change of a nonvelar sound into a velar sound due to the influence of a neighboring velar sound

“gone” → [gɑŋ]

Nasal Assimilation

Change of a nonnasal sound into a nasal sound due to the influence of a neighboring nasal sound

“bunny” → [muni]

Liquid Assimilation

Change of a nonliquid sound into a liquid sound due to the influence of a neighboring liquid

“yellow” → [lEloU]

Other

43

44

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Box 2–2.  Assimilations That Are Not Typically Considered Speech Sound Errors Assimilation refers to adaptive articulatory changes in which one sound becomes similar to, or possibly identical to a neighboring sound. Such changes may affect one or more sounds of a word. For example, under most circumstances the final sound in “news” is a [z]. However, in the context of “newspaper” this voiced sound, influenced by the voiceless [p], becomes voiceless, thus [nuspeIp2]. There are several words on standardized speech assessment tests that often result in assimilation processes. These are normal occurrences and should not typically be counted as errors in speech sound production. Several of these were noted in Table 2–3, Allophonic Variations (see pages 18–19). The following are possibilities from standardized speech assessments that could be assimilations. 1. [tɹ] blends are often assimilated to [tS ]:  Children and adults will often produce the [tɹ] blend as [tS ], with the [ɹ] being either partially or totally reduced. Word Examples:  truck, train, tree, tractor, trunk, triangle, trash 2. Two-syllable words with [n] + [t] or [n] + [d] reduced to [n]:  Nasals have a tendency to completely assimilate the stops that are produced at the same place of articulation. This results in a complete loss of the [t] or [d]. Word Examples:  presents, Santa, handbag, handcuff, bandage, kindness, reindeer, center, county, painter, printer, wanted 3. Yellow:  This word belongs in a category by itself. The authors have heard so many children say lelow [lEloU] for “yellow.” This is a regressive assimilation; that is the medial [l] impacts the beginning [ j ]. The child is not having difficulties with [ j ]. Have the child say “yes,” it has to do with this particular word. 4. Assimilations due to word length:  Many children have difficulties with production of 3+ syllable words. They will often shorten the word [dZæm@z] for “pajamas” (deleting the unstressed syllable) or make errors that they normally wouldn’t make if the word were one syllable. Try the sound in a one-syllable context. Word Examples:  helicopter, strawberry, umbrella, elephant, music box, cowboy hat, screwdriver, television, vegetable

Table 2–15.  Summary Sheet for Phonological Processes Syllable Structure Processes

Number of Occurrences AND/OR

Sounds Affected*

Substitution Processes Continued

Syllable Reduction

Gliding of Liquids

Cluster Reduction/ Simplification

Gliding of Fricatives

Deletion of Initial Consonant

Vowelization/ Vocalization also includes Derhotacization

Deletion of Medial Consonant

Backing

Deletion of Final Consonant

Voicing

Reduplication

Devoicing

Other

Glottal Replacement

Substitution Processes

Number of Occurrences AND/OR

Sounds Affected

Liquidization

Velar Fronting

Other

Palatal Fronting

Assimilation Processes

Stopping (Fricatives and Affricates)

Labial Assimilation

Stopping (other, e.g., /m/ → /b/)

Velar Assimilation

Deaffrication

Nasal Assimilation

Affrication

Liquid Assimilation

Stridency Deletion

Other

Denasalization *For some analyses you might want to know the number of occurrences of the process, whereas for others the sounds affected by the process may be important. Therefore, both columns have been established in this table. The clinician can use one or both for the analysis procedure.

45

Table 2–16.  Summary Sheet for Phonological Processes:  Andy

Syllable Structure Processes

Number of Occurrences

Syllable Reduction

Gliding of Liquids

Cluster Reduction/ Simplification

6

Deletion of Initial Consonant

14

Deletion of Medial Consonant Deletion of Final Consonant

Vowelization/ Vocalization also includes Derhotacization

6

Other

Glottal Replacement Liquidization 8

Palatal Fronting

Other Assimilation Processes

15

Labial Assimilation

Stopping (other, e.g., /m/ → /b/)

Velar Assimilation

Deaffrication

Nasal Assimilation

Affrication

Liquid Assimilation

Stridency Deletion (see Stopping)

Other

Denasalization

1

Voicing Devoicing

Stopping (Fricatives and Affricates)

4

Gliding of Fricatives

Reduplication

Velar Fronting

Number of Occurrences

Backing

Substitution Processes

46

Substitution Processes Continued

4

2  Ways to Collect and Analyze Data

47

Measuring Intelligibility For our purposes, intelligibility is a judgement made by a clinician based on how much a child’s speech can be understood. Most measurements of speech intelligibility are rather subjective and are somehow correlated to the percentage of words a listener can understand. However, there are a number of factors that do affect intelligibility, not only the number of errors, but also the type and consistency of errors (Bernthal, Bankson, & Flipsen, 2017). Based on the age of the child, we expect a different degree of intelligibility. There are numerous sources that state how much of a child’s conversational speech should be understood at a particular age. The following guidelines are given based on Bowen (2011) and Coplan and Gleason (1988), but other older and newer references are very similar: 18 months:  Approximately 25% of a child’s speech is intelligible, 24 months:  Approximately 50% of child’s speech is intelligible, 36 months:  Approximately 75% of a child’s speech is intelligible, and by 48 months the child is 100% intelligible, although speech sound errors may be present. As clinicians, we may want to establish treatment goals based on intelligibility rather than targeting specific speech sounds. For example, if we are working on changing the overall phonological system of a highly unintelligible child, which is the case with specific phonological treatment protocols, then an intelligibility goal may be our best option. However, intelligibility measures will need to be considered that are somewhat objective and can be repeated at periodic intervals. For this purpose, there are several types of options, some are more objective than others. For example Fudala and Stegall (2017) offer the following scale: Level 6:  Sound errors are absent or minimally noticed in continuous speech. Level 5:  Speech is intelligible, although noticeably in error. Level 4:  Speech is intelligible with careful listening. Level 3:  Speech intelligibility is difficult, even with careful listening. Level 2:  Speech is usually unintelligible. Level 1:  Speech is unintelligible. Another rating scale that attempts to quantify the parents’ perception of intelligibility is the Intelligibility in Context Scale (McLeod, Harrison, & McCormack, 2012). This is a seven- item scale that is directed toward parents, but could be used by teachers and clinicians to quantify intelligibility. It is available for free in 60 languages from: http:// www.csu.edu.au/research/multilingual-speech/ics

48

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

These types of scales seem to offer a simple solution for rating intelligibility. However, there are several problems inherent in this or any type of scale (Flipsen, 2010). First, different listeners, especially in the mid-range of the scale, will choose different levels, thus, there is a large degree of variability between listeners. A good example, the parents will rank their child more toward the top of the intelligibility scale, they listen to the child every day and have learned how to interpret certain words and phrases. And you, as a clinician, will change your scale ratings based on the same principle. As you listen to this child repeatedly, you also will be able to understand more, even without treatment changes. And with such a scale, how much does it take for the child to move from Level 3: Speech intelligibility is difficult, even with careful listening to Level 4: Speech is intelligible with careful listening. Thus, these scales are not sensitive enough to track change in a child’s speech over time (Flipsen, 2010). Another option is the clinician actually trying to calculate the percentage of words understood. However, first it must be determined at which linguistic level intelligibility will be analyzed: (1) Single words, spontaneous or imitated, (2) Sentences, spontaneous or imitated, or (3) Conversational speech. The degree of unintelligibility will probably dictate the level. For children who are highly unintelligible imitated words or sentences may be ideal. However, for less severe speech sound disorders, conversational speech may provide the best measurement. Ideally, different levels should be sampled. Over the years, many different types of intelligibility probes have been developed. Several of them are out of print or unavailable. The sentences from the Beginner’s Intelligibility Test (Osberger, Robbins, Todd & Riley, 1994) are available at the following link: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3386604/ This test, which was originally designed for testing the intelligibility of hearing-impaired children, can be used for any population. There are four sets of sentences that can be used randomly in an imitative task, or if the child can read, they can be read by the child. The sentences are tape recorded and two independent listeners state how many of the words can be understood. According to Osberger et. al, (1994) a percentage is then calculated based on the number of words the child says (not on the total number of words from the sentences). A slightly different method, supported by Lagerberg, Åsberg, Hartelius and Persson (2014) and Martin (2017), could be easily implemented by clinicians. It uses the syllables that can be understood, versus the total number of syllables to calculate a percentage. Many clinicians may already be doing this, it is an easy way to objectively measure intelligibility. A 10 × 10 grid is used, in which + is used for an understood syllable, while a – is used for an unintelligible one. This could be used as the child looks at a picture (if the child is highly unintelligible, some context may be beneficial). When the grid is full, the number of + marks is added and that is your percentage of intelligibility. See Appendix 2–6 for a grid and some basic instructions.

Percentage of Consonants Correct There have also been a number of methods developed for quantifying intelligibility in children’s conversational speech. The Intelligibility Index (Shriberg. Austin, Lewis,

2  Ways to Collect and Analyze Data

49

McSweeney, & Wilson, 1997) using the Percentage of Consonants Correct, has been used extensively in research and is an objective way to calculate intelligibility. Box 2–3 gives a step-by-step approach for using the Percentage of Consonants Correct (PCC). The following severity division is provided: 90% mild 65% to 85%

mild-moderate

50% to 65%

moderate-severe

50%

severe (Shriberg et al., 1997).

Box 2–3.  Percentage of Consonants Correct (PCC) Ideally, the calculation is based on a 5- to 10-minute conversational speech sample. Depending on the intelligibility of the child and the time constraints of the clinician, this could be reduced. It should be at least a 100-word sample. Only consonants are scored with this measure. The examiner is required to make a correct/incorrect judgment of individual consonants. Incorrect productions include: 1. Deletion of a consonant, 2. Substitution of a consonant, 3. Partial voicing of a prevocalic consonant such as partial voicing of “Sue,” 4. Any distortions, 5. Addition of a sound 6. Initial [h] deletion or n/ŋ in a one-syllable word. Examples: [aUs] for “house” and [rIn] for “ring.” Note: [n] for [ŋ] is considered correct in unstressed syllables, such as [ɹnIn] for “running.” In addition, acceptable allophonic variations are also considered correct, such as the flap in “ladder” [læɾ2]. 7. Rhotic diphthongs such as in “farm” or “bear” (rhotic diphthongs) are counted as consonants; however, [2] and [] alone are counted as vowels and not included in the consonant count. Do not score utterances that are unintelligible or consonants that are repeated two or more times, such as [bɑ bɑ], only the first consonant counts. Calculation:  (Number of correct consonants) ÷ (Number of correct + incorrect consonants) × 100 Source:  Adapted from Shriberg et al. (1997).

50

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Percentage of Whole Words Correct Another way to objectively quantify intelligibility is with the Percentage of Whole Words Correct (Proportion of Whole-Word Correctness) outlined in Ingram and Ingram (2001) and Ingram (2002). It is a relatively easy way to measure intelligibility and accuracy of a child’s productions. This type of measurement is not time consuming and can be used as documentation for accuracy goals (Taps, 2017). The procedure is as follows: n Examine each of the child’s words to determine if it contains an error. Therefore,

the child’s utterance is compared to the standard adult pronunciation. n If there is a complete match, the word is marked correct. n Once the number of correct words is determined the proportion of whole words

(PWW) is calculated. n Calculation:  (Number of whole words correct) ÷ (Total of words in the sample) n For example, if a child’s 100 word sample contains 25 correct whole words then

the Percentage of Whole Words Correct is 25 ÷ 100 = .25 or 25%. Gordon-Brannan and Hodson (2000) provide the following intelligibility/severity scale for prekindergarten children: 91% to 100% for children with “adult-like” speech; 83% to 90% for children in the “mild” category; 68% to 82% for children with moderate intelligibility/speech involvement; 16% to 67% for children in the “severe” (i.e., least intelligible) category. The Percentage of Whole Words Correct (PWWC) was calculated for Andy (see Table 2–6, page 24). Andy has 4 words correct out of 39. He accurately says one, two, toe, and my. His PWWC is 10%. Andy is below the severe categorization noted by Gordon-Brannon et al. (2000). To increase accuracy would be a good goal for Andy. This chapter reviewed several ways data could be collected for a child with a speech sound disorder. The following chapters discuss in detail what type of diagnostic information could be gathered to support a specific treatment protocol. The reader may be directed back to this chapter, or other analysis procedures are provided within the chapter.

References Bernthal, J. E., Bankson, N. W., & Flipsen Jr., P. (2017). Articulation and phonological disorders: Speech sound disorders in children (8th ed.). Boston, MA: Pearson. Blache P. (2000). Constraints, linguistic theories, and natural language processing. In D. N. Christodoulakis (Ed.), Natural language processing — NLP 2000. NLP 2000. lecture notes in Computer Science, 1835. Berlin, Heidelberg: Springer.

2  Ways to Collect and Analyze Data

51

Bowen, C. (2011). Children’s speech sound disorders. Retrieved from http://www.speech-languagetherapy.com/ Coplan, J., & Gleason, J. R. (1988). Unclear speech: Recognition and significance of unintelligible speech in preschool children. Pediatrics, 82, 447–452. Crystal, D. (1982). Profiling linguistic disability. London, UK: Edward Arnold. Crystal, D. (1986). Prosodic development. In P. Fletcher & M. Garman (Eds.), Language acquisition (2nd ed., pp. 174–197). Cambridge, UK: Cambridge University Press. Edwards, H. T. (2003). Applied phonetics: The sounds of American English (3rd ed.). Boston, MA: Cengage Learning. Facultad de Filologia, Universidad de Sevilla. (2011). Retrieved from http://www.siff.us.es/fil/ publicaciones/apuntes/teresals/apartado%209-0.pdf Flipsen, P., Jr. (2006). Syllables per word in typical and delayed speech acquisition. Clinical Linguistics and Phonetics, 20, 303–312. Flipsen, P., Jr. (2010). Measuring intelligibility in children: Why and how. Poster presented at the national convention of the American Speech-Language-Hearing Association, San Diego, CA. Fudala, J., & Stegall, S. (2017). Arizona Articulation Proficiency Scale (4th ed.). Torrance, CA: Western Psychological Services. Gordon-Brannan, M., & Hodson, B. W. (2000). Intelligibility/severity measurements of prekindergarten children’s speech. American Journal of Speech-Language Pathology. 9, 141–150. Ingram, D. (2002). The measurement of whole-word productions. Journal of Child Language, 29, 713–733. Ingram, D., & Ingram, K. D. (2001). A whole-word approach to phonological analysis and intervention. Language, Speech, and Hearing Services in Schools, 32, 271–283. Khan, L. M., & Lewis, N. P. (2015). Khan-Lewis Phonological Analysis (3rd ed.). Boston, MA: NCS Pearson. Klieve, S. A. (1998). Perception of prosody assessment (Unpublished master’s thesis). University of Melbourne, Australia. Lagerberg, T. B., Åsberg, J., Hartelius, L., & Persson, C. (2014). Assessment of intelligibility using children’s spontaneous speech: Methodological aspects. International Journal of Language and Communication Disorders, 49, 228–239. Loos, E. E., Anderson, S., Day, D. H., Jordan, P. C., & Wingate, J. D. (2013) Dictionary of linguistic terms, studies in linguistics. Retrieved from http://www.glossary.sil.org Martin, J. (2017). How to calculate speech intelligibility. Retrieved from http://jacimartin.weebly​ .com/uploads/2/8/7/2/28728745/jminstructions.pdf McLeod, S., & Arciuli, J. (2009). School-aged children’s production of /s/ and /r/ consonant clusters. Folia Phoniatrica et Logopaedica, 61, 336–241. McLeod, S., van Doorn, J., & Reed, V. A. (2001) Consonants cluster development in two-year-olds: General trends and individual differences. Journal of Speech, Language, Hearing Research, 44, 1144–1171. McLeod, S., Harrison, L. J., & McCormack, J. (2012). Intelligibility in context scale. Bathurst, Australia: Charles Sturt University. Retrieved from http://www.csu.edu.au/research/multilingual-speech/ics Osberger, M. J., Robbins, A. M., Todd, S. L., & Riley, A. (1994). Speech intelligibility of children with cochlear implants. Volta Review, 96, 169–180. Peppé, S. (2015). Profiling elements of prosody in speech-communication (PEPS-C) (2015 version). Tynon, Scotland: Author. Retrieved from http://www.peps-c.com/index.html Peppé, S., & McCann, J. (2003). Assessing intonation and prosody in children with atypical language development: The PEPS-C test and the revised version. Clinical Linguistics and Phonetics, 17, 345–354.

52

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Shriberg, L. D. (1993). Four new speech and prosody-voice measures for genetics research and other studies in developmental phonological disorders. Journal of Speech and Hearing Research, 36, 105–140. Shriberg, L. D., Austin, D., Lewis, B. A., McSweeney, J. L., & Wilson, D. L. (1997). The percentage of consonants correct (PCC) metric: Extensions and reliability data. Journal of Speech, Language, and Hearing Research, 40, 708–722. Shriberg, L. D., Kwiatkwoski, J., Rasmussen, J. Lof, G. L., & Miller, J. F. (1990). Prosody-Voice Screening Profile. Tucson, AZ: Communication Skill Builders. Taps, J. (2017). Sample complex phonological goals. Personal correspondence, August 4, 2017. Velleman, S. (2016). Speech sound disorders. Philadelphia, PA: Wolters Kluwer.

2  Ways to Collect and Analyze Data

53

Appendix 2–1 Additional Commonly Used Transcription Symbols

Name

Explanation

When is this Production Used?

Symbol and Examples

Glottal Stop

This sound can occasionally release vowels in stressed syllables, or separate two vowels in words following one another. It is produced by the vocal folds coming together and then being suddenly released after a buildup of pressure.

Children with misarticulations might use the glottal stop to initiate words, or as a sound substitution.

“Nana up” [nn@ ʔp] “cow” [ʔaU] “Hi” [ʔaI]

Flap or Tap

This sound is produced by a brief contact between the tongue and the alveolar ridge. This articulation could occur when either [t] or [d] are produced between two vowels such as in “butter” or “ladder” in casual conversation.

This is a normal variation in production for speakers. It should not be counted as an error if used in the noted situation.

[bt2] with a [t] [bɾ2] with a flap or tap

There are a few consonants that can function as a syllable nucleus. These consonants are referred to as syllabics. They are transcribed with a small vertical line under the consonant in question. Edwards (2003) states that [t] preceding the syllabic [l] can typically cause the [t] to be produced as [ɾ], whereas [t] before the syllabic [n] will result in a [ʔ].

This is a normal variation in production for speakers. It should not be noted as an error if used in the noted situations.

“bottle” [bɑɾl]

To mark where the syllable boundary occurred is done by placing a period between the syllables.

Used in longer words, especially if the syllable boundaries seem to be unusual.

“elephant” [[email protected]@nt] or possibly [[email protected]@nt] “chicken” [tʃI.k@n]

Syllabics

Noting Syllable Boundaries

[ld2] with a [d] [lɾ2] with a flap or tap

“button” [bʔn]

54

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 2–2 Minimal Pairs Ordered by Vowel

/ɑ, a, ɔ/

/i/

/I/

/e/

/E/

/æ/

/o/

/p/ pea

/p/ pin

/p/ pay

/p/ pen

/p/ pat

/p/ pot

/p/

/b/ bee

/b/ bin

/b/ bay

/b/ Ben

/b/ bat

/b/ bought

/b/ bow

/t/ tea

/t/ tin

/t/

/t/ ten

/t/

/t/ taught

/t/ toe

/d/ Dee

/d/

/d/ day

/d/ den

/d/

/d/ dot

/d/ doe

/k/ key

/k/ kin

/k/ Kay

/k/ Ken

/k/ cat

/k/ caught

/k/

/g/

/g/

/g/ gay

/g/

/g/

/g/ got

/g/ go

/m/ me

/m/

/m/ may

/m/ men

/m/ mat

/m/

/m/ mow

/n/ knee

/n/

/n/ nee

/n/

/n/ gnat

/n/ knot

/n/ no

/w/ we

/w/ win

/w/ way

/w/ when

/w/

/w/ watt

/w/ whoa

/j/

/j/

/j/ yay

/j/

/j/

/j/ yacht

/j/ yo

/h/ he

/h/

/h/ hay

/h/ hen

/h/ hat

/h/ hot

/h/ hoe

/f/ fee

/f/ fin

/f/ Fay

/f/

/f/ fat

/f/ fought

/f/ foe

/v/ V

/v/

/v/

/v/

/v/ vat

/v/

/v/

/s/ see, sea

/s/ sin

/s/ say

/s/

/s/ sat

/s/ sought

/s/ sew, so

/z/ Z

/z/

/z/

/z/ Zen

/z/

/z/

/z/

/θ/

/θ/ thin

/θ/

/θ/

/θ/

/θ/ thought

/θ/

/ð/ thee

/ð/

/ð/ they

/ð/ then

/ð/ that

/ð/

/ð/ though

/ʃ/ she

/ʃ/ shin

/ʃ/ Shay

/ʃ/

/ʃ/

/ʃ/ shot

/ʃ/ show

/l/ Lee

/l/ Lynn

/l/ lay

/l/

/l/

/l/ lot

/l/ low

/ɹ/

/ɹ/

/ɹ/ ray

/ɹ/ wren

/ɹ/ rat

/ɹ/ rot

/ɹ/ row

/tS/

/tS/ chin

/tS/

/tS/

/tS/ chat

/tS/

/tS/

/dZ/ gee

/dZ/ gin

/dZ/ J

/dZ/ Jen

/dZ/

/dZ/ jot

/dZ/ Joe

55

2  Ways to Collect and Analyze Data

/U/

/u/

/ɑI/

/ɑU/

/p/ put

/p/ poo

/p/ pie

/p/ pow

/b/

/b/ boo

/b/ buy

/t/

/t/ two

/d/

/ɔI/

//

//

/p/

/p/ putt

/p/ purr

/b/ bow

/b/ boy

/b/ but

/b/ burr

/t/ tie

/t/

/t/ toy

/t/

/t/

/d/ do

/d/ dye

/d/

/d/

/d/

/d/

/k/

/k/ coo

/k/

/k/ cow

/k/ coy

/k/ cut

/k/

/g/

/g/ goo

/g/ guy

/g/

/g/

/g/ gut

/g/ grr

/m/

/m/ moo

/m/ my

/m/

/m/

/m/ mutt

/m/

/n/

/n/ new

/n/ nye

/n/ now

/n/

/n/ nut

/n/

/w/

/w/

/w/ why

/w/ wow

/w/

/w/ what

/w/ were

/j/

/j/ you

/j/

/j/ yow

/j/

/j/

/j/

/h/

/h/ who

/h/ hi

/h/ how

/h/

/h/ hut

/h/ her

/f/ foot

/f/ foo

/f/

/f/

/f/

/f/

/f/ fur

/v/

/v/

/v/ Vie

/v/ vow

/v/

/v/

/v/

/s/ soot

/s/ Sue

/s/ sigh

/s/ sow

/s/ soy

/s/

/s/ sir

/z/

/z/ zoo

/z/

/z/

/z/

/z/

/z/

/θ/

/θ/

/θ/ thigh

/θ/

/θ/

/θ/

/θ/

/ð/

/ð/

/ð/ thy

/ð/ thou

/ð/

/ð/

/ð/

/ʃ/

/ʃ/ shoe

/ʃ/ shy

/ʃ/

/ʃ/

/ʃ/ shut

/ʃ/ sure

/l/

/l/ Lu

/l/ lie

/l/

/l/

/l/

/l/

/ɹ/ root

/ɹ/

/ɹ/ rye

/ɹ/

/ɹ/ Roy

/ɹ/ rut

/ɹ/

/tS/

/tS/ chew

/tS/ chai

/tS/ chow

/tS/

/tS/

/tS/

/dZ/

/dZ/

/dZ/

/dZ/

/dZ/ joy

/dZ/ jut

/dZ/

56

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 2–3 Minimal Word Pairs of Frequently Occurring Substitutions/Phonological Processes Fronting:  Minimal Pair Words Contrasting /k/ - /g/ to /t/ - /d/ /k/ versus /t/

/k/ versus /t/

/g/ versus /d/

/g/ versus /d/

cake cop cape cub key kite cool car corn

ache back bake beak bike knock lake like neck

gate gown go got gull

bag beg bug leg sag

take top tape tub tea tight tool tar torn

ate bat bait beat bite knot late light net

date down doe dot dull

bad bed bud led sad

Fronting:  Minimal Pair Words Contrasting /s/ - /z/ to /θ/ - /ð/ /s/ versus /θ/

/z/ versus /ð/

/s/ versus /θ/

/z/ versus /ð/

sank sick sink sing saw sigh sin song sought sum

Zen

bass Bess face mass miss moss mouse pass

breeze close seize she’s Sue’s tease

thank thick think thing thaw thigh thin thong thought thumb

then

bath Beth faith math myth moth mouth path

Palatal Fronting:  Minimal Pair Words Contrasting /ʃ/ to /s/ /ʃ/ versus /s/

/ʃ/ versus /s/

shack shag shame shave she

bash clash gash leash mesh

sack sag same save see

bass class gas lease mess

breathe clothe seethe sheathe soothe teethe

57

2  Ways to Collect and Analyze Data

/ʃ/ versus /s/

/ʃ/ versus /s/

shed sheep sheet shell shine ship shock shoe shoot show shy

plush

said seep seat cell sign sip sock Sue suit sew sigh

plus

Stopping:  Minimal Pair Words Contrasting /s/ and /z/ to /t/ and /d/ /s/ versus /t/

/s/ versus /t/

/z/ versus /d/

/z/ versus /d/

sell cent sack sag sail sank sea seam sew sip sock

ace base brass case kiss hiss lice mice nice peace rice

Z zing zip zoo zoom zipper

as bees buzz cries dries knees rose size toes ways trays

tell tent tack tag tail tank tea team toe tip talk

ate bait brat Kate kit hit light might night Pete write

D ding dip do doom dipper

add bead bud cried dried need rode side towed wade trade

Stopping (+ Fronting):  Minimal Pair Words Contrasting /ʃ/ and /Z/ to /t/ and /d/ /ʃ/ versus /t/

/Z/ versus /d/

/ʃ/ versus /t/

/Z/ versus /d/

shack shag shake shape sharp she shed

no words found

bash cash fish flash hash mash rash

rouge beige

tack tag take tape tarp tea Ted

bat cat fit flat hat mat rat

rude bade

continues

58

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 2–3.  continued

/ʃ/ versus /t/

/ʃ/ versus /t/

shell ship shop shoe shoot

rush wish

tell tip top two toot

rut wit

Gliding:  Minimal Pair Words Contrasting /l/ to /ɹ/, /w/, and /j/ /l/ versus /ɹ/

/l/ versus /2/

/l/ versus /w/

/l/ versus /j/

lane led lick long lie light lead lock

Dale feel male mole owl tile

life lake leave leap leak light let

loose lard Lou less let

rain red Rick wrong rye right read rock

dare fear mare more our tire

wife wake weave weep weak white wet

use yard you yes yet

Gliding:  Minimal Pair Words Contrasting /ɹ/ to /l/, /w/, and /j/ /ɹ/ versus /l/

/2/ versus /l/

/ɹ/ versus /w/

/ɹ/ versus /j/

race rain red Rick wrong rye right read rock

boar dare fear mare more our tire

rag rail rake rate red ray right rent ring ripe ride raced rest round rake run

rung ram rank rot rear roar

lace lane led lick long lie light lead lock

bowl Dale feel male mole owl tile

wag whale wake wait wed way white went wing wipe wide waste west wound wake won

young yam yank yacht year you’re

2  Ways to Collect and Analyze Data

59

Appendix 2–4 Screener for Prosodic Features Intonation Here is a picture which could be used. Listen to the child’s intonation even if the response is partially or totally unintelligible.

Child uses rising versus falling pitch contours to signal statement (falling) versus question (rising). 1. Tell me about the dog. Tell me more (falling pitch contour). 2. I am going to ask you a question about the dog. Is he blue? Does he have long legs? Now you ask me a question about the dog (rising pitch contour). 3. Now let’s pretend like you are very surprised to see this dog. Look it’s a dog. (You use exaggerated intonation and see if the child can imitate). What is that? or Where is that dog? (These type of sentences show a similar pattern to statements but a sharper fall in pitch at the end.)

continues

60

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 2–4.  continued

4. Use the following pictures to elicit a statement which includes a list of items with rising intonational patterns on each item listed: I want bananas, apples, and cake. Now your turn, what do you want?

Stress 1. Look at the words from your articulation test or that you elicited from the speech sample. Does the stress seem to be appropriate? You can also have the child name the items from the previous pictures. For example, “apple” (stress on first syllable), “banana” (stress on second syllable), “cupcake” (both syllables stressed).

2  Ways to Collect and Analyze Data

61

2. Does it appear that the child uses a wide variety of stress patterns? This will apply only if the child uses multisyllabic words. monkey vs. guitar

spider vs. asleep

rabbit vs. balloon

apple vs. cartoon

telephone vs. kangaroo

lemonade vs. banana

Or does the child seem to use equal and even stress on most words? 3. Sentence stress:  Does the child use stress to differentiate the appropriate word in a sentence. These pictures could be used: Is the baby sleeping? Which baby? The baby or the kitten? Point to one and have child indicate: The kitten is sleeping. The baby is sleeping. The baby is crawling (versus sleeping).

Duration Rate:  Is the rate appropriate? Too slow, too fast? Are the pauses appropriate? Do the words sometimes seem to run on without pauses? Are the individual sounds an appropriate duration? Too long, too short?

62

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 2–5 Probes for Monitoring Therapy Progress [p]

[b]

[t]

[d]

[k]

[g]

paw

big

toe

do

key

go

pig

bite

two

day

comb

game

pea

bean

tan

deep

cow

get

pen

ball

tail

dime

cup

good

poke

boat

ten

down

can

gate

up

tub

bat

bad

back

bag

lip

cab

boot

bed

book

hog

top

knob

hot

hide

duck

dog

tape

mob

out

wood

look

dig

mop

web

nut

need

pack

leg

[m]

[n]

[ŋ]

mad

no

man

now

map

night

me

knot

mean

need

came

pin

wing

dime

bone

long

name

down

hang

time

one

king

gum

pain

bang

[f]

[v]

[s]

[z]

[ ʃ ]

[Z]

fall

van

sew

zoom

shake

Jacques

fan

vote

sing

zoo

shoe

fed

volt

sock

zip

show

fig

vault

soup

zap

shut

2  Ways to Collect and Analyze Data

[Z]

[f]

[v]

[s]

[z]

[ ʃ ]

fight

veil

some

zing

ship

beef

cave

bus

bees

cash

beige

cough

dive

face

boys

leash

rouge

half

five

gas

does

brush

leaf

give

mess

is

wash

tough

have

yes

noise

wish

[θ]

[ð]

[tʃ ]

[dZ]

thank

that

chain

gym

think

the

cheap

jack

thick

they

chop

jam

thin

then

chin

jeep

thing

them

chick

jet

bath

smooth

beach

age

math

bathe

catch

cage

moth

breathe

peach

edge

tooth

clothe

match

huge

teeth

loathe

witch

fudge

[ l ]

[r]

[ j]

[w]

[h]

low

row

you

we

high

lamb

rain

yell

wet

have

like

road

yolk

week

hill

late

ring

young

wall

him

lap

rock

your

wide

home

ball

bear

bull

four

pail

her

will

more

fall

deer

63

64

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 2–6 Intelligibility Grid Mark a + for each syllable that is intelligible and a − for each unintelligible one. The grid contains 100 squares, so when it is full count the number of + marks. This is your percentage of intelligibility.

The following guideline is from Shriberg et al. (1997). 86–90% mild 65–85% mild-moderate 50–65% moderate-severe 50% severe Note.  For best results, this grid should be filled out by a person who is unfamiliar with the child, especially on repeat measures. For example, a teacher or another SLP could help out.

3 Minimal Pair Therapy Treatment Minimal Pair Therapy

Age Range 3+ years of age

Severity of Phonological Disorder Mild to mild-moderate

Key Features n  Targets speech

sound errors that

involve a loss of contrast between sounds or word shapes n Uses minimal pairs to teach new contrasts

Diagnostic Information Needed Identification of child’s phonemic inventory, sound error patterns,

and/or phonemic collapses

Target Selection Traditionally based upon developmental norms and impact on intelligibility

target contrasted with child’s error n  Phases of Treatment: familiarization, perception, production, and carryover training

Basic Structure of Therapy n Create minimal

pairs that include

65

66

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Minimal pair therapy is an effective and versatile intervention method for children who exhibit speech sound disorders (Barlow & Gierut, 2002). This approach provides a framework for target selection, as well as guidelines for the structure and implementation of therapy sessions. Children are presented with word pairs that vary by only one phoneme in an effort to highlight the semantic consequences of misproductions. The goal is an increase in the use of contrastive sounds and word shapes for the purpose of functional communication. Minimal pair therapy may be used alone or in conjunction with other intervention approaches, thus providing flexibility for clinicians when working with children who exhibit mild to moderate speech sound disorders (Baker, 2010). Minimal pair therapy has a long and diverse history within the field of speech-language pathology. Minimal pairs have been a clinical tool and a subject of research for over 25 years (Baker, 2010). As a result, this approach has been referred to by a variety of names, including minimal pairs therapy (Velleman, 2016), method of meaningful minimal contrast (Weiner, 1981), conventional minimal pair treatment (Barlow & Gierut, 2002), and minimal opposition contrast therapy (Gierut, 1990; BaumanWaengler, in press). In this chapter, the term “minimal pair therapy” will be used to refer to the conventional use of minimal pairs, and to differentiate it from recent variations that have developed their own unique methodologies and appellations, including maximal oppositions (Gierut, 1990; see Chapter 5), treatment of the empty set (Gierut, 1991; see Chapter 5), and multiple oppositions (Williams, 2000; see Chapter 4). Two principles are at the core of this intervention approach: 1. The use of minimal pairs, 2. The goal of increasing contrast in a child’s developing phoneme system. Let’s look more closely at each of these elements in order to more fully understand minimal pair therapy.

Minimal Pairs:  What are minimal pairs and why do we use them? Minimal pair therapy involves the use of two words that differ by only one phoneme (e.g., “tea” and “key” or “map” and “mat”). Each duo is known as a “minimal pair,” and these pairs form the basis of intervention. Minimal pairs typically vary by one consonant phoneme, but they may also be formed by words that vary by a vowel phoneme (e.g., “boat” and “bat” or “like” and “look”). Near minimal pairs are often employed as well. Near minimal pairs share the same vowel sound, but differ by the presence or absence of a consonant sound (e.g., “me” and “meat” or “key” and “ski”; Baker, 2010). For the sake of convenience, these various types of word duos will all be referred to as “minimal pairs” in this chapter. Table 3–1 provides a summary of the types of minimal pairs. Typically, minimal pair targets are chosen that represent an error produced by the child. Let’s suppose that a child demonstrates the error pattern of velar fronting, including consistent /t/ for /k/ substitutions. She produces “key” as “tea” because she does not have the /k/ in her phonemic inventory. She may be able to physically produce the

3 Minimal Pair Therapy

67

Table 3–1.  Types of Minimal Pairs Types of Minimal Pairs

Examples

Differ by One Consonant

sea & tea; pan & fan

Differ by One Vowel

coat & cat; lake & lock

Near Minimal Pairs:  Differ by Presence or Absence of One Sound

bee & beet; two & stew

phone [k], but is unable to use /k/ contrastively in words to signal a change in meaning. Therefore, when presented with a picture of a “key,” the child says [ti]. She also says [ti] when presented with a picture of a cup of tea. Thus, the two words “key” and “tea” are produced as homonyms (i.e., two words that sound the same, but have different meanings), creating potential confusion for a listener. Minimal pair therapy capitalizes upon this potential confusion. A key feature of this intervention approach is to highlight the functional consequences of a child’s misproductions. During therapy activities, the child is asked to produce both target words of the minimal pair set (e.g., “tea” and “key”) within the context of communicating with another person. For example, the child might be directing the clinician to toss a beanbag at one of two picture cards placed on the floor (“Toss the bag on key”), or asking for a card during a game of Go Fish (“Do you have a key?”). If the child produces the word “key” as “tea,” then there will be a direct impact upon the activity. The clinician will throw the beanbag on the wrong picture or give the child the wrong card. The child will be faced with the semantic consequence of using the incorrect sound to produce the target word. According to the Pragmatic Principle of Informativeness (Greenfield & Smith, 1976), children will attempt to resolve these miscommunications. In the context of minimal pair therapy, a child will modify speech production in an effort to repair the communication breakdown and achieve the desired outcome of the exchange (e.g., receiving the correct card to complete a pair in Go Fish). The clinician may facilitate the correction of the sound error through a verbal prompt, such as, “Tea? Did you want a picture of a cup of tea? The thing that you drink? Or did you want key? The thing you use to unlock the door?” Some children need help learning to use sounds contrastively, but have already achieved mastery of the physical production of the sounds in question. For example, a child might produce [s] correctly in words such as “sun” and “bus.” He demonstrates the ability to produce the [s] phone and uses it appropriately as a singleton consonant. However, the same child might omit the /s/ from all consonant blends, saying “top” for “stop” and “mut” for “must.” Thus, he does not use /s/ as a phoneme when producing consonant blends. This child needs to learn how to contrast singletons and clusters in the initial and final position of words (in word pairs such as “two” and “stew,” and “bus” and “bust”). This child does not need help with the physical production of the [s] phone.

68

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Other children may have simultaneous difficulty with both the physical production of a sound and its contrastive use. For example, a child who produces “tea” for “sea” may have difficulty using these two sounds contrastively, but may also need assistance to achieve the place, manner, and voicing necessary to produce the [s] phone. In its originally conceived format, minimal pair therapy did not include any cueing for the physical production of the sound (i.e., the phone). However, some studies have suggested that inclusion of cues for correct place, manner, and voicing of the target sound will result in more rapid progress (e.g., Saben & Ingham, 1991). For this reason, many descriptions of minimal pair therapy now allow for the inclusion of articulatory cues to assist children who need help achieving the phonetic production features of target sounds.

Contrast:  How does minimal pair therapy increase the use of contrast? Minimal pair therapy belongs to the family of phonological interventions known as “contrast approaches.” The primary goal of these therapies is the development of contrast within a child’s phonological system. Other contrast approaches include the maximal oppositions approach (Gierut, 1990; see Chapter 5) and multiple oppositions therapy (Williams, 2000; see Chapter 6). The broad intent of all minimal pair therapy is to increase contrast, but the specific objective may vary. Depending on the intervention framework, the goal may be to teach: 1. Contrast between classes of sounds (e.g., velars versus alveolars, stops versus fricatives) 2. Contrast between specific features, such as continuants (fricatives, glides, liquids) and noncontinuants (stops, affricates) 3. Contrast between syllable or word shapes (e.g., CV versus CVC, CVC versus CCVC) In each instance, specific sounds and word pairs (i.e., minimal pairs) will be chosen as the vehicle to teach the contrast needed. Table 3–2 provides examples of the sounds and words that might be utilized to teach each of the skills listed above. It should be kept in mind that the end goal is not accurate production of the specific sounds and words you have chosen, but rather achievement of the contrast necessary to communicate effectively. Each sound pair gives the child new knowledge that will generalize to other similar contrasts. When the phonemes /k/ and /t/ are presented in the context of contrasting word pairs, the child learns about the contrast between velar and alveolar sounds and this knowledge will hopefully generalize to other velar/ alveolar pairs (e.g., /g/ vs. /d/) (Barlow & Gierut, 2002). Additional targets that have been found to be appropriate for minimal pair therapy include: 1. Contrast between specific consonant phonemes, such as sound preference substitutions (Bauman-Waengler, 2020)

3 Minimal Pair Therapy

69

Table 3–2.  Minimal Pair Examples

To teach contrast between . . . Classes of Sounds

Example Velars versus Alveolars

Choose a pair of sounds to target . . .

Choose minimal pair words . . .

/k/ versus /t/

key & tea can & tan kite & tight

Features of Sounds

Continuant versus Non-continuant

/f/ versus /p/

fan & pan foot & put fast & past

Word Shapes

CCVC versus CVC

/st/ versus /t/

stop & top steak & take stick & tick

2. Contrast between vowel sounds (Gibbon & Beck, 2002) 3. Remediation of atypical phonological processes (Dodd & Iancano, 1989; Leahy & Dodd, 1987) When teaching contrast, clinicians should remember that there are both receptive and expressive modalities (Velleman, 2016). Children must learn to receptively discriminate between features, sounds, and syllable shapes in order to differentiate spoken words from each other. They must also expressively use features, sounds, and syllable shapes contrastively in the context of word productions to be effective communicators. Minimal pair therapy incorporates both receptive and expressive knowledge of contrasting words as distinct phases of the intervention process (see Phases of Treatment, p. 79).

Selection of Minimal Pairs:  How will we choose our minimal pairs? Now let’s turn our attention to the selection of minimal pairs. Various methods have been described in the literature for how to identify potential targets, including: 1. Identification of phonological processes, with selection of sound and word pairs to eliminate error patterns (Weiner, 1981), 2. Identification of phonemic collapses and homonymy, with selection of sound classes and word pairs to increase feature contrasts and reduce homonymy (Barlow & Gierut, 2002), 3. Identification of sound errors, with selection of sound pairs that contrast by the fewest feature differences (i.e., minimal contrasts) (Blache, Parsons, & Humphreys, 1981; Lowe, 1994).

70

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

In this chapter, we will utilize the first option — identification of phonological processes, with selection of sound and word pairs to eliminate error patterns. This method for selection of minimal pairs is widely used and corresponds with much of the commercially available intervention materials. The second possibility — identification of phonemic collapses and homonymy to identify minimal pairs — is illustrated during the discussion of the multiple opposition approach in Chapter 4. Finally, the third option — the selection of sound and word pairs that represent minimal feature contrasts — is addressed in Chapter 5 as the maximal opposition approach is explained.

Research Supporting Minimal Pair Therapy Baker (2010) surveyed 42 studies addressing minimal pair therapy that were documented in peer reviewed publications, including two randomized controlled trials (Dodd et al., 2008; Ruscello, Cartwright, Haines, & Schuster, 1993), 25 quasi-experimental studies, and 15 “exploratory” case studies. There was much variation across the studies in terms of methodology, subjects, and specifics of intervention. Yet, taken as a whole, clinicians may find assurance that minimal pair therapy has been shown to be an effective phonological intervention approach. Baker (2010) provides a comprehensive review of research findings and concludes that, “across the 42 studies, the majority reported that intervention was effective” (p. 51). Table 3–3 provides a summary of selected, relevant research articles. More recent research has investigated alternative options for the selection of minimal pairs, including the use of maximal oppositions (Gierut, 1990; see Chapter 5), treatment of the empty set (Gierut, 1991; see Chapter 5), and multiple oppositions (Williams, 2000; see Chapter 4). These newer variations may be more effective and/or more efficient at promoting system-wide change for children with moderate-to-severe phonological disorders (Barlow & Gierut, 2002). More specifically, several aspects of conventional minimal pair therapy have been subsequently questioned by researchers, including the importance of homonymy in effecting change, the use of stimulable sounds, and the implementation of a developmental approach to target selection. In two studies comparing the use of homonymous and nonhomonymous word pairs, the researchers concluded that homonymy was not necessary to promote phonological change (Gierut, 1991; Gierut & Neumann, 1992). Although minimal pair therapy originally recommended the use of sounds that children were physically capable of producing (Lowe, 1994), later studies have supported the use of nonstimulable sounds as intervention targets (e.g., Powell, Elbert, & Dinnsen, 1991). Finally, many investigations have concluded that later developing phonemes may yield greater generalization and more efficient results within the context of a minimal pair framework (e.g., Gierut, Morrisette, Hughes, & Rowland, 1996; Tyler & Figurski, 1994). Despite recent developments in target selection for minimal pairs, the more conventional approach described in this chapter remains an appropriate, evidence-based alternative for some children with phonological disorders (Baker, 2010).

Table 3–3.  Examples of Research Supporting Minimal Pair Therapy

Authors, Date Baker & McLeod, 2004

Blache, Parsons, & Humphreys, 1981

Number of Subjects

Age Range

2 children

4;4 & 4;9

7 children

5;4 to 6;7

Severity Moderate-severe phonological impairment Percentage of Consonants Correct PCC = 48% to 59% (Shriberg, Austin, Lewis, McSweeney, & Wilson, 1997) At least 4 sounds in error on the Developmental Test of Articulation (Hejna, 1959)

Results Approach:  Minimal pair therapy Targets:  Elimination of phonological processes Results:  Both children eliminated cluster reduction with greater than 90% accuracy, but at different rates (12 sessions in 7 weeks vs. 32 sessions over 5 months). Approach:  Minimal pair therapy included both perception and production training Targets:  Sound pairs representing feature contrasts produced as substitution errors Results:  The therapy was effective at decreasing targeted substitution patterns. Generalization occurred to other sound pairs in the same feature class; some generalization also to untreated feature substitutions.

Crosbie, Holm, & Dodd, 2005

18 children

4;8 to 6;5

With consistent or inconsistent speech sound disorder

Approaches:  Comparison of minimal pair therapy with core vocabulary approach Results:  “Phonological contrast therapy” (e.g., minimal pair therapy) resulted in greater change for children with consistent speech sound disorders. Core vocabulary resulted in greater change for inconsistent speech sound disorder.

Dodd & Iacano, 1989

7 children

3;0 to 4;9

Presence of unusual (nondevelopmental) phonological processes

Approach:  Minimal pair therapy with children who demonstrated unusual phonological processes Results:  All children demonstrated positive qualitative changes in speech production; 6/7 children exhibited improvement on PCC (Shriberg et al., 1997) scores obtained during spontaneous speech. continues

71

Table 3–3.  continued

Authors, Date

Number of Subjects

Age Range

Dodd et al., 2008

19 children

3;11 to 6;0

Severity Moderate to severe phonological delay or disorder; consistent error patterns

Results Approach:  Minimal pair therapy, comparing use of minimal contrast pairs with non-minimal contrast pairs Targets:  Initial consonant clusters included Results:  Both groups made progress on speech accuracy and suppression of phonological processes. There was no significant difference between groups.

Ruscello, Cartwright, Haines, & Shuster, 1993

12 children

4;1 to 5;8

Composite score of less than or equal to 15% on KhanLewis Phonological Analysis (Khan & Lewis, 2015)

Approaches:  Comparison of clinician administered minimal pair therapy only versus clinician administered minimal pair therapy combined with a parent administered minimal pair computer program Results:  In terms of magnitude of change, both groups made similar gains.

Tyler, Edwards, & Saxman, 1987

4 children

Weiner, 1981

2 children

3;1, 3;8, 4;1, 5;1

Moderate to severe phonological disorder

Approaches:  Comparison of minimal pair therapy with cycles approach (Hodson & Paden, 1991) Results:  Both approaches shown to be effective and efficient. Cycles concluded to be more appropriate for children with a large number of phonological processes; minimal pair therapy recommended for children with one or a few phonological processes.

4;4 & 4;10

Demonstration of at least six phonological processes

Approach:  Minimal pair therapy, production training only Targets:  Elimination of phonological processes using words representing child’s production and adult form. Later developing sounds were excluded. Results:  The therapy was effective at reducing the occurrence of targeted phonological processes. Generalization occurred to untreated words produced with the same error pattern.

72

3 Minimal Pair Therapy

73

What Is the Age and Severity of Children Who Could Most Benefit from Minimal Pair Therapy? Research subjects receiving minimal pair therapy have included children ranging from two years to ten years of age, with 4 to 5 years being the most common age (Baker, 2010). According to Baker (2016), the minimal pair approach is most appropriate for children who are approximately three years old or older and who demonstrate only one or a few speech sound errors that involve a loss of contrast. These children will typically be diagnosed with a mild or mild-to-moderate phonological disorder. It is not an appropriate intervention approach for children with severe phonological impairments or who demonstrate an inconsistent speech disorder (Crosbie, Holm, & Dodd, 2005), as other treatment protocols have been shown to be more efficient for these children. Minimal pair therapy is designed to target substitution and omission patterns, but not sound distortions or assimilations. Therefore, children who demonstrate primarily the latter may not benefit from this approach (Bauman-Waengler, in press).

Initial Data Collection and Selecting Targets Identification of Error Patterns:  Phonological Processes Target selection for minimal pair therapy begins with identification of the child’s phonemic inventory and sound error patterns. This information may be obtained from the results of a standardized speech assessment, spontaneous speech sample analysis, or use of probe lists (see Chapter 2, Appendix 2–5). Table 3–4 provides a sound production summary for Andrea, age 4 years, 7 months, whose data we will use as an illustration. Intervention targets for minimal pair therapy are typically chosen based upon a child’s sound errors. A common framework utilized for describing error patterns is phonological process analysis. A description of a child’s phonological processes may lead directly to selection of appropriate minimal pairs for intervention. Although Table 3–4 includes substitution and omission errors for Andrea, we will use a summary of her phonological processes to determine specific intervention targets (Table 3–5). Andrea demonstrates the following phonological processes: cluster reduction, final consonant deletion, velar fronting, stopping, gliding, and vowelization/vocalization. It is important to make note of the particular phonemes affected by each process. For example, in Table 3–5 we see that the process of final consonant deletion does not impact Andrea’s production of /m/ or /n/ in the postvocalic position, and that the process of stopping does not impact production of /f/. We will only include sounds in our minimal pairs that are impacted by the error patterns demonstrated. For each error pattern identified, a sound pair is selected to teach the new skill representing the target sound matched against the child’s error production. For example, we might select the phoneme pair of /k/ versus /t/ to address velar fronting, as Andrea substitutes /t/ for /k/. Minimal pair words will be selected that include this contrast, such as “key” and “tea.” Once Andrea learns to successfully contrast the targets in the context of word productions, this knowledge will, in theory, generalize to other sound

74

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 3–4.  Sound Productions for Andrea, Age 4;7 Early Sounds

I

M

F

Stops

Later Sounds

I

M

F

Fricatives

p

3

3

Ø

f

3

3

Ø

b

3

3

Ø

v

b

b

Ø

t

3

3

Ø

s

t

t

Ø

d

3

3

Ø

z

d

d

Ø

k

t

t

Ø

θ

t

t

Ø

g

d

d

Ø

ð

d

d

Ø

ʃ

t

t

Ø

Nasals m

3

3

3

Liquids

n

3

3

3

ɹ

w

w

ə

ŋ

----

n

Ø

l

w

w

ə

Glides

Affricates

w

3

3

----

tS

t

t

Ø

j

3

3

----

dZ

d

d

Ø

3

3

----

Fricative h

pairs that represent the same contrast — in this case, other velar versus alveolar contrasts (e.g., /g/ vs. /d/). If the child’s error production happens to be an omission of a sound, such as occurs in the case of final consonant deletion, then the target sound will be contrasted with its absence. For example, final /p/ might be contrasted with the absence of a final sound in word pairs such as “type” and “tie.” Remember, though, that the phoneme /p/ is not in and of itself the target; rather, the true goal is the production of a final sound in each word. If Andrea learns to produce the exemplar of final /p/ in the context of minimal pairs, the expectation is that she will generalize this skill to the production of other phonemes in the postvocalic position, and will thus eliminate the process of final consonant deletion. She will have achieved contrast between words without a final consonant (e.g., CV words) and words with a final consonant (e.g., CVC words). Table 3–6 provides a list of Andrea’s phonological processes and sounds impacted, as well as sample sound pairs and corresponding minimal pair word possibilities. Which phonological processes should we begin working on? Traditionally, minimal pair therapy progressed in a developmental order, targeting early developing sounds

3 Minimal Pair Therapy

75

Table 3–5.  Andrea’s Phonological Processes Syllable Structure Processes

Sounds Affected

Substitution Processes Continued

Sounds Affected

Syllable Reduction

None

Gliding of Liquids

/ɹ/ → /w/, /l/ → /w/

Cluster Reduction/ Simplification

All /s/ clusters

Gliding of Fricatives

None

Deletion of Initial Consonant

None

Vowelization/ Vocalization also includes Derhotacization

Word-final /l/ and /ɹ/

Deletion of Medial Consonant

None

Backing

None

Deletion of Final Consonant

All sounds except /m, n/

Voicing

None

Reduplication

None

Devoicing

None

Glottal Replacement

None

Liquidization

None

Other Substitution Processes Velar Fronting

/k/ → /t/, /g/ → /d/, /ŋ/ → /n/

Palatal Fronting

None

Labial Assimilation

None

Stopping (Fricatives and Affricates)

/v/ → /b/, /s/ → /t/, /z/ → /d/, /ʃ/ → /t/, /θ/ → /t/, /ð/ → /d/, /tʃ/ → /t/, /dZ/ → /d/

Velar Assimilation

None

Stopping (other, e.g., /m/ → /b/)

None

Nasal Assimilation

None

Deaffrication

None

Liquid Assimilation

None

Affrication

None

Other

Stridency Deletion

See stopping

Denasalization

None

Assimilation Processes

before later developing sounds and beginning with those error patterns that are typically suppressed at a young age. In addition, priority was given to sound error patterns that were deemed to have the greatest negative impact on intelligibility (Lowe, 1994). Of the phonological processes demonstrated by Andrea, final consonant deletion would likely be the first to be suppressed, according to most developmental norms (e.g., Grunwell, 1987). Omission of sounds also has a significant negative effect on speech intelligibility.

76

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 3–6.  Andrea’s Sample Sounds and Words Phonological Process

Sounds Affected

Deletion of Final Consonants

All sounds except /m, n/

Sample Sound Pairs final Ø versus /p/

Sample Minimal Pair Words ma/mop bee/beep who/hoop paw/pop tie/type

Velar Fronting

/k/ → /t/, /g/ → /d/, /ŋ/ → /n/

initial /t/ versus /k/

tea/key tight/kite tan/can tape/cape top/cop

Stopping

/v/ → /b/, /s/ → /t/, /z/ → /d/, /ʃ/ → /t/, /θ/ → /t/, /ð/ → /d/, /tS/ → /t/, /dZ/ → /d/

initial /t/ versus /s/

tea/sea toe/sew tip/sip team/seam tent/cent

Cluster Reduction

All /s/ clusters

initial /t/ versus /st/

two/stew tack/stack tough/stuff top/stop team/steam

Gliding

/l/ → /w/, /ɹ/ → /w/

initial /w/ versus /l/

weep/leap wet/let white/light way/lay whip/lip

Therefore, deletion of final consonants would be a good initial target for Andrea’s minimal pair intervention. In addition, velar fronting would be an appropriate target, as it, too, is typically suppressed by children of Andrea’s age. There are many options for how to progress through Andrea’s list of error patterns. Research supports the use of a vertical goal attack strategy (i.e., targeting one error sound or pattern at a time until a predetermined level of mastery is reached), a simultaneous goal attack strategy (i.e., working on more than one target simultaneously), or a cyclical goal attack strategy, such as is used in Hodson and Paden’s Cycles Approach (Hodson & Paden, 1991; see Chapter 9).

3 Minimal Pair Therapy

77

Establishing Treatment Goals Prior to establishing treatment goals, we must consider data collection. Baker (2016) recommends gathering several types of data over the course of minimal pair intervention. Assessment data is gathered to identify initial targets, baseline data is obtained on targets at the beginning of treatment, then both quantitative and qualitative treatment data is collected throughout the duration of therapy and, finally, generalization data may be obtained on untreated targets. Baker (2016) recommends that generalization data be collected every fourth session. We have already looked at the initial assessment data for Andrea, including her phonemic inventory and list of phonological processes. Now we must administer a sound probe to establish her baseline performance on proposed therapy targets (e.g., final consonant deletion, velar fronting). Probe words will not receive direct instruction during therapy, and should therefore not be included as intervention stimuli. Table 3–7 provides an example of a probe created for Andrea. Table 3–7.  Baseline and Generalization Probe for Andrea Minimal Pair Therapy:  Baseline & Generalization Probe Baseline Words (also generalization words) Goal #1: Final Consonant Deletiona

a b

Goal #2: Velar Frontingb

boat −

coo −

neat −

Kay −

wood −

cone −

hide −

cub −

home +

cap −

time −

go −

peep −

guy −

win −

goop −

tub −

gum −

bib −

gone −

Baseline

Date: 3/1

1/10 = 10%

Date: 3/1

Generalization

Date:

Date:

Generalization

Date:

Date:

Generalization

Date:

Date:

Correct production (+) indicates production of any final consonant sound. Correct production (+) indicates production of any velar sound.

0/10 = 0%

78

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

In this probe, all productions are recorded as correct (+) or incorrect (−). Clinicians may choose to include additional information regarding the nature of errored productions, for example, by noting specific sound substitutions. Andrea’s probe will be administered prior to the initiation of minimal pair treatment to establish a baseline, and then again periodically to assess progress and monitor generalization to untreated words. Clinicians may also want to monitor progress on the treatment words that will be established in the section below. We are now prepared to turn our attention to the development of specific treatment goals for Andrea. If we choose a developmental approach to select targets, we will prioritize the error patterns that are typically suppressed at a younger age, including deletion of final consonants and velar fronting. In addition, we will use earlier developing sounds as vehicles to teach the contrasts needed to suppress these error patterns. We have selected the phoneme final /p/ to contrast with its absence in an effort to eliminate Andrea’s use of final consonant deletion, and the sound pair of /k/ versus /t/ to eliminate the occurrence of velar fronting. Table 3–8 provides examples of how these intervention targets might be presented as treatment goals. It is important to note that there are many possibilities for writing appropriate minimal pair therapy goals. The examples provided in Table 3–8 do not require accurate production of a specific phoneme, as these phonemes are not in and of themselves the goal of intervention. Instead, the goal is for Andrea to demonstrate a skill (production of a final consonant or production of a velar sound) that will increase contrast within her phonological system. Specific sound targets (e.g., final /p/ or initial /k/) are the vehicle to eliminate an error pattern, and additional sound targets may be added over the course of an intervention program if generalization has not been achieved. Each treatment goal must have corresponding baseline data, in order for progress to be accurately determined. As illustrated in the examples above (see Tables 3–7 and 3–8), the nature of the baseline data will be guided by the content of the goal. Clinicians will Table 3–8.  Example Goals for Andrea; Minimal Pair Therapy Goal Area #1:  Final Consonant Deletion; /p/ versus Ø Baseline #1:  Andrea produced early developing consonant sounds in the final position of CVC (consonant-vowel-consonant) words in 1 out of 10 trials, in the context of spontaneously naming pictures, during administration of a baseline probe. Goal #1:  Andrea will produce early developing consonant sounds in the final position of CVC (consonant-vowel-consonant) words in 8/10 trials, in the context of spontaneously naming pictures. Goal Area #2:  Velar Fronting; /k/ versus /t/ Baseline #2: Andrea produced a velar sound (i.e., /k, g/) in the initial position of CV (consonant-vowel), and CVC (consonant-vowel-consonant) words in 0 out of 10 trials, in the context of spontaneously naming pictures, during administration of a baseline probe. Goal #2: Andrea will produce a velar sound (i.e., /k, g/) in the initial position of CV (consonant-vowel) and CVC (consonant-vowel-consonant) words in 6 out of 10 trials, in the context of spontaneously naming pictures.

3 Minimal Pair Therapy

79

need to design baseline probes to provide the appropriate information needed for the goals they develop.

Beginning Therapy Target Word Selection The overarching goal of minimal pair therapy is to achieve the contrast between sounds or word shapes that signals a change in meaning between two words. Therefore, intervention always begins at the word level, not the sound level. Most descriptions of minimal pair treatment include the use of five sets of minimal pair words. For Andrea, we have selected the following words to reduce the occurrence of final consonant deletion: ma/mop, be/beep, who/hoop, paw/pop, and tie/type. In addition, we have chosen these word pairs to decrease Andrea’s use of velar fronting: tea/key, tight/kite, tan/can, tape/cape, and top/cop (see Table 3–6). It is advisable to select words that contain a variety of vowel sounds as well as consonant sounds (other than the target sound) that are already in the child’s pretreatment phonemic inventory. During therapy, target words are represented in picture form. Therefore, a stimulus card must be created to represent each word. Clinicians may want to create two copies of each picture for use in games such as matching and Go Fish. See Appendix 2–3 for lists of words representing common substitution and omission error patterns. Additional minimal pair word lists may be obtained for free online, and there are commercially produced materials available.

Phases of Treatment Many variations of minimal pair therapy have been described in the literature. The following is a delineation of the phases of treatment typically included in this approach, with explanations of possible modifications.

Phase 1:  Familiarization Intervention begins by familiarizing the child with the treatment words. If we want children to use targets appropriately in the context of producing words, we must first make sure that they understand the vocabulary. Therefore, during this phase, the child listens as the clinician reviews each picture, providing a verbal model of the stimulus word and a brief description of its meaning. For example, Andrea might be presented with pictures representing the minimal pair of “tie” and “type,” in order to facilitate production of final /p/. During the familiarization phase, the clinician will present each picture and describe its meaning: “This is tie. The boy needs to tie his shoes. This is type. The girl likes to type on her computer.” The goal is to ensure that the child understands the words represented by the stimulus pictures.

Phase 2:  Perception During this phase of treatment, the emphasis is on developing the child’s ability to discriminate between the word pairs. The child must demonstrate the ability to receptively

80

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

perceive the contrast between target sounds or word shapes within the context of spoken words. The clinician creates an opportunity for the child to discriminate between the phonologically similar words that are presented. The child might be asked to simply listen to each word and find the appropriate picture from a field of cards laid on a table. Or we might try to make this phase more fun and interactive by asking the child to throw a beanbag at the correct picture taped to the wall or to jump on the correct picture that has been placed on the floor. The number of pictures presented at one time may be altered according to the activity and the child’s abilities. The clinician will provide positive feedback for correct identifications and instructional feedback for incorrect responses. For example, if the child correctly points to the picture representing “type” when directed to do so, the clinician might simply say, “Yes, that’s right, the girl types on her computer.” However, if the child responds incorrectly by indicating the picture depicting “tie,” we could prompt “Tie or type? Listen again. Pick up the picture that shows type.” Remember to present the cards in random order to ensure that the child is correctly perceiving the targeted contrast. If a child has continued difficulty during this phase of intervention, clinicians should consider more direct work on speech discrimination prior to proceeding to the production phase (Baker, 2016; Rvachew, 1994b). Bauman-Waengler (in press) indicates that an appropriate criteria for discrimination training is seven correct consecutive responses. If a child fails to reach this criteria, focused perception work may be necessary. The SAILS (Speech Assessment and Interactive Learning System, AVAAZ Innovations, Rvachew, 1994a) computer program, developed by Susan Rvachew, is an excellent option for improving speech perception abilities.

Phase 3:  Production Now the child practices the target by producing it in the minimal pair words. In its simplest version, the child tells the clinician which card to pick up from a field of two minimal pairs. If the child produces the target incorrectly — for example, saying “tea” instead of “key”— the adult will pick up the incorrect picture. The clinician might say, “Oh, a cup of tea? Is this what you meant to say? Or did you mean key? Tell me again.” The child is thus confronted with the functional impact of his misproduction. An essential element of the production phase is that there must be a natural consequence for errored productions. That is, if the child says the word incorrectly, the resulting semantic confusion should be highlighted by an immediate and direct effect. Perhaps the clinician produces the wrong card during a game of Go Fish, or places a sticker on the wrong picture, or tosses a beanbag onto the wrong word, when directed by the child. In each of these scenarios, the child is confronted with the consequences that result from producing a target word as its homonym. The child must produce the appropriate contrast between the two words in order to be a successful communicator within the context of the game. In conventional minimal pair therapy, no instruction or feedback is given regarding articulatory accuracy. Because the goal is for the child to produce a contrast between the target sounds or words, and thus reduce homonymy, positive reinforcement is given when the child successfully produces the target feature or contrast, even if it is not accurate in terms of articulation (Barlow & Gierut, 2002; Velleman, 2016). However, some

3 Minimal Pair Therapy

81

studies have indicated that the use of articulatory cues during the production phase will enhance outcomes (e.g., Saben & Ingham, 1991). In fact, some variations of minimal pair therapy include articulatory practice of the targets as a distinct phase of intervention. Baker (2010) describes one such alternative format as “Perception-Production Minimal Pair Intervention.” Within this framework, after the familiarization and perception training is completed, the child is asked to first imitate the target words and then produce them independently while naming the picture stimuli. The focus during these additional phases is on learning the physical production features of the target sounds. Instructional feedback is given for incorrect productions, including cues for articulatory features (place, manner, voicing), as needed. Once a predetermined level of accuracy has been reached during both the imitation and naming phases, therapy progresses to production of the minimal pairs within the context of directing the clinician. Keep in mind that this additional sound production practice will be conducted at the word level, rather than at the sound level as occurs during traditional articulation therapy.

Phase 4:  Carryover Training Once the child achieves accurate production of the targets at the word level, the linguistic complexity slowly increases to ensure carryover to connected speech. The clinician might ask the child to produce the targets using a carrier phrase, such as “point to key” or “give me key” and then progress to longer utterances, such as “I need a key,” “Where is the key?” or “Please give me a key.” Parent involvement in the practice of target words may assist in generalization of skills to the home environment. Baker (2010) found that 37% of minimal pair therapy research investigations included a parent administered component, either during the actual therapy sessions or through administration of home practice. Clinicians might send home minimal pair picture cards once mastery has been met during the production phase, and encourage families to play games with them, such as hide and seek or Go Fish. One of the advantages of minimal pair therapy is its flexibility. Clinicians may choose to implement it as its own approach in a manner such as that described in the steps above, or it may be incorporated into other intervention programs (Baker, 2010). For example, Hodson recommends utilizing minimal pair therapy within the secondary targets of the Cycles Approach (Hodson, 2007; see Chapter 9) and the use of minimal pairs is an important component of metaphon therapy (Howell & Dean, 1994).

Progress Monitoring Generalization of skills is the most important indicator of the success of a minimal pair intervention program, according to Baker and McLeod (2004). Thus, clinicians will need to periodically probe for generalization of targeted skills. See Table 3–7 for an example of a probe that may be used to assess progress and generalization of skills. Baker (2010) recommends that if a child is making minimal progress after 4 to 6 weeks, modifications to the intervention plan should be considered. These adjustments might include changing the goal attack strategy, increasing the frequency of sessions, expanding the

82

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

number of words or trials per session, or even considering another intervention method altogether (Baker, 2010). Research has demonstrated that other phonological intervention approaches may be more efficient and/or effective than conventional minimal pair therapy (Barlow & Gierut, 2002). Therefore, if this method is chosen, it is particularly important to carefully monitor progress and to consider alternative interventions if a child’s improvement is slow.

Group Therapy Minimal pair therapy works well within the context of group intervention due to its emphasis on communicative effectiveness. Although it is often challenging to keep several students engaged simultaneously, the opportunity for children to produce target words during interactions with peers can be a therapeutic advantage rather than a drawback. Let’s consider some possibilities for group activities that are appropriate for the different phases of minimal pair intervention.

Phase 1: Familiarization The main purpose of the familiarization phase is to ensure that a child understands the vocabulary represented by the stimulus cards. To this end, the clinician may review the pictures, providing a short description of each target word, as the child listens attentively. Such a straight-forward approach may work well in the context of individual therapy. However, as any clinician who has worked with groups of young children can tell you, attention spans seem to be inversely related to the number of children participating in any given activity. Expecting three or four children to listen patiently while all of the intervention targets are reviewed is likely to be unrealistic. Therefore, clinicians may want to employ more interactive activities which allow for familiarization with the vocabulary to occur. Some possibilities are listed here: 1. Mix the stimulus cards for all of the children together and play a game such as Go Fish or matching. Ask each child to describe the picture that is requested or chosen. 2. Briefly describe one stimulus card at a time to the group, while keeping the picture hidden. The child who guesses the word correctly, keeps the card; the child with the most cards at the end of the game wins. 3. Ask each child to select a stimulus card and keep it hidden from the group. The other children take turns asking scripted questions about the card (e.g., “What color is it?”; “What do you use it for?”; “Where do you find it?”; “What is it made of?”) until someone guesses the word correctly. Remember that during any of these activities, the emphasis will be on learning the vocabulary words and not on accurate sound production. Because of this, any of these

3 Minimal Pair Therapy

83

games may simultaneously be used to improve the semantic skills of children in the group who have language deficits.

Phase 2:  Perception The perception phase of minimal pair therapy focuses on the child’s ability to discriminate between the two words represented by each picture duo. Activities will involve someone directing the child to identify a specific picture. In its simplest form, a child might be asked to point to a card when it is named by the clinician. However, this phase of therapy can be made more fun and interactive by directing the child to perform actions or complete tasks that involve the target word. Examples of discrimination activities include directing the child to: 1. Throw a beanbag on the correct stimulus card. 2. Place a sticker on the correct picture. 3. Jump on or over the correct card placed on the floor. 4. Feed the correct card to a stuffed animal or puppet. 5. Shine a flashlight at the correct picture. In each instance, both pictures of a minimal pair set must be visible to the child so that accurate discrimination is verified. Rather than the therapist always being the director, other children in the group who are able to produce the target sounds may take on this role. In fact, children with expressive language goals who have good sound production skills may benefit from these activities by being given the opportunity to practice sentence formulation.

Phase 3:  Production As the name implies, this phase of minimal pair intervention should be designed to provide maximum production practice of the target words. Baker (2010) recommends that every child says each of the five target words at least 20 times for a total of 100 trials per session. In the context of group therapy, this goal will be easier to achieve if activities are simple and fast paced. Children must have frequent turns to produce target words. An easy way to structure an efficient minimal pair therapy session is to utilize the same activity for the perception phase and production phase. Once the clinician has directed the child to complete a task (e.g., throwing a beanbag) and the child has demonstrated appropriate discrimination skills, it will then be the child’s turn to direct the adult to complete the same task. When the child becomes the director of another’s actions, she must now correctly produce the target sound in order to effectively communicate her message. One advantage of group work is that children may direct each other in the context of interactive games, providing necessary sound production practice and immediate pragmatic consequences for misproductions.

84

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Phase 4: Carryover Training The final phase of minimal pair therapy may be the most conducive to a group therapy format, as it involves practicing target words in utterances of increasing complexity, in a variety of situations with different communication partners. The same activities utilized during the production phase may be modified to increase the linguistic demands by lengthening the utterances produced by the child from single words to longer and longer sentences. For example, the child might begin the production phase by saying “key” to request or direct another person’s actions. During carryover training, the child might be required to say, “Hand me the key,” “Do you have the key?” or “Throw the beanbag on the key.” Generalization to daily interactions with peers may be facilitated by the communication opportunities created during these group therapy sessions.

Case Study Initial Data Collection and Identification of Targets Step 1:  Identification of Error Patterns and Potential Targets Table 3–9 presents a summary of sound productions obtained from the results of a standardized speech assessment and spontaneous speech sample, for James, age 7;0. James was diagnosed with a phonological disorder at the age of four. After three years of speech and language therapy, most of his errors have been resolved. He currently demonstrates prevocalic voicing of /f/, /s/, and /θ/, as well as gliding of /l/ and /ɹ/. An extensive description of James’ sound errors utilizing a conventional phonological process analysis form (such as the one we used for Andrea, see Table 3–5) will not be necessary, as he exhibits only two error patterns. Table 3–10 presents a summary of James’ error patterns and sounds impacted, as well as possible sound pairs and sample minimal pair words. If we use a developmental approach to target selection, we might begin therapy by addressing the phonological process of prevocalic voicing, as it is typically suppressed at a younger age than gliding, and we might use the sound pair of /f/ versus /v/, as /f/ is an earlier developing phoneme affected by this error pattern. We will select at least five word pairs to represent the /f/ versus /v/ contrast (see Appendix 2–2). For James, we have chosen the following words: fan/van, fail/veil, face/vase, fast/vast, fairy/very. Once James has achieved a predetermined level of mastery (e.g., 70% across three consecutive sessions) during production of these minimal pairs, a progress monitoring probe will be administered to assess generalization. If, as hoped, James’ new ability to contrast initial /f/ versus /v/ has generalized to other initial and medial voiced versus voiceless strident contrasts (i.e., /s/ versus /z/ and /θ/ versus /ð/), we will proceed to the next phonological process, implementing a vertical goal attack strategy. Continuing to follow a developmental target selection model, we will address the error pattern of gliding by presenting the /l/ versus /w/ contrast to James in word pairs such as light/white, lead/weed, leak/week, lay/way, and late/wait. If necessary, the /ɹ/ versus /w/ contrast may be targeted next, in word pairs such as red/wed, rate/wait, reed/weed, ray/way, and right/white.

Table 3–9.  Sound Productions for James, Age 7;0 Early Sounds

I

M

F

Stops

Later Sounds

I

M

F

Fricatives

p

3

3

3

f

v

v

3

b

3

3

3

v

3

3

3

t

3

3

3

s

z

z

3

d

3

3

3

z

3

3

3

k

3

3

3

θ

ð

ð

3

g

3

3

3

ð

3

3

3

ʃ

3

3

3

Nasals m

3

3

3

Liquids

n

3

3

3

ɹ

w

w

----

ŋ

----

3

3

l

w

w

----

Glides

Affricates

w

3

3

----

tS

3

3

3

j

3

3

----

dZ

3

3

3

3

3

----

Fricative h

Table 3–10.  James’ Sample Sounds and Words

Error Pattern

Sounds Affected

Sample Word Pairs

Sample Minimal Pair Words

Prevocalic Voicing

Initial and medial

initial /f/ vs. /v/

face/vase

/f/ → /v/

initial /s/ vs. /z/

sip/zip

/s/ → /z/

initial /θ/ vs. /ð/

thigh/thy

Initial /l/ → /w/

Initial /l/ vs. /w/

let/wet

Initial /ɹ/ → /w/

Initial /ɹ/ vs. /w/

ray/way

/θ/ → /ð/ Gliding

85

86

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

References Baker, E. (2010). Minimal pair intervention. In A. L. Williams, S. McLeod, & R. J. McCauley (Eds.), Interventions for speech sound disorders in children (pp. 41–72). Baltimore, MD: Brookes. Baker, E. (2016). Contrastive approaches to phonological intervention: Minimal pairs and multiple oppositions. ASHA Online Conference: Improving Intelligibility in Children with Speech Sound Disorders. Baker, E., & McLeod, S. (2004). Evidence-based management of phonological impairment in children. Child Language Teaching and Therapy, 20(3), 261–285. Barlow, J. & Gierut, J. (2002). Minimal pair approaches to phonological remediation. Seminars in Speech and Language, 23(1), 57–67. Bauman-Waengler, J. (in press). Articulation and phonology in speech sound disorders (6th ed.). Boston, MA: Pearson. Blache, S. E., Parsons, C. L., & Humphreys, J. M. (1981). A minimal-word-pair model for teaching the linguistic significant difference of distinctive feature properties. Journal of Speech and Hearing Disorders, 46, 291–296. Crosbie, S., Holm, A., & Dodd, B. (2005). Intervention for children with severe speech disorder: A comparison of two approaches. International Journal of Language and Communication Disorders, 40, 467–491. Dodd, B., Crosbie, S., McIntosh, B., Holm, A., Harvey, C., Liddy, M., . . . Rigby, H. (2008). The impact of selecting different contrasts in phonological therapy. International Journal of SpeechLanguage Pathology, 10(5), 334–345. Dodd, B., & Iacono, T. (1989). Phonological disorders in children: Changes in phonological process use during treatment. British Journal of Disorders of Communication, 24, 333–351. Gibbon, F. E., & Beck, J. M. (2002). Therapy for abnormal vowels in children with phonological impairment. In M. J. Ball & F. E. Gibbon (Eds.), Vowel disorders (pp. 217–248). Woburn, MA: Butterworth-Heinemann. Gierut, J. A. (1990). Differential learning of phonological oppositions. Journal of Speech and Hearing Research, 33, 540–549. Gierut, J. A. (1991). Homonymy in phonological change. Clinical Linguistics and Phonetics, 5, 119–137. Gierut, J. A., Morrisette, M. L., Hughes, M. T., & Rowland, S. (1996). Phonological treatment efficacy and developmental norms. Language, Speech, and Hearing Services in Schools, 27, 215–230. Gierut, J. A., & Neumann, H. J. (1992). Teaching and learning /th/: A non-confound. Clinical Linguistics and Phonetics, 6(3), 191–200. Gillon, G. (2018). Phonological awareness: From research to practice (2nd ed.). New York, NY: Guilford Press. Greenfield, P., & Smith, J. (1976). Communication and the beginnings of language: The development of semantic structure in one-word speech and beyond. New York, NY: Academic Press. Grunwell, P. (1987). Clinical phonology. Gaithersburg, MD: Aspen. Hejna, R. (1959). Developmental test of articulation. Ann Arbor, MI.: Speech Materials. Hodson, B. W. (2007). Evaluation and enhancing children’s phonological systems: Research and theory to practice. Greenville, SC: Thinking Publications. Hodson, B. W., & Paden, E. P. (1991). Targeting intelligible speech: A phonological approach to remediation (2nd ed.). Austin, TX: Pro-Ed. Howell, J. & Dean, E. (1994). Treating phonological disorders in children: Metaphon — theory to practice (2nd ed.). London, UK: Whurr. Khan, L. M., & Lewis, N. P. (2015). Khan-Lewis Phonological Analysis (3rd ed.). Boston: NCS Pearson.

3 Minimal Pair Therapy

87

Leahy, J. & Dodd, B. (1987). The development of disordered phonology: A case study. Language and Cognitive Processes, 2, 115–132. Lowe, R. J. (1994). Phonology: Assessment and intervention applications in speech pathology. Baltimore, MD: Williams & Wilkins. Powell, T. W., Elbert, M., & Dinnsen, D.A. (1991). Stimulability as a factor in the phonological generalization of misarticulating preschool children. Journal of Speech and Hearing Research, 34, 1318–1328. Ruscello, D. M., Cartwright, L. R., Haines, K. B., & Shuster, L. I. (1993). The use of different service delivery models for children with phonological disorders. Journal of Communication Disorders, 26, 193–203. Rvachew, S. (1994a). Speech Assessment and Interactive Learning System (SAILS). AVAAZ Innovations. Rvachew, S. (1994b). Speech perception training can facilitate sound production learning. Journal of Speech and Hearing Research, 37, 347–357. Saben, C. B. & Ingham, J. C. (1991). The effects of minimal pairs treatment on the speech-sound production of two children with phonological disorders. Journal of Speech and Hearing Research, 34, 1023–1040. Shriberg, L. D., Austin, D., Lewis, B. A., McSweeney, J. L., & Wilson, D. L. (1997). The percentage of consonants correct (PCC) metric: Extensions and reliability data. Journal of Speech, Language, and Hearing Research, 40, 708–722. Tyler, A. A., Edwards, M. L., & Saxman, J. H. (1987). Clinical application of two phonologically based treatment procedures. Journal of Speech and Hearing Disorders, 52, 393–409. Tyler, A. A., & Figurski, G. R. (1994). Phonetic inventory changes after treating distinctions along an implicational hierarchy. Clinical Linguistics & Phonetics, 8, 91–107. Velleman, S. L. (2016). Speech sound disorders. Philadelphia, PA: Wolters Kluwer. Weiner, F. (1981). Treatment of phonological disability using the method of meaningful contrasts: Two case studies. Journal of Speech and Hearing Disorders, 46, 97–103. Williams, A. L. (2000). Multiple oppositions: Theoretical foundations for an alternative contrastive intervention approach. American Journal of Speech-Language Pathology, 9, 282–288.

4 Multiple Oppositions Intervention Treatment

Key Features

Multiple Oppositions

n Individual

Age Range 3 to 6½ years

Severity of Phonological Disorder Moderate to severe Demonstrate extensive phonemic collapses and/or high degree of homonymy

phoneme collapses and rule sets are determined. n  Selects groups of sounds that represent phoneme collapses. n  Targets include families of homonyms (e.g., top, cop, shop, chop, and stop are

all pronounced the same).

Diagnostic Information Needed Extensive speech production sample; phonetic and phonemic inventories, including substitution and omission patterns and distribution of phonemes

Systemic analysis results in mapping of phonemic collapses and identification of individual rule sets

Target Selection Targets selected through application of a distance metric, which yields continues

89

90

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

groups of sounds that are maximally distinct from the error sound and represent maximal classifications from each other

Basic Structure of Therapy n Select target sounds representing phoneme collapses and rule sets

n Develop contrastive word sets containing target sounds n  Phases of Treatment are specified: familiarization,

production, imitation, interactive play, use of communicative contexts and conversational recasts

The multiple oppositions approach, as described by Williams (2000a, 2000b, 2005a, 2005c, 2010), is a variation of minimal pair therapy that incorporates larger treatment sets. Rather than targeting one contrast pair at a time, a whole family of homonyms is targeted simultaneously. Each child’s unique phonological system is analyzed to identify phonemic collapses, and targets are developed to systematically increase contrasts and reduce homonymy across the child’s phonological system. Multiple oppositions therapy offers an innovative approach to selecting targets and developing a data-driven treatment program for children with moderate to severe speech sound disorders. There are three aspects of the multiple oppositions approach that differentiate it from other phonological interventions, including: 1. The use of contrastive sets of words to stimulate change across a child’s disordered phonological systems, 2. The identification of each child’s unique rule sets that govern the organization of individual phonologies, utilizing a phonemic mapping analysis method, 3. The selection of specific sound targets according to a distance metric that creates an expanded frame of learning for optimal growth. These three components are explored in this chapter, highlighting the unique features of this therapeutic program. In addition, the multiple oppositions approach provides specific guidelines for data analysis, target selection, implementation of therapy, and progress monitoring (Williams, 2000a, 2000b, 2005a, 2005c, 2010). Each of these components will be presented, so that clinicians may successfully employ a multiple oppositions approach to achieve significant change for children on their caseloads.

Contrastive Sets:  What is a contrastive multiple oppositions set? A contrastive multiple oppositions set is a group of words that is produced by a particular child as homonyms. For example, a child with a speech sound disorder might say “top” when attempting to name pictures representing the vocabulary targets of “cop,” “shop,” “chop,” and “stop.” Despite understanding that the pictures represent

4 Multiple Oppositions Intervention

91

four different words, the child produces them all as the same word, “top,” thus creating homonyms that do not exist in the adult lexicon. During multiple oppositions therapy, the child is presented with the challenge of producing each word in a contrastive set differently, in order to reduce homonymy. The emphasis on homonymy differentiates multiple oppositions from maximal oppositions and treatment of the empty set (Williams, 2003; Gierut, 1990, 1991; see Chapter 5). Like conventional minimal pair therapy, this approach contrasts the target phonemes (i.e., the phonemes to be learned) with the error sound produced by the child. In our example above, the sound produced in error by the child is [t]. Therefore, /t/ will be contrasted with each of the target phonemes (/k, ʃ, tS, st/) in the context of multiple oppositions word sets (e.g., top, cop, shop, chop, and stop). We anticipate that all of the target words will be produced by the child as homonyms (i.e., “top”). In the context of therapy activities, the child will experience the misunderstandings that result from her misproductions, just like in conventional minimal pair therapy. The child will be motivated to differentiate production of these words by realizing the necessary contrast between the targets, and will experience the resulting increase in speech intelligibility. In order to achieve this goal, multiple oppositions therapy employs larger treatment sets than conventional minimal pair therapy. Rather than targeting one pair of words at a time (e.g., “top” and “cop”), a whole family of homonyms is targeted simultaneously (e.g., “top,” “cop,” “shop,” “chop,” and “stop”). This allows children to learn a variety of contrasts more quickly, and, thus, effect change across disordered phonological systems efficiently.

Rule Sets and Phonemic Maps:  What is a rule set? What is a phonemic map? A rule set describes a child’s unique strategy of compensation for a limited phonemic inventory. Many children with speech sound disorders have very few sounds available to create all the words they need to communicate. They often collapse many adult phonemes to one sound that they are able to produce. In the contrastive word set presented above (“top,” “cop,” “shop,” “chop,” and “stop”), the child has collapsed the phonemes /k/, /ʃ/, /tS/, and /st/ to /t/. This collapse forms one rule set and helps us to understand the internal guidelines the child has developed to create a correspondence between his disordered phonology and the adult system (Williams, 2003). This rule set, or phonemic collapse, may be represented visually as a phonemic map (Table 4–1). A map of phonemic collapses allows us to discover a child’s individual rule set. Rather than describing sound errors using predetermined categories, such as phonological processes, we look at each child’s limited sound inventory in relation to adult phonology and discover individual patterns of compensation (Williams, 2003). If we utilize a phonological process analysis, we might describe the substitution errors in our example as velar fronting (/t/ for /k/), stopping (/t/ for /ʃ/ and /t/ for /tS/), and cluster reduction (/t/ for /st/). However, Williams (2000b) encourages us to look more holistically at the child’s system. By making note of all the sounds that collapse to one phoneme and addressing this pattern as an interrelated rule set, rather than as unrelated errors, we will effect

92

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 4–1.  Phonemic Map Example

t k t

ʃ



tS

st

greater change across a child’s phonological system. In fact, Williams (2010) argues that it is easier for children to learn new contrasts within the larger context of a whole rule set with the result being a systematic reorganization of the child’s sound system. Most children with moderate or severe speech sound disorders demonstrate multiple rule sets. That is, many patterns of phonemic collapses may be identified for each child to describe his or her disordered phonemic system. Collapses most often involve consonant sounds, but may also impact production of vowels. For example, a child might collapse the front vowels /i, I, E, æ/ to the central vowel //, creating homonym families such as “bead,” “bid,” “bed,” and “bad,” all produced as “bud.” Furthermore, consonant collapses may occur in the initial, medial, or final positions of words, and may involve singleton consonants and consonant clusters. Multiple oppositions may be used for all of these different types of phonemic collapses. It is a strength of this approach that it is able to address a wide variety of errors, including substitutions (of consonants or vowels), as well as syllable structure errors, such as initial consonant deletion, final consonant deletion, and cluster reduction (Williams, 2010). The examples and forms included in this chapter address consonants in the initial and final positions of words, as these rule sets are the most common. Clinicians are encouraged to modify and adapt the information and materials provided to meet the needs of their students.

Distance Metric:  What is a distance metric? A distance metric is a system described by Williams (2003, 2005b, 2006a) to select specific targets for remediation. Many rule sets, or phonemic collapses, demonstrated by children include a long list of sounds collapsed to just one phoneme. The distance metric will allow us to reduce a large group of affected sounds to fewer targets. In our previous example, we looked at the collapse of /k, ʃ, tS, st/ to the phoneme /t/. In fact, it is likely that this is part of a larger rule set that includes many more consonant singletons and clusters, such as /k, s, ʃ, tS, tr, tw, kl, kɹ, kw, sk, st/ all collapsing to /t/. Let’s see how the distance metric helps us to narrow down this long list of potential targets. Williams (2010) recommends the selection of two to four targets based upon the following two criteria:

4 Multiple Oppositions Intervention

93

1. Maximal classification: Choose targets that are different from each other in terms of place, manner, and voicing. In addition, select targets that differ by their complexity (i.e., singletons vs. consonant clusters). 2. Maximal distinction: Choose targets that are maximally distinct from the child’s error production. Phonemes that vary by place, manner, and voicing are considered to be maximally distinct. Thus, the phonemes we choose will be different from each other (maximal classification) and also different from the error sound (maximal distinction). Looking at our example, if we elect to contrast /t/ with the phonemes /k, ʃ, tS, st/, this will allow us to incorporate contrasts across place (alveolar versus velar and palatal) and manner (stop versus fricative and affricate), as well as to contrast a singleton (/t/) with a cluster (/st/). According to Williams (2010), the purpose of the distance metric is to “expand the relevant frame of learning required” and to “increase the saliency of the contrasts” (p. 84). The selection of targets representing maximal classifications and maximal distinctions is also designed to facilitate phonological reorganization and generalization to untreated phonemes (Williams, 2000a; 2005a). It is interesting to note that targets are not selected based upon characteristics of the sounds in and of themselves, but rather how sounds relate to each other and to the child’s own phonemic collapses (Williams, 2010). In addition, the attention to both the phonetic characteristics of the sounds and their phonemic function creates a holistic approach for the treatment for speech sound disorders.

Research Supporting Multiple Oppositions Intervention Williams (2010) reviewed seven studies of multiple oppositions intervention, including: one controlled study without randomization, (Pagliarin, Mota, & Keske-Soares, 2009), two quasi-experimental studies, (Williams, 2006c, cited in Williams, 2010; Williams & Kalbfleisch, 2001), and four case studies (Cathell & Ruscello, 2004; Liles & Williams, 2006, cited in Williams, 2010; Marcum & Williams, 2005, cited in Williams, 2010; Williams, 2000a). She concluded that the “available evidence base indicates that multiple oppositions is a promising intervention with probable efficacy for creating system-wide change that increases speech intelligibility in children with multiple sound errors” (Williams, 2010, p. 82). More recently, Allen (2013) conducted a randomized controlled study with 54 preschool children. Results supported the efficacy of multiple oppositions and the benefits of a more intensive intervention schedule (i.e., three times per week). In addition, this study demonstrated that multiple oppositions may be effectively implemented in natural work settings by SLPs and SLPAs (Allen, 2013). Table 4–2 provides a summary of selected, relevant research articles.

94 54 children 39 boys; 15 girls

5 children 2 boys; 3 girls

9 children 5 boys; 4 girls

Ceron & KeskeSoares, 2012

Pagliarin et al., 2009

Number of Subjects

Allen, 2013

Authors, Date

4;2 to 6;6 years

4;2 to 8;11 years Mean age 6;1

3;0 to 5;5 years Mean age 4;4

Age Range

3 groups of 3 children, each group with 1 child with a severe phonological disorder, 1 moderate-severe, and 1 mild-moderate, as determined by PCC (Shriberg, Austin, Lewis, McSweeney, & Wilson, 1997)

Phonological disorders of varying severity: severe (2 children); moderate-severe (2 children); mild-moderate (1 child)

Exclusion of childhood apraxia of speech

At least 6 sounds in error across 3 different manner classes

Severity

Table 4–2.  Examples of Research Supporting Multiple Oppositions

Results:  All interventions were effective in treating children with varying degrees of phonological disorder; best results obtained for children with severe and moderate-severe disorders. Multiple oppositions yielded better gains in phonemic inventory and number of distinctive features (other two approaches had better gains in phonetic inventory) for children with severe and moderate-severe disorders.

Approaches:  Comparison of multiple oppositions, minimal oppositions, maximal oppositions/treatment of the empty set, across three severity levels with Brazilian Portuguese–speaking children

Results:  Children demonstrated increases in phonetic and phonemic inventories, and a corresponding increase in speech intelligibility.

Approach:  Multiple oppositions with Brazilian Portuguese–speaking children

Results:  3x per week sessions for 8 weeks yielded significantly better gains than 1× per week for 24 weeks. Higher dose frequency recommended for multiple oppositions intervention.

Approach:  Comparison of multiple oppositions intervention using two different therapy schedules

Results

95

6 children

Williams, 2006c*

4;0 to 6;0 years Mean age = 4;9

Mean age = 5;4

3;5 years

4;6 to 6;5 years Mean age 4;8

Age Range

*Poster sessions presented at international conferences.

Williams & Kalbfleisch, 2001* 14 children 9 boys; 5 girls

1 child

Williams, 2000b

(cited in Williams, 2010)

10 children 8 boys; 2 girls

Number of Subjects

Williams, 2000a

Authors, Date

At least 6 sounds in error across 3 different manner classes

Moderate to severe phonological impairment

Moderate-to-severe speech sound disorder

Results:  Majority of targets showed significant improvement and system-wide change was observed for each child, after 21 sessions or less.

Approach:  Multiple oppositions

Results:  Multiple oppositions resulted in greater systemwide change, including improvement in untreated sound production (i.e., gains in new knowledge). Minimal pairs resulted in greater stabilization of inconsistently produced sounds (i.e., stabilization of existing knowledge).

Approaches:  Multiple oppositions compared with minimal pair therapy

Results:  There was significant improvement on target sounds with multiple oppositions (much greater than with minimal pairs). Also, system-wide change was observed, including improvement on untreated sounds.

Approaches:  Multiple oppositions compared with minimal pair intervention

Results:  Multiple oppositions proved to be time efficient (average length of 105 treatment sessions) and to result in significant increase in percentage of “known” aspects of child’s sound system, particularly for children with profound speech sound disorders. Multiple oppositions concluded to be not well suited for mild phonological impairments; minimal pair therapy and NSI better suited for less severe impairments.

At least 6 sounds in error across 3 different manner classes.

Moderate to severe functional speech disorder

Approaches:  Multiple oppositions, minimal pairs, and naturalistic speech intelligibility training (NSI)

Results

Moderate to profound phonological disorder

Severity

96

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

What Is the Age and Severity of Children Who Could Most Benefit from Multiple Oppositions Intervention? Research studies conducted with English speakers included children ranging in age from 3 to 6½ years, most with moderate or severe speech sound disorders. A Brazilian study (Ceron & Keske-Soares, 2012) included children as old as 8 years, 11 months, and demonstrated that multiple oppositions intervention was effective with Portuguese speaking subjects. Most studies implemented by Williams (e.g., Williams, 2000a; Williams & Kalbfleisch, 2001) included children who exhibited at least six sounds in error across different place and manner classes (indicating at least a moderate speech sound disorder). All of these subjects exhibited typical intelligence and hearing acuity, as well as normal oral-motor structure and function. Many of the participants also demonstrated expressive language delays (Williams, 2010). Multiple oppositions may also be appropriate for children with mixed receptive-expressive language impairment, although studies indicate that outcomes may be more modest for this group of children (e.g.,Williams & Kalbfleisch, 2001). According to Baker (2016), multiple oppositions intervention is appropriate for children with moderate-severe or severe phonological impairment. It is ideal for children presenting with phonemic collapses — that is, many sounds and/or clusters replaced by one phoneme. Their speech will be characterized by limited phonemic inventories and significantly reduced intelligibility (Williams, 2010). Williams (2010) indicates that multiple oppositions may also be appropriate for children who exhibit other subtypes of speech sound disorders with appropriate modifications, including children with childhood apraxia of speech or Down syndrome, but not children whose speech difficulties arise from an organic cause, such as a cleft palate.

Initial Data Collection and Selecting Targets Target selection for the multiple opposition approach is based upon a systemic analysis of a child’s productions, including the use of a unique analysis system for mapping phoneme collapses (Williams, 2003, 2010). These maps result in a description of sound errors from the child’s perspective, and reveal compensatory strategies, including any idiosyncratic error patterns (Williams, 2003, 2010). The following steps for analyzing a child’s disordered speech sound system have been adapted from Williams’ more comprehensive discussion in Speech Disorders Resource Guide for Preschool Children (2003): 1. Collect an extensive speech production sample. 2. Compile a phonetic inventory. 3. Complete a sound distribution form, including substitution and omission errors. 4. Map phoneme collapses. 5. Identify patterns and principles that may describe the child’s phoneme collapses and production of consonant clusters.

4 Multiple Oppositions Intervention

97

Williams has developed an analysis system that is specific to the multiple oppositions approach called the Systemic Phonological Analysis of Child Speech (SPACS; Williams, 2003). This assessment includes both independent and relational components, including a phonetic inventory, a distribution of sounds protocol, and a form to map phoneme collapses. The data needed to complete the SPACS may be obtained from a 245-word list, developed by Williams (2003) that assesses the child’s production of all English phonemes at least five times in each word position. This list is known as the Systemic Phonological Protocol (SPP; Williams, 2003) and was utilized by Williams for her intervention studies. Both the SPP (the 245 word list), and the SPACS (the analysis form), may be found in Williams’ text, Speech Disorders Resource Guide for Preschool Children (2003). Williams (2003, 2010) indicates that the SPACS may also be completed using a smaller sample of single word productions, such as obtained from administration of a standardized speech assessment, but cautions that “a thorough understanding of the child’s sound system requires a minimum of 100 words” (Williams, 2003, p. 62). For our purposes, we will use data obtained from the results of a standardized speech assessment, spontaneous speech samples, and probe lists to complete forms from Chapter 2, which provide similar information to those included in the SPACS. Table 4–3 provides a phonetic inventory (independent analysis) and Table 4–4 a sound production summary (including distribution of phonemes, substitutions, and omissions) for Oscar, age 3 years, 5 months, whose data we will use as an illustration. In addition, we will need in-depth information regarding Oscar’s production of consonant clusters in order to complete the phonemic mapping of his sound system. Therefore, we’ve included an inventory of his cluster productions, with omissions and substitutions noted, in Table 4–5. As we can see in Table 4–3, Oscar demonstrates many sound substitution errors and frequent omission of consonants in clusters and in the final position of words. If we describe these errors using a phonological process framework, Oscar exhibits velar fronting, stopping, gliding, cluster reduction, and final consonant deletion. Rather than using these predetermined categories of error patterns, Williams encourages us instead to look at the child’s system in relation to adult phonology by mapping Oscar’s phoneme collapses. Table 4–6 provides an illustration of how to create such a map using Oscar’s data. When mapping phoneme collapses, Williams (2003) recommends organizing the sounds first by manner, and then by place of production, in order to highlight underlying patterns within the child’s system. As we can see in Table 4–6, Oscar demonstrates several phoneme collapses. He has the following initial sounds in his inventory: /b, p, d, t, w, m, n, j, h/. However, when looking at error patterns, he collapses all other initial singleton consonants to only five phonemes: /b, p, d, t, w/. If we look more closely for organizing principles across these phoneme collapses, we find that the /b/ and /p/ collapses are roughly mirrors of each other, as are the /d/ and /t/ collapses. Voiced, labial obstruents (obstruents include fricatives, stops, and affricates), and clusters collapse to /b/, whereas unvoiced, labial obstruents and clusters collapse to /p/. In a similar manner, voiced, non-labial obstruents and clusters collapse to /d/, whereas unvoiced nonlabial obstruents and clusters collapse to /t/. Finally, liquid singletons and /s/ clusters containing sonorants collapse to /w/. We also need to address Oscar’s phoneme collapses in the final position of words. At the end of words, he omits all fricatives and

Table 4–3.  Independent Analysis for Oscar, Age 3;5 Early Sounds

I

M

F

Stops

Later Sounds

M

Fricatives

p

++++ ++++ ++++ ++++ +++

++++ ++

++++ ++

f

b

++++ ++++ ++++ +++

++++ +++

++++ +++

v

t

++++ ++++ +++

+++

++++ ++

s

d

++

++++ ++++ +

++++ +++

z

k

θ

g

ð

Nasals

ʃ

m

++++

+++

++++

Liquids

n

++++

++++

++++

ɹ

ŋ

l

Glides

Affricates

w

I

++++ ++++ +++

tʃ

++++ ++

j

dZ

Fricative h

+++

+

Consonant Clusters: Initial Medial Final

98

none noted

none noted









none noted

F

Table 4–3.  continued Vowels:  Put a + below the vowel sound if present in the inventory: i

I

eI

E



u

U

oU

ɑ/ɔ/a

+++++

++++++

+++

++

+++++

+++++

+

++++++ ++

+++

/2

/@ ++++++++

aI

aU

ɔI

++++

+++

++

Table 4–4.  Sound Productions for Oscar, Age 3;5 Early Sounds

I

M

F

Stops

Later Sounds

I

M

F

Fricatives

p

3

3

3

f

p

p

Ø

b

3

3

3

v

b

b

Ø

t

3

3

3

s

t

t

Ø

d

3

3

3

z

d

d

Ø

k

t

t

Ø

θ

t

t

Ø

g

d

d

Ø

ð

d

d

Ø

ʃ

t

t

Ø

Nasals m

3

3

3

Liquids

n

3

3

3

ɹ

w

w

@

ŋ

----

n

Ø

l

w

w

@

Glides

Affricates

w

3

3

----

tS

t

t

Ø

j

3

3

----

dZ

d

d

Ø

3

3

----

Fricative h

99

100

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 4–5.  Initial Consonant Cluster Inventory for Oscar Initial Cluster

Child’s Production

Word Examples

Initial Cluster

Child’s Production

Word Examples

*pj*

p

pure, puke

*fj*

p

few, fury

pl

p

play, plow

fl

p

flu, flea

pɹ

p

pray, print

fɹ

p

free, fry

*bj*

b

beauty

*vj*

b

view, viewer

bl

b

blue, blow

sk

t

sky, skate

bɹ

b

bread, bring

sl

w

slow, sled

tɹ

t

true, tree

sm

m

smell, small

tw

t

tweet, twine

sn

n

snow, snack

dɹ

d

draw, dry

sp

p

spy, spoon

*dw*

d

dwarf, dwell

st

t

stay, star

*kj*

t

cue, cute

sw

w

swim, sweep

kl

t

clay, claw

skɹ

w

scrape, scrunch

kɹ

t

cry, crow

skw

w

square, squeeze

kw

t

queen, quack

spl

w

split, splash

gl

d

glow, glue

spɹ

w

spring, sprout

gɹ

d

grow, gray

stɹ

w

stray, stripe

*mj*

m

mule, music

ʃɹ

t

shred, shrank

θɹ

t

throw, three

*Indicates those consonant clusters with relatively few word examples.

affricates, as well as the stops /k/ and /g/. This collapse is represented in Table 4–4 by showing that the target sounds are produced as the empty set (Ø). As predicted by Williams (2010), we have discovered a logical structure to the child’s disordered phoneme system. We have now completed the five steps of data analysis outlined on pp. 82–83 and p. 96, including compilation of a phonetic inventory (see Table 4–3), completion of a sound distribution form, including substitution and omission errors (see Tables 4–4 and 4–5), mapping of phoneme collapses (see Table 4–6), and identification of patterns that may describe the child’s organizational structure. Now we will turn our attention to target selection. Understanding the child’s unique rule sets through the mapping of phoneme collapses will guide our selection of intervention targets. We have mapped five phoneme collapses for Oscar affecting the initial production of consonants, and one that impacts

101

4 Multiple Oppositions Intervention

Table 4–6.  Phoneme Collapse Mapping (Consonants) for Oscar

Table 4–6. Phoneme Collapse Mapping (Consonants) for Oscar

PHONEME COLLAPSE MAPPING (CONSONANTS) PHONEME COLLAPSE MAPPING (CONSONANTS) Word-Final Collapses

Word-Initial Collapses Word-Initial Collapses b

p

k

t

f

%

bj

k

pj

(

bl

s

pl

f

b!

"

p!

v

vj

#

fj

d

$

fl

z

%

t!

f!

#

tw

sp

v b

z

d

!!

t

p

Ø

s

l

&

&

kj

!

"

'

kl

sl

$

d!

k!

sw

'

dw

kw

%l

sk

skw

%!

st

spl

"!

sp!

#!

st!

w

sk!

the final position (see Table 4–6). We must now decide which collapses will receive priority in therapy. Williams (2010) employs the metaphor of a puzzle to represent a child’s disordered phonological system, and advocates for presenting the child with the corners of the puzzle in order that they might fill in the missing pieces (i.e., the missing contrasts). In Oscar’s case, we see that four of his collapses (i.e., /b, p, d, t/) are organized around voiced versus unvoiced and labial versus nonlabial distinctions. If

Word-Final Collapse

102

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

we select sound families to represent these parameters, we will create an appropriate framework for Oscar’s phonological reorganization. Specifically, if we create homonym families that incorporate the /p/ and the /d/ collapse, Oscar will have the opportunity to experience sound contrasts that incorporate voiced, unvoiced, labial and non-labial features (alternatively, selection of the /b/ and /t/ collapses would achieve this same goal). In addition, it would be advisable to create word families to represent the /w/ collapse, as it represents additional contrasts that are missing from Oscar’s current system. Metaphorically, we are addressing the corners of Oscar’s phonological puzzle, so he may fill in the missing pieces in a systematic manner. We have identified four phonemic collapses to use in the initial round of intervention, including the /d/, /p/, /w/ collapses in the initial position of words and the collapse to the empty set that occurs for Oscar in the final position of words. Now we must select phonemes to specifically target within each collapse, employing the distance metric described earlier. As an example, we see that the following phonemes are produced by Oscar as /d/: /d, g, z, ð, dZ, dɹ, gl, gɹ/. How will we narrow this group of phonemes down to the two to four needed to create our word families? Remember that Williams (2005) recommends the selection of target phonemes that are maximally distinct from the error and represent maximal classification — that is, the targets represent different categories of place, manner, and voicing. In addition, we are advised to include a singleton versus cluster distinction. If we apply the distance matrix to Oscar’s /d/ collapse, we might choose the sounds /g, ð, dZ, gl/ from the longer list of sounds produced as /d/ (i.e., /d, g, z, ð, dZ, dɹ, gl, gɹ/). The four targets selected satisfy the distance metric because they differ from each other in terms of place and manner of production, and they also represent maximal distinctions from the error phoneme /d/. Our chosen targets include a velar stop (/g/), an interdental fricative (/ð/) and a palatal affricate (/dZ/), and are thus distinct from Oscar’s error sound /d/, which is an alveolar stop. In addition, we’ve included a consonant cluster (/gl/). Because Oscar’s /d/ phoneme collapse includes only voiced elements, we are unable to incorporate a voiced-voiceless contrast in our target selection. However, we have included a variety of other contrasts representing differences across place, manner, and syllable complexity (i.e., singleton vs. cluster). This group of sounds — /d/ contrasted with /g, ð, dZ, gl/ — will form one goal area for Oscar, as it represents one rule set, or phoneme collapse, that must be remediated in order to decrease homonymy and bring Oscar’s phonemic system in closer alignment with typical adult phonology. Now let’s apply the distance matrix to all four of the phoneme collapses we’ve selected for Oscar, before moving on to create our word families. As listed in Table 4–7, we might select three phonemes to address the smaller groups that collapse to initial /p/ (/p/ vs. /f, pl, sp/) and initial /w/ (/w/ vs. /l, ɹ, sl/), and four to facilitate production of final consonants (Ø vs. /k, v, ʃ, dZ/). In each case, phonemes have been selected that represent maximal classifications from each other and are maximally distinct from the error sound. These four groups of targets will be presented to Oscar in the form of contrastive word families (e.g., doe, go, though, Joe, glow). Through the wide variety of contrasts represented by these groups of sounds, Oscar will achieve optimal reorganization of his disordered phonological system.

4 Multiple Oppositions Intervention

103

Table 4–7.  Selected Targets for Oscar’s Phoneme Collapses Error Sound

Target #1

Target #2

Target #3

Target #4

Rule Set #1

/d/ initial

/g/

/ð/

/dZ/

/gl/

Rule Set #2

/p/ initial

/f/

/pl/

/sp/

Rule Set #3

/w/ initial

/l/

/ɹ/

/sl/

Rule Set #4

Ø (final)

/k/

/v/

/ʃ/

/dZ/

Target Word Selection: Multiple Opposition Contrastive Sets Contrastive word sets, incorporating the target phonemes, form the foundation of multiple opposition therapy. Oscar will be presented with groups of words that include missing phonemic contrasts that he must master in order to accurately produce the different target vocabulary and avoid producing homonyms. For example, when targeting Oscar’s /w/ collapse, we might present him with pictures representing the words “wait,” “late,” “rate,” and “slate.” These words were chosen using the sounds selected according to Williams’ distance metric (see Table 4–7). According to our maps of Oscar’s sound collapses (see Table 4–6), we expect that Oscar will collapse the phonemes /l/, /ɹ/, and /sl/ to /w/ and produce all four words as “wait,” and thus, as homonyms. During the phases of multiple opposition therapy, Oscar will be given the phonetic and phonemic knowledge required to successfully differentiate these words. For each phoneme collapse and corresponding group of targets, we will create five contrastive word sets to utilize in therapy activities. Table 4–8 provides a list of word examples chosen to address Oscar’s rule sets. Each group of five word sets forms a treatment set for that particular phoneme collapse. When choosing target words, it is advisable to use simple word shapes, such as consonant-vowel ( CV) or consonant-vowel-consonant ( CVC), and to incorporate a variety of vowel sounds. Williams (2010) promotes the use of real words whenever possible. However, in order to create five contrastive word groups it is sometimes necessary to incorporate proper nouns or even nonsense words. For example, the first word family for the /d/ collapse includes the name Joe and the second family includes the nonsense word “gee” (see Table 4–8). When using made-up words, it is best to give the word a meaning, such as the name of a character or an action, so that the target may be visually represented in a picture. An excellent resource for therapy stimuli is Sound Contrasts in Phonology (SCIP), created by A. Lynn Williams (2006b). Originally created as a CD-ROM for computers, SCIP is now available as an app for iPads. It includes pictures of treatment targets, representing both real and nonsense words, and containing both consonant and vowel contrasts. The stimuli may be used for a variety of contrast therapy approaches, including multiple oppositions, maximal oppositions, treatment of the empty set, and minimal pair therapy. Premade therapy materials for multiple oppositions are scarce, but clinicians

Table 4–8.  Contrastive Multiple Opposition Sets for Oscar Contrastive Multiple Opposition Sets Treatment Set #1 for Rule Set #1:  /d/ (initial) versus /g, ð, dZ, gl/ Error

Target #1

Target #2

Target #3

Target #4

/d/ (initial)

/g/

/ð/

/dZ/

/gl/

Word Set 1

doe

go

though

Joe

glow

Word Set 2

“D”

gee

thee

“G”

glee

Word Set 3

day

gay

they

“J”

glay

Word Set 4

den

gen

then

Jen

glen

Word Set 5

dot

got

thot

jot

glot

Treatment Set #2 for Rule Set #2: /p/ (initial) versus /f, pl, sp/ Error

Target #1

Target #2

Target #3

/p/ (initial)

/f/

/pl/

/sp/

Word Set 1

pay

Faye

play

spay

Word Set 2

pea

fee

plea

spee

Word Set 3

pot

fought

plot

spot

Word Set 4

pat

fat

plat

spat

Word Set 5

pine

fine

pline

spine

Treatment Set #3 for Rule Set #3: /w/ (initial) versus /l, ɹ, sl/ Error

Target #1

Target #2

Target #3

/w/ (initial)

/l/

/ɹ/

/sl/

Word Set 1

wait

late

rate

slate

Word Set 2

why

lie

rye

sly

Word Set 3

white

light

right

slight

Word Set 4

wick

lick

Rick

slick

Word Set 5

whip

lip

rip

slip

Treatment Set #4 for Rule Set #4: Ø (final) versus /k, v, ʃ, dZ/ Error

Target #1

Target #2

Target #3

Target #4

Ø (final)

/k/

/v/

/ʃ/

/dZ/

Word Set 1

baa

back

bav

bash

badge

Word Set 2

paw

pock

pov

posh

podge

Word Set 3

way

wake

wave

washe

wage

Word Set 4

hi

hike

hive

hishe

hidge

Word Set 5

pay

pake

pave

pashe

page

104

4 Multiple Oppositions Intervention

105

may compile appropriate contrastive word families by using stimuli designed for minimal pair and maximal opposition intervention. Caroline Bowen’s website (http:// www.speech-language-therapy.com) includes word lists and picture stimuli for minimal opposition and maximal opposition pairs. In addition, of course, creating your own materials is easier than ever with the Internet. After brainstorming appropriate word families for each phonemic collapse, a few minutes spent cutting and pasting images found through Google or Boardmaker will be time well spent. The extra effort of making stimuli to address a child’s specific needs will likely be rewarded by more rapid progress toward production goals.

Establishing Treatment Goals Ongoing data collection is a vital component of the multiple oppositions approach (Baker, 2016; Williams, 2010). Baseline data is taken on each contrastive word set prior to the initiation of therapy and treatment data on these same words is gathered during each session. In addition, generalization probes are administered throughout the program. The first step is to create a data collection sheet for each treatment set that may be used to record baseline, treatment, and generalization data (Baker, 2016). Table 4–9 provides an example of a probe created for Oscar’s first treatment set, which addresses the rule set of /d/ versus /g, ð, dZ, gl/. The first section of the probe includes all 25 words contained in the five multiple opposition contrastive sets created for this phoneme collapse (first listed in Table 4–8). Baseline data will be taken on 20 of these words (remember, we do not track data on the words representing the error sound — in this case, /d/). The second section of the probe includes 10 generalization words for each of the target sounds (i.e., /g, ð, dZ, gl/). Baseline data will also be gathered for these generalization words. Thus, our baseline probe for each sound will include 15 words — one word from each of the five contrastive word sets and 10 additional generalization words. The baseline for the target /g/ will include the five contrastive set targets (go, gee, gay, gen, got) and the ten additional baseline words (game, get, good, gate, guy, goat, gum, gut, give, gas). Ideally, all words will be presented to the child in the form of picture stimuli and baseline data will reflect spontaneous naming of the stimulus cards. As can be seen in Table 4–9, Oscar produced 3/15 of the initial /g/ baseline words correctly, yielding a 20% accuracy rate prior to the start of treatment. This information will be included as the baseline data when writing goals for intervention, and will allow us to track progress appropriately. It should be noted that there are two types of generalization data gathered during the course of multiple oppositions intervention. The 10 generalization words included in our sample probe for Oscar will be administered periodically to assess carryover to untreated words. In addition, the use of target sounds will be sampled during conversational speech to monitor generalization to spontaneous production across a variety of contexts. Williams (2010) refers to the first type of data as a “narrow generalization measure” and the speech samples as a “broad generalization measure” (p. 87). As indicated previously, the remediation of each phoneme collapse (or rule set) will become a separate goal of intervention. In Oscar’s case, we’ve selected four rule sets

Table 4–9.  Baseline and Generalization Probe for Oscar Multiple Oppositions:  Baseline & Generalization Probe Treatment Set #1 for Rule Set #1:  /d/ (initial) versus /g, ð, dZ, gl/ Baseline words (also treatment words) Error

Target #1

Target #2

Target #3

Target #4

/d/

/g/ +/−

/ð/ +/−

/dZ/ +/−

/gl/ +/−

Word Set 1

(doe)

go +

though −

Joe −

glow −

Word Set 2

(“D”)

gee −

thee −

“G” −

glee −

Word Set 3

(day)

gay −

they −

“J” +

glay −

Word Set 4

(den)

gen −

then −

Jen −

glen −

Word Set 5

(dot)

got −

thot −

jot −

glot −

Baseline

Date: 3/1

1/5 = 20%

0/5 = 0%

1/5 = 20%

0/5 = 0%

Treatment

Date: 3/3

5/10 = 50%  I

3/10 = 30%  I

5/10 = 50%  I

2/10 = 20%  I

Treatment

Date: 3/8

7/10 = 70%  I

5/10 = 50%  I

7/10 = 70%  I

4/10 = 40%  I

Treatment

Date: 3/10

7/10 = 70%  I

6/10 = 60%  I

8/10 = 80%  I

4/10 = 40%  I

Treatment

Date: 3/15

3/10 = 30% S

7/10 = 70%  I

4/10 = 40%  S

6/10 = 60%  I

Target #1

Target #2

Target #3

Target #4

/g/ +/−

/ð/ +/−

/dZ/ +/−

/gl/ +/−

game −

the −

gym +

glad −

get +

that −

jack −

glam −

good −

them −

jam +

glide −

gate −

these −

jeep −

glue −

guy +

there −

jet −

globe −

goat −

this −

jar −

gleam −

gum +

thy −

jelly −

glib −

gut −

those −

jeer −

glass −

give −

than −

joy +

glare −

gas −

thine −

jut −

glut −

3/15 = 20%

0/15 = 0%

3/15 = 20%

0/15 = 0%

Baseline Words (also generalization words)

Baseline

Date: 3/1

Generalization

Date:

Generalization

Date:

Generalization

Date:

I = Imitation; S = Spontaneous

106

4 Multiple Oppositions Intervention

107

to target initially, including the initial /p/, /d/, and /w/ collapses, as well as the final collapse to the empty set, so we might logically develop four goal areas, as follows: Goal Area #1:  initial /d/ versus /g, ð, dZ, gl/ Goal Area #2:  initial /p/ versus /f, pl, sp/ Goal Area #3:  initial /w/ versus /l, ɹ, sl/ Goal Area #4:  final /k, v, ʃ, dZ/ Let’s use the first phoneme collapse (represented by Goal Area #1) as an example for writing a possible treatment goal. We will write the intervention goal to target the specific phonemes to be taught. These will then serve as the vehicle for increasing the use of contrastive features and decreasing the occurrence of homonymy. Table 4–10 presents a sample goal for Oscar.

Beginning Therapy Phases of Treatment Williams (2000a, 2003, 2010) outlines a treatment paradigm for multiple oppositions that includes several phases of therapy, as well as specific criteria for target progression and discontinuation. Generally speaking, as a word set moves through the stages of treatment, the emphasis shifts from phonetic learning to phonemic learning, and from imitative productions to the spontaneous use of words in more natural contexts. Each phase below is implemented for a goal area using the entire treatment set, which includes all contrastive word sets created for a given phoneme collapse (i.e., rule set). For example, if we begin therapy with Oscar by targeting the /d/ phoneme collapse, we will need to have 25 different picture stimuli, including five pictures for each of the five contrastive word sets created. We will likely want to make at least two copies of each stimuli to use in matching games (Baker, 2016).

Phase 1:  Familiarization and Initial Production of Contrasts Step 1:  Familiarization with the Rule and Sounds.  The first step of intervention is to introduce the child to a rule set, which includes the targets and their contrastive differences Table 4–10.  Sample Goal for Oscar: Multiple Oppositions Goal Area #1:  initial /d/ versus /g, ð, dZ, gl/ Baseline #1:  Oscar produced /g/ and /dZ/ in the initial position of single words with 20% accuracy, and /ð/ and /gl/ in the initial position of single words with 0% accuracy, in the context of spontaneously naming picture stimuli during administration of a baseline probe. Goal #1:  Oscar will produce /g/, /ð/, /dZ/, and /gl/ in the initial position of single words with 70% accuracy, in the context of spontaneously naming pictures.

108

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

(Baker, 2016; Williams, 2010). Thus, we will begin by introducing the phonemes /d, g, ð, dZ, gl/ to Oscar. He needs to learn about each sound and how it differs from the others. We might say, “the /d/ sound is produced in the front of the mouth, whereas the /g/ sound is made in the back” or “the /d/ sound is a short sound, whereas the /ð/ sound is a long sound.” Clinicians are encouraged to use metaphors and gestures to illustrate the features of each target sound. Step 2:  Familiarization with the Pictures and Vocabulary.  Next we must ensure that Oscar is familiar with the picture stimuli and the vocabulary represented by each. This step is particularly important because the word families created for multiple oppositions therapy often include proper nouns and nonsense words. While the child listens, the clinician will give a short explanation of each picture or tell a little story to make the word memorable. For example, we might say, “This is a doe. She is a mommy deer. Bambi was a fawn, and her mom was a doe.” or “This is Joe. He’s my uncle and he’s a fireman. He’s Uncle Joe.” A short game of pick-up, in which the clinician names the cards in random order and the child must pick up the correct card, will confirm that the child has become familiar with the pictures and vocabulary. Step 3:  Initial Production of Contrasts.  Now Oscar will be ready to attempt production of the target words. Baker (2016) recommends that this initial attempt is preceded by a clinician model. Then, the child and clinician produce the target simultaneously. Maximum cueing should be provided to support success. The metaphors and gestures introduced during the sound familiarization step will now be helpful as cues for accurate motoric production of the targets. However, keep in mind that this is the child’s first attempt and that we are not looking for perfect sound production at this point in the therapeutic process (Baker, 2016).

Phase 2:  Production of Contrasts and Interactive Play Step 1:  Imitation. During this stage of intervention, Oscar will produce each word after the clinician’s model. He will imitate 20 target words at least two times each, for a total of 40 productions (remember, we have a total of 25 pictures, but we do not count the five pictures that begin with the error sound /d/). Our focus during the imitation phase is on teaching Oscar the phonetic aspects of the sound contrasts, including the appropriate place, manner, and voicing required for accurate production. Clinicians are again encouraged to use hand cues and metaphors to support the child’s learning. The goal of this phase is for the child to establish a new motor pattern for accurate physical production of the sound. Therefore, this stage of therapy should be comprised of focused, intensive practice with high response rates from the child to facilitate automaticity (Williams, 2010). Treatment data should be gathered during every session. The sound probe developed for Oscar will allow us to track his progress (see Table 4–9). When he achieves at least 70% accuracy for two consecutive treatment sessions (producing each target word twice), intervention proceeds to the spontaneous production phase (Williams, 2010). Note that the different targets may reach the 70% accuracy level at varying times. We see on Oscar’s data collection sheet (see Table 4–9) that he achieved at least 70% accuracy during imitative productions (indicated by the letter I on the probe) for two consecu-

4 Multiple Oppositions Intervention

109

tive sessions on two target phonemes, /g/ and /dZ/, after only three treatment dates. The words with these phonemes can be moved to the next step, spontaneous productions. The words with the other two targets, /ð/ and /gl/, will continue to be practiced while imitating the clinician, until they, too, reach 70% accuracy for progression to the next step. Step 2:  Spontaneous Production.  Now that the child has acquired the phonetic knowledge needed to accurately produce the targets imitatively, the focus of intervention will shift to learning phonological contrasts. Oscar will be asked to spontaneously name the pictures and must achieve the necessary contrasts between the target sounds. In a similar fashion to the production phase of conventional minimal pair therapy, we will provide the child with feedback regarding the miscommunications that occur when contrasts are not produced accurately. For example, if Oscar says “doe” when presented with the picture of “Joe,” we might say “Doe? Is this a mommy deer? Or did you mean Joe? My Uncle Joe?” Our cueing has shifted from providing information about the phonetic characteristics of the sounds to providing feedback regarding the phonological contrasts and their impact on communication. When the child reaches an accuracy level of 90% across two consecutive treatment sessions, naming all pictures in the set twice, intervention proceeds to Phase 3: Contrasts Within Communicative Contexts (Williams, 2010). Step 3:  Interactive Play.  Before proceeding to the third phase of intervention, Williams (2010) recommends engaging in a short period of interactive play with the child. Play allows the child to be exposed to the target contrasts in a more naturalistic context. We might find toys or activities to represent the target words. For example, we could use a plastic deer to be the “doe” and a firefighter figure to be “Joe.” During pretend play, we could talk about a forest fire that “glows” and how “Joe” wants to “go” quickly to save the “doe,” “though” his firetruck is out of gas and he must “go” to the gas station first. In fact, words such as “though” are often easier for a child to understand in the context of play rather than through pictures. This period of interaction provides children with the opportunity to hear and use the new contrasts in natural contexts that may facilitate generalization (Baker, 2016).

Phase 3:  Contrasts Within Communicative Contexts During the third phase of multiple oppositions therapy, the child continues to produce the target words spontaneously, in the context of structured, interactive games. The pictures no longer need to remain in their sets, but may be mixed up in games such as Go Fish or Concentration (Baker, 2016; Williams, 2006a). This provides an opportunity for more distributed practice as the child is required to achieve appropriate phonemic contrasts in a random order. In addition to gathering data on the treatment words, generalization probes will be administered every third session. Using the data collection form created at the beginning of the therapy process, the 10 generalization words created for each target sound will be presented to the child (see Table 4–9 for Oscar’s sound probe). This use of the untreated word lists as a narrow generalization measure will begin during the second phase of therapy (Williams, 2010). If the child achieves 90% accuracy on a given target sound in the untrained words, then the clinician will check for accuracy in a conversational

110

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

speech sample, thus completing the broad generalization measure (Williams, 2010). If the child demonstrates at least 50% accuracy on the target sound during spontaneous conversation, then treatment on that sound will be terminated. If not, then the next phase of intervention will be initiated.

Phase 4:  Conversational Recasts A conversational recast strategy will be introduced when a child: (1) has achieved 90% accuracy on production of a target sound during administration of the generalization probe, and (2) is not yet using the sound accurately in conversational speech in at least half of the opportunities presented. This phase of intervention is based upon the naturalistic speech intelligibility training developed by Camarata (1993, 1995), in which the clinician recasts the child’s incorrect productions of the target sound during natural activities. Similar to the Interactive Play stage of Phase 2, we will now create opportunities for the child to produce target sounds and contrasts in the context of play or familiar routines. It is important to structure the activities to allow for frequent opportunities for target production (Williams, 2006a). For example, if Oscar was failing to generalize the phoneme /g/ to conversational speech, we might play a stop and “go” game with toy animals (including the “doe”) that requires him to frequently direct the clinician and the animals to stop or “go.” When an error occurs, we will recast the production for Oscar (“Oh, you want me to go.”). Williams (2006a) also suggests the use of nonverbal cues as needed to assist accurate phonetic production during this phase. However, we want to fade cues over time to facilitate increasing independence. In fact, Williams (2010) advises that throughout a multiple oppositions program, clinicians should gradually decrease modeling support, while gradually increasing situational complexity in order to facilitate generalization. During this final phase of intervention, we will continue to periodically check for generalization in the context of spontaneous conversation. As soon as 50% accuracy has been achieved, we may discontinue intervention on this specific target. Please note that the phases described above, along with criteria for target progression, are visually represented in a flowchart titled “Treatment Paradigm,” created by Williams (2000b; reprinted in Williams, 2010).

Home Program One study conducted by Williams (2003) included a parent-administered program as a component of the multiple oppositions intervention. Stimulus cards were sent home and parents were asked to elicit 10 to 20 responses during short turn-taking activities. Parents were trained in how to model the target sounds and recast any errors produced by the child (Williams, 2003). The inclusion of such a home program may yield positive benefits and should be considered when implementing the multiple oppositions approach.

Goal Attack Strategy How and when do we move from one goal area to the next? For example, when do we stop working on the /d/ phoneme collapse and begin intervention on the /p/ phoneme

4 Multiple Oppositions Intervention

111

collapse for Oscar? There is not a simple answer to this question and clinicians will need to rely somewhat on their clinical judgment. Williams utilized a variety of goal attack strategies in research studies, including vertical, horizontal, and cyclical. Currently, there is no research comparing the different options. However, Williams (2000a) proposed employing different goal attack strategies as a treatment set progresses through the phases of intervention, beginning with a vertical strategy to establish accuracy and then employing a horizontal or cyclical approach to promote generalization. In practice, this might involve focused practice on the /d/ collapse for Oscar until he reaches the third or fourth phase of intervention, then a shift to focused practice on the /p/ collapse while continuing to incorporate less intensive work on the /d/ word sets. As each new treatment set is introduced over time, the previous sets could continue to be incorporated through periods of interactive play, structured games, or conversational recast intervention, until the generalization criteria for each sound target is met. Will Oscar’s treatment program be complete once all four treatment sets have met the generalization criteria of 50% accuracy during conversational speech samples? We hope the answer is yes. Our assumption is that resolution of the four targeted phoneme collapses will stimulate phonological reorganization throughout Oscar’s disordered system. Ideally, we will not need to directly address the other two phoneme collapses initially identified (i.e., the /b/ and /t/ collapses). However, only time and continued periodic assessment will tell. If Oscar continues to demonstrate extensive phonemic collapses, we will want to create new treatment sets to address any remaining rule sets. Perhaps instead he may exhibit a few remaining singular phoneme collapses (e.g., /t/ for /tS/) that could be more appropriately addressed through another contrast intervention approach, such as minimal pair therapy (Baker, 2010; see Chapter 3) or maximal opposition therapy (Gierut, 1990; see Chapter 5).

Group Therapy Multiple oppositions therapy may be successfully implemented in the context of group therapy, with some forethought and preparation. Because targets are selected based upon unique phoneme collapses, each child will have his own picture cards representing individualized contrastive word sets. In addition, each word set will likely be at a different stage of intervention. For example, a child may have three target word sets, one that is being practiced through conversational recasts (Phase 4), one that is ready for production within communicative contexts (Phase 3), and one that is just being introduced (Phase 1). In order to facilitate multiple oppositions therapy in a group setting, clinicians will need a strategy for tracking the progression of students’ word sets through the phases of intervention. This might be accomplished through a wall chart or perhaps most easily, by keeping each word set in its own plastic bag, with the student’s name written on the outside. As a particular group of pictures moves from one phase of intervention to the next, this could be marked on the outside of the bag. Let’s now look at some ideas for group therapy activities that may be utilized at each stage of multiple oppositions intervention.

112

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Phase 1: Familiarization This initial phase requires the most focused attention from the clinician and the child. Therefore, therapists may want to work one on one with a child when introducing a new contrastive word set. It’s important to keep in mind that the familiarization phase of multiple opposition therapy involves more elements than the same phase of conventional minimal pair therapy. Both include familiarization with the target vocabulary words, but multiple opposition also includes reviewing the features of the sounds, explaining the rules that differentiate the sounds, and finally, facilitating the actual first productions of sound targets. This work will likely necessitate some individual attention. It may be possible for clinicians to schedule short one on one sessions for this purpose. If not, other children in a group therapy session may engage in interactive games with previous word sets that have been promoted to Phase 3, while the clinician gives focused attention to an individual child learning a new word set.

Phase 2: Imitation and Spontaneous Production The second phase of multiple opposition therapy involves the naming of picture cards in both imitative and spontaneous contexts. Ideally, clinicians will present tasks that promote a high density of responses from each child in the group. Therefore, activities should be fast moving and simple to learn, with the naming of cards as the primary focus. Avoid games with complicated rules, long wait times, or too many distracting materials. Although creativity is always desirable, it is important to remember that a high number of responses are most often achieved through the simplest of games. Here are some ideas for appropriate activities that will engage students, while allowing for maximum sound production practice. Ask the child to first name the stimulus item and then: 1. Feed the stimulus cards to a puppet or stuffed animal. 2. Place the stimulus cards into a container, such as a mailbox or train car. 3. Use the stimulus cards to build a street for toy cars, an airport runway, a road to a castle, or a path to a jar containing a snack (Hammer, 2014). 4. Place each card on the floor and leap over it. The child who is able to leap over the most cards, placed end to end, wins the game (Hammer, 2014). Distribute the stimulus cards on the floor around the room and then ask the child to: 1. Stand on a picture, say the word, and throw a ball into a hoop or a beanbag into a container. 2. Drop or throw an object (e.g., a plastic spider, a coin, or a sticky ball) onto a picture and then say the word.

4 Multiple Oppositions Intervention

113

Hide the cards around the room or inside objects (e.g., plastic eggs) and then ask the child to: 1. Find a card and say the word. 2. Shine a light on the card with a flashlight and say the word. 3. One child directs another child to the card using the word in a sentence (e.g., “Go get the key that is under the table.”). Keep in mind that it is possible for a given target to move from imitation to spontaneous production during the same activity. Once the child has achieved 70% accuracy when imitating words (across two consecutive treatment sessions), it is time to require spontaneous productions. Therefore, an efficient system for data collection is important for group sessions. Older students may be able to track their own data or each other’s, if given instruction on how to complete the data sheets. Assigning this responsibility to children may have the added benefit of facilitating self-awareness and motivation.

Phase 3: Contrasts Within Communicative Contexts The third phase of multiple opposition therapy emphasizes the production of words during communicative exchanges. Activities will therefore include an interaction between the child and clinician or between two children, as an essential component. Similar to the production phase of conventional minimal pair therapy, misproductions of words should result in a breakdown of communication during the activity, to highlight the natural consequences of producing target words as homonyms. Games that involve a child requesting an object or directing another person will achieve this goal. Here are some ideas for games to utilize during Phase 3: Requesting games: 1. Go Fish card game:  each child must ask for desired pictures. The child asks, “Do you have _____?” 2. Shopping:  each child asks the cashier (another student) for the cost of each item represented on a stimulus card (Richard, 2017). The child asks, “How much does _____ cost?” 3. Restaurant:  each child asks for desired items from a server (another student). The child asks, “May I have ______?” Directing games:  each child directs the clinician or another child to do the following: 1. Pin the tail on a specific stimulus card. Child says, “Pin the tail on the _____.” 2. Roll a ball toward a bowling pin with a specific stimulus card taped to it. Child says, “Hit the ______.”

114

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

3. Dot paint a specific picture of a target word. Child says, “Paint the _____.” 4. Jump on a specific stimulus card. Child says, “Jump on the ______.”

Phase 4: Conversational Recast As you will recall, this final phase of multiple opposition therapy is only utilized if a child is failing to generalize a new skill to conversational speech. Recasts are provided by the clinician during play activities and familiar routines that have been structured to promote the frequent production of target sounds. Group therapy provides the perfect opportunity for children to practice new sound production skills in the context of natural communication interactions with peers, while the clinician gradually fades support. Here are a few ideas for Phase 4: 1. Storytelling:  Place 3 to 4 stimulus cards on the table and ask each child, in turn, to make up one sentence of a silly story (Richard, 2017). 2. Engage in interactive, pretend play that incorporates objects and actions representing the target words. 3. Use topic cards to stimulate conversation among older students. Such interactions will also provide an opportunity to probe for generalization. Although working with multiple students simultaneously may be a challenge, group therapy also provides opportunities for children to use new skills in natural contexts that will promote generalization.

Case Study Initial Data Collection and Selecting Targets The following presents initial data collection for Jasmine, age 4;9, including a phonetic inventory (independent analysis), a summary of her sound productions, and an initial consonant cluster inventory. Information was obtained from the results of a standardized speech assessment, spontaneous speech sample, and probe lists. Refer to Tables 4–11, 4–12, and 4–13. The first step in creating an appropriate multiple opposition program for Jasmine will be to map her phoneme collapses in an effort to discover her unique rule sets. Table 4–14 presents a visual representation of Jasmine’s phoneme collapses. We can see that in the initial position, Jasmine has one very extensive rule set that involves the substitution of /h/ for all singleton fricatives, affricates, and liquids, as well as clusters containing a fricative combined with a glide or liquid. She also demonstrates a collapse to /t/ which includes singleton /k/ and clusters containing /t/ or /k/.

Table 4–11.  Independent Analysis for Jasmine, Age 4;9 Early Sounds

I

M

F

Stops

Later Sounds

I

M

F

Fricatives

p

++++ ++++ ++++

++++

++++ ++

f

b

++++ ++++ ++

+++

++++ +++

v

t

++++ ++++ ++++ ++

+++

++++ ++

s

d

++++ ++++ +

++++

++++ +++

z

k

θ

g

ð

Nasals

ʃ

m

++++

+++

++++

Liquids

n

++++

+++

++++

ɹ

ŋ

l

Glides

Affricates

w

++++ ++

+++

tʃ

j

+++

++

dZ

++++ ++++ ++++ ++++ ++++ +++

++++ ++++ ++

Fricative h

continues

115

Table 4–11.  continued Consonant Clusters: Initial Medial Final pj none noted none noted bj





tw





dw mj Vowels:  Put a + below the vowel sound if present in the inventory: i

I

eI

E



u

U

oU

ɑ/ɔ/a

+++++

++++++

+++

++

+++++

+++++

+

++++++ ++

+++

/2

/@ ++++++++

aI

aU

ɔI

++++

+++

++

Table 4–12.  Sound Productions for Jasmine, Age 4;9 Early Sounds

I

M

F

Stops

I

M

F

Fricatives

p

3

3

3

f

h

h

p

b

3

3

3

v

h

h

b

t

3

3

3

s

h

h

t

d

3

3

3

z

h

h

d

k

t

t

t

θ

h

h

t

g

d

d

d

ð

h

h

d

ʃ

h

h

t

Nasals m

3

3

3

Liquids

n

3

3

3

ɹ

h

h

@

ŋ

----

n

n

l

h

h

@

Glides

Affricates

w

3

3

----

tS

h

h

t

j

3

3

----

dZ

h

h

d

3

3

----

Fricative h

116

Later Sounds

Table 4–13.  Initial Consonant Cluster Inventory for Jasmine Initial Cluster

Child’s Production

Word Examples

Initial Cluster

Child’s Production

Word Examples

*pj*

3

pure, puke

*fj*

DNT

few, fury

pl

p

play, plow

fl

h

flu, flea

pɹ

p

pray, print

fɹ

h

free, fry

*bj*

3

beauty

*vj*

DNT

view, viewer

bl

b

blue, blow

sk

t

sky, skate

bɹ

b

bread, bring

sl

h

slow, sled

tɹ

t

true, tree

sm

m

smell, small

tw

3

tweet, twine

sn

n

snow, snack

dɹ

d

draw, dry

sp

p

spy, spoon

*dw*

3

dwarf, dwell

st

t

stay, star

*kj*

t

cue, cute

sw

w

swim, sweep

kl

t

clay, claw

skɹ

t

scrape, scrunch

kɹ

t

cry, crow

skw

t

square, squeeze

kw

t

queen, quack

spl

p

split, splash

gl

d

glow, glue

spɹ

p

spring, sprout

gɹ

d

grow, gray

stɹ

t

stray, stripe

*mj*

3

mule, music

ʃɹ

h

shred, shrank

θɹ

h

throw, three

*Indicates those consonant clusters with relatively few word examples. DNT = Did not test

117

Table 4–14.  4–14. Phoneme Collapse Mapping (Consonants) for Jasmine Table Phoneme Collapse Mapping (Consonants) for Jasmine PHONEME COLLAPSE MAPPING (CONSONANTS)

PHONEME COLLAPSE MAPPING (CONSONANTS)

Word-Initial Collapses Word-Initial Collapses h

t

f

k

v

t!

s

tw

z

kj

"

h

t

kl

#

k!

$

kw

l

sk

!

st

%

d

sk!

&

'

skw

d!

st!

fl

d

f!

'l

sl

'!

p pl

p

p!

$!

spl

"!

sp!

Word-Final Collapses Word-Final Collapses

t

t

d

k

'

s

z

"

#

$

&

%

118

d

4 Multiple Oppositions Intervention

119

A mirror collapse also exists to /d/ which includes singleton /g/ and clusters containing /d/ or /g/. Finally, there is another small collapse to /p/ involving blends containing the /p/. The collapses to /t/ and /d/ are also found in the final position of words for Jasmine, as illustrated in her phoneme map (see Table 4–14). We must now decide which rule sets to prioritize in intervention and select appropriate target sounds to represent the chosen phoneme collapses. The most obvious candidate to begin therapy will, or course, be Jasmine’s pervasive substitution of /h/ for many other sounds. In addition, we will also want to address her collapse to /t/ or /d/ in either the initial or final position. As these collapses represent similar phonemic feature differences — the /t/ rule set addressing the voiceless sounds and the /d/ rule set addressing their voiced cognates — we will begin with one of these rule sets and anticipate generalization of learned contrasts to the other. Let’s begin with the /t/ collapse in the final position. We have now identified two rule sets to begin treatment — the initial /h/ collapse and the final /t/ collapse. Our next step will be to select two to four target phonemes to represent each of these rule sets. Table 4–15 provides possible targets to begin treatment. The phonemes listed in Table 4–15 were selected according to the distance metric. That is, there are maximal classifications between the phonemes chosen and, in addition, the targets are each maximally distinct from the error sound. Our first rule set includes /h/, a voiceless glottal fricative and it will be contrasted with /z/, a voiced alveolar fricative, /tS/, a voiceless post-alveolar affricate, /ɹ/, a voiced palatal liquid, and /fl/, a cluster containing a voiceless labiodental fricative and a voiced alveolar liquid. Our second error sound, final /t/, a voiceless alveolar stop, will be contrasted with /k/, a velar stop, /θ/, an interdental fricative, and /tS/, a post-alveolar affricate (no voiced sounds were available in this phonemic collapse). These two groups of phonemes, taken together, include many feature differences that will provide Jasmine with new phonological knowledge needed to achieve appropriate contrasts and improve her speech intelligibility. We will now generate five sets of contrastive multiple opposition words to represent each rule set. Remember, these words would be produced by Jasmine as homonyms without adult assistance. Table 4–16 provides a list of possible treatment words. The final step of preparation prior to beginning intervention is to administer a baseline probe using data collection forms created for each treatment set. Table 4–17 provides an example probe developed for the target phonemes selected for Jasmine’s first rule set (/h/ vs. /z, tS, ɹ, fl/). Once baseline data has been gathered, therapy may begin! Table 4–15.  Selected Targets for Jasmine’s Phoneme Collapses Error Sound

Target #1

Target #2

Target #3

Target #4 /fl/

Rule Set #1

/h/ initial

/z/

/tS/

/ɹ/

Rule Set #2

/t/ final

/k/

/θ/

/tS/

Table 4–16.  Contrastive Multiple Opposition Sets for Jasmine Contrastive Multiple Opposition Sets Treatment Set #1 for Rule Set #1:  /h/ (initial) versus /z, tS, ɹ, fl/ Error

Target #1

Target #2

Target #3

Target #4

/h/ (initial)

/z/

/tS/

/ɹ/

/fl/

Word Set 1

hip

zip

chip

rip

flip

Word Set 2

hi

zye

chai

rye

fly

Word Set 3

who

zoo

chew

Roo

flu

Word Set 4

head

Zed

Ched

red

fled

Word Set 5

hoe

Zoe

Cho

row

flow

Treatment Set #2 for Rule Set #2:  /t/ (final) versus /k, θ, tS/ Error

Target #1

Target #2

Target #3

/t/ final

/k/

/θ/

/tS/

Word Set 1

wit

wick

with

witch

Word Set 2

bat

back

bath

batch

Word Set 3

pit

pick

pith

pitch

Word Set 4

mat

Mac

math

match

Word Set 5

pat

pack

path

patch

120

Table 4–17.  Baseline and Generalization Probe for Jasmine Multiple Oppositions: Baseline & Generalization Probe Treatment Set #1 for Rule Set #1:  /h/ (initial) versus /z, tS, ɹ, fl/ Baseline words (also treatment words) Error

Target #1

Target #2

Target #3

Target #4

/h/ (initial)

/z/

/tS/

/ɹ/

/fl/

Word Set 1

hip

zip −

chip −

rip −

flip −

Word Set 2

hi

zye −

chai −

rye −

fly −

Word Set 3

who

zoo −

chew +

Roo −

flu −

Word Set 4

head

Zed −

Ched −

red −

fled −

Word Set 5

hoe

Zoe −

Cho −

row −

flow −

Baseline

Date: 10/8

0/5 = 0%

1/5 = 20%

0/5 = 0%

0/5 = 0%

Treatment

Date:

Treatment

Date:

Treatment

Date:

Target #1

Target #2

Target #3

Target #4

/z/ +/−

/tS/ +/−

/ɹ/ +/−

/fl/ +/−

“Z” −

chow +

rat −

flee −

zoom −

cheap +

ray −

flop −

zig −

chin −

reap −

flame −

zag −

chap −

rot −

fleet −

zap −

cheat −

room −

flap −

zipper −

chain +

right −

flat −

zest −

chill −

rude −

flight −

zinc −

cheek −

run −

flake −

zing −

cheese −

rope −

flag −

zero −

cheer −

root −

flare −

0/15 = 0%

3/15 =20%

0/15 = 0%

0/15 = 0%

Baseline Words (also generalization words)

Baseline

Date: 10/8

Generalization

Date:

Generalization

Date:

Generalization

Date:

Generalization

Date:

I = Imitation; S = Spontaneous

121

122

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

References Allen, M. M. (2013). Intervention efficacy and intensity for children with speech sound disorders. Journal of Speech, Language, and Hearing Research, 56, 865–877. Baker, E. (2010). Minimal pair intervention. In A. L. Williams, S. McLeod, & R. J. McCauley (Eds.), Interventions for speech sound disorders in children (pp. 41–72). Baltimore, MD: Brookes. Baker, E. (2016). Contrastive approaches to phonological intervention: minimal pairs and multiple oppositions. ASHA Online Conference: Improving Intelligibility in Children with Speech Sound Disorders. Camarata, S. M. (1993). The application of naturalistic conversation training to speech production in children with speech disabilities. Journal of Applied Behavior Analysis, 26, 173–182. Camarata, S. M. (1995). A rationale for naturalistic speech intelligibility training. In M. E. Fey, J. Windsor, & S. F. Warren (Eds.), Language intervention: Preschool through the elementary years. (pp. 63–84). Baltimore, MD: Paul H. Brookes. Cathell, V., & Ruscello, D. (2004, November). Sound system disorders: Teaching broad versus deep. Paper presented at the American Speech-Language Hearing Association, Philadelphia, PA. Ceron, M. I., & Keske-Soares, M. (2012). Análise do progresso terapêutico de crianças com desvio fonológico após aplicação do Modelo de Oposições Múltiplas [Analysis of the therapeutic progress of children with phonological disorders after the application of the multiple oppositions approach]. Jornal da Sociedade Brasileira de Fonoaudiologia, 24(1), 91–95. Gierut, J. A. (1990). Differential learning of phonological oppositions. Journal of Speech and Hearing Research, 33, 540–549. Gierut, J. A. (1991). Homonoymy in phonological change. Clinical Linguistics and Phonetics, 5, 119–137. Hammer, D. W. (2014, October). Using a multi-sensory therapy approach for speech sound disorders. Workshop presented at the Region 10 Coordinating Council, Fontana, CA. Pagliarin, K. C., Mota, H. B., & Keske-Soares, M. (2009). Análise da eficácia terapêutica em três modelos fonológicos de abordagem contrastive [Therapeutic efficacy analysis of three contrastive approach phonological models]. Pro-Fóno Revista de Atualização Científica, 21(4), 297–302. Richard, J. T. (2017, March). Home practice principles for children with speech sound disorders: Facilitating learning across contexts. Workshop presented at the California Speech-Language-Hearing Association Annual Convention, Pasadena, CA. Shriberg, L. D., Austin, D., Lewis, B. A., McSweeney, J. L., & Wilson, D. L. (1997). The percentage of consonants correct (PCC) metric: Extensions and reliability data. Journal of Speech, Language, and Hearing Research, 40, 708–722. Williams, A. L. (2000a). Multiple oppositions: Case studies of variables in phonological intervention. American Journal of Speech-Language Pathology, 9, 289–299. Williams, A. L. (2000b). Multiple oppositions: Theoretical foundations for an alternative contrastive intervention approach. American Journal of Speech-Language Pathology, 9, 282–288. Williams, A. L. (2003). Speech disorders resource guide for preschool children. Clifton Park, NY: Thomson Delmar Learning. Williams, A. L. (2005a). Assessment, target selection, and intervention: Dynamic interactions within a systemic perspective. Topics in Language Disorders, 25(3), 231–242. Williams, A. L. (2005b). From developmental norms to distance metrics: Past, present, and future directions for target selection practices. In A. G. Kamhi & K. E. Pollock (Eds.), Phonological disorders in children: Clinical decision making in assessment and intervention (pp. 101–108). Baltimore, MD: Brookes.

4 Multiple Oppositions Intervention

123

Williams, A. L. (2005c). A model and structure for phonological intervention. In A. G. Kamhi & K. E. Pollock (Eds.), Phonological disorders in children: Clinical decision making in assessment and intervention. Baltimore, MD: Brookes. Williams, A. L. (2006a). A systematic perspective for assessment and intervention: A case study. Advances in Speech-Language Pathology, 8(3), 245–256. Williams, A. L. (2006b). SCIP – Sound contrast in phonology: Evidence-based treatment program. Greenville, SC: Super Duper Publications. Williams, A. L. (2010). Multiple oppositions intervention. In A. L. Williams, S. McLeod, & R. J. McCauley (Eds.), Interventions for speech sound disorders in children (pp. 73–94). Baltimore, MD: Brookes. Williams, A. L., & Kalbfleisch, J. (2001, August). Phonological intervention using multiple oppositions. Poster session presented at the 25th World Congress of the International Association of Logopedics and Phoniatrics, Montreal, Canada.

5 Maximal Opposition Therapy Treatment

Key Features

Maximal Oppositions

n Utilizes minimal pairs contrasting two phonemes representing maximal distinctions n Variation called the Treatment of the Empty Set targets two unknown sounds in minimal pairs (rather than one known and one unknown sound).

Age Range 3½ to 6+ years

Severity of Phonological Disorder Demonstrate exclusion of at least 6 sounds from phonemic inventory Multiple sound errors across a variety of sound classes

Diagnostic Information Needed Initial consonant phoneme inventory Identification of known and unknown sounds

Target Selection Optimal targets incorporate contrasts across many distinctive or descriptive features (i.e., place, manner, voicing) and at least one

major class feature (i.e., sonorant, consonantal, vocalic)

Basic Structure of Therapy n Create five

minimal pairs containing target phonemes, using real words or nonsense words n  Phases of Treatment: imitation, spontaneous production

125

126

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Many practicing clinicians incorporate traditional minimal pair therapy into their current intervention programs. However, the maximal opposition approach, a variation of minimal pair therapy, offers the potential for more efficient intervention for children with moderate to severe speech sound disorders. Research (e.g., Gierut, 1992) has demonstrated that slight adjustments to target selection may produce greater system-wide change in children’s phonological systems. Possible modifications include the selection of minimal pairs which incorporate the following: 1. Two phonemes that represent maximal distinctions, rather than minimal distinctions, 2. Two phonemes that differ by a major class feature, rather than by only a nonmajor class feature, 3. Two phonemes that are unknown to the child, rather than one known and one unknown sound. These new options for creating minimal pairs are explored in this chapter. Selection of intervention targets utilizing these three principles of maximal oppositions may yield faster progress within a therapeutic framework that is already familiar to many SLPs. With high caseload numbers and heavy workloads, who wouldn’t want therapy to generate quicker results? Maximal oppositions treatment gives busy clinicians an evidence-based method for achieving this goal.

Maximal Distinctions:  What does it mean to select two phonemes that represent maximal distinctions, rather than minimal distinctions? Traditional minimal pair therapy involves contrasting two phonemes that are quite similar to each other, such as /t/ and /k/. Maximal opposition therapy, on the other hand, utilizes pairs of sounds that are very distinct. Both approaches begin by choosing a target phoneme that is not currently in a child’s inventory. In the traditional version of minimal pairs, the target phoneme is typically contrasted with the error — that is, the sound that the child uses as a substitute for the target sound. For example, if the child is unable to use /k/ in words and instead substitutes /t/ for /k/, then /k/ would be contrasted with /t/ in word pairs such as “key” and “tea.” In maximal opposition therapy, the selection of the second sound is based not upon an error pattern, but by finding a phoneme that is very different from the target sound. If the target is /k/, then the contrasting sound might be /m/ or /w/, and the resulting word pairs might include “me” and “key” or “we” and “key.” Thus, a central component of maximal opposition therapy is the selection of phonemes that contrast with each other across multiple features. Original descriptions and subsequent investigations have shown that descriptive features are a comprehensive way of determining maximal oppositions between targets (e.g., Barlow & Gierut, 2002; Elbert & Gierut, 1986; Gierut, 1989). Descriptive features constitute a well-known system for describing the differences between sounds. They include the three parameters of place, manner, and voicing to describe the physical production of speech sounds. Slight

5 Maximal Opposition Therapy

127

variations may be found among place-manner-voicing charts, but for our purposes we will use the one represented in Table 5–1. Each sound may be characterized according to the three parameters of place, manner, and voicing. For example, /t/ is an alveolar (place), stop (manner) that is voiceless (voicing). Contrasts between sounds may be described according to these variables as well. For example, /t/ and /k/ are minimally contrastive because they only vary by one descriptive feature — place. They share the other two descriptive features of voicing and manner, as both are categorized as voiceless stops. On the other hand, /m/ and /k/ are maximally contrastive because they vary across all three variables. The target /m/ is a voiced sound, whereas /k/ is voiceless; /m/ is a bilabial, whereas /k/ is a velar; /m/ is a nasal, whereas /k/ is a stop.

Major Class Feature Difference:  What does it mean to select two phonemes that differ by major class features, rather than only by nonmajor class features? Maximal opposition sound pairs may be selected using the place-manner-voicing paradigm as described above. However, some research on maximal oppositions utilized a more complex distinctive features system developed by Chomsky and Halle (1968). Distinctive features are the smallest components of sounds that serve to differentiate one phoneme from another and create two distinct words. For example, the distinctive feature of voicing differentiates /p/ from /b/, and thus the word “pay” from the word “bay.” Chomsky and Halle’s matrix of distinctive features captures how phonemes differ along various parameters related to the production characteristics of sounds (Table 5–2 provides samples of distinctive features). It is a binary system that categorizes each sound as either possessing a particular feature (indicated by a + sign) or not possessing a particular feature (indicated by a − sign). For example, the phoneme /m/ has the features of “voiced” and “nasal” and these are indicated in Table 5–2 by a + sign, but /p/ does have not have the features of “voiced” or “nasal,” indicated in Table 5–2 by a − sign. An important component of Chomsky and Halle’s system is the inclusion of features considered to constitute a major class distinction. The matrix developed by these authors in The Sound Pattern of English (1968) included five categories of features: major class, cavity, manner of articulation, source, and prosodic. The first category — major class features — carries special significance for our discussion of maximal opposition therapy. Research has demonstrated that phoneme pairs that include a difference across a major class feature will yield better therapeutic results than a phoneme pair that does not include such a distinction (e.g., Gierut, 1990; 1992). Major class features establish three basic sound classes: sonorant, consonantal, and vocalic. Sonorant sounds are relatively louder than other sounds and are produced with a more open expiratory passage. Vowels are sonorants, as well as specific consonants. Consonantal sounds are produced with a high degree of oral obstruction. Vocalic sounds are characterized by a low degree of oral obstruction. These features are represented in the first three rows of the Chomsky and Halle (1968) matrix in Table 5–2. In addition, Table 5–3 provides a list of sound classes included in each major class feature category.

w

m

f

v

Labiodental

θ ð

Interdental

s

t

l

n

z

d

Alveolar

Z dZ

ʃ tS

Postalveolar

Coronal

j

ɹ

Palatal k

Velar

+

+

+

−

−

−

+

Nasal

Sonorant

Consonantal

Continuant

Strident

Coronal

Anterior

+

+

−

−

+

+

+

+

n

−

−

−

−

+

+

+

+

ŋ

+

−

−

−

+

−

−

−

p

+

−

−

−

+

−

−

+

b

+

+

−

−

+

−

−

−

t

Source:  Adapted from Bauman-Waengler, in press.

+

Voiced

m

+

+

−

−

+

−

−

+

d

−

−

−

−

+

−

−

−

k

−

−

−

−

+

−

−

+

g

−

−

−

+

−

+

−

−

h

+

−

+

+

+

−

−

−

f

+

−

+

+

+

−

−

+

v

+

+

+

+

+

−

−

−

s

+

+

+

+

+

−

−

+

z

−

+

+

+

+

−

−

−

ʃ

−

+

+

+

+

−

−

+

Z

+

+

−

+

+

−

−

−

θ

+

+

−

+

+

−

−

+

ð

−

+

+

−

+

−

−

−

tS

−

+

+

−

+

−

−

+

dZ

Table 5–2. Samples of Distinctive Features of English Consonant Sounds According to Chomsky and Halle (1968)

*Sonorant

Glide*

Liquid*

Nasal*

Affricate

Fricative

p

Stop

b

Bilabial

MANNER

Labial

PLACE

Table 5–1.  Place-Manner-Voicing Descriptive Features: American English Consonants

Sonorant

128 −

−

−

+

−

+

−

+

w

ŋ

g

−

−

−

+

−

+

−

+

j

−

+

−

+

+

+

−

+

ɹ

Glottal

h*

Dorsal

+

+

−

+

+

+

−

+

l

5 Maximal Opposition Therapy

129

Table 5–3.  Three Major Class Features +Sonorant

−Sonorant

+Consonantal

−Consonantal

+Vocalic

−Vocalic

Vowels

Stops

Stops

Vowels

Vowels

Stops

Glides

Fricatives

Fricatives

Glides

Liquids

Fricatives

Nasals

Affricates

Affricates

/h/

Affricates

Liquids

Nasals

Nasals

/h/

Liquids

/h/

This table shows us, for example, that fricatives are all classified as (+) consonantal, but as (−) sonorant and (−) vocalic. Liquids, on the other hand, are categorized as (+) sonorant, (+) vocalic, and (+) consonantal. Fricatives and liquids, therefore, contrast with each other across two major class features. Fricatives are not sonorant, whereas liquids are sonorant; fricatives are not vocalic, whereas liquids are vocalic. Phonemes selected from these two sound classes would thus constitute favorable target pairs within a maximal opposition approach. For example, the sound pairs of /f/ (a fricative) versus /l/ (a liquid), or /s/ (a fricative) versus /ɹ/ (a liquid), would vary by two major class features each. To summarize, the maximal opposition approach guides us to select target phoneme pairs that are maximally distinct from one another. For optimal therapeutic results, we should select sounds with the most feature differences, including at least one major class feature distinction. Although some studies utilized a distinctive feature matrix developed by Chomsky and Halle (1968), we encourage clinicians to use the less complex place-manner-voicing descriptive feature system (see Table 5–1). Selection of two sounds that vary across all three parameters of place, manner, and voicing will yield maximally contrastive pairs. And for even better results, attention should be given to selecting phonemes that represent a major class feature distinction (see Table 5–3). Later in this chapter, a step-by-step process will be presented for choosing optimal maximal opposition targets (see Initial Data Collection and Selecting Targets, p. 135)

Known Versus Unknown Sounds:  What does it mean to select two phonemes that are unknown to the child, rather than one known and one unknown sound? In all forms of minimal pair therapy, the first phoneme chosen as a target is not in the child’s current inventory. It is an unknown sound. In traditional minimal pair therapy, the opposing phoneme is typically the sound used by the child as a substitute for the target phoneme (e.g., /t/ for /k/). Thus, one of the sounds of the minimal pair is already in the child’s phonemic inventory. It is a known sound. Therefore, the resulting word pair contains one unknown sound (e.g., /k/) and one known sound (e.g., /t/). Initial research on maximal opposition pairs also included one new target sound

130

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

contrasted with a sound already produced by the child. However, the known sound was not the substitute phoneme, but rather a maximally distinct sound identified from the child’s pretreatment phonemic inventory (e.g., /m/). Later studies by Gierut (1991, 1992) utilized two sounds that the child was unable to produce. This was termed Treatment of the Empty Set (i.e., treatment of two new unknown phonemes) and was shown to be more effective than the previous approaches which had included only one new target sound. Maximal Opposition therapy and Treatment of the Empty Set may be combined to create a new version of minimal pair therapy. Clinicians may choose sound pair targets that are maximally distinct, incorporate a major class distinction, and represent two unknown sounds in order to achieve optimal therapeutic results.

Nonhomonymous Word Pairs An important difference between maximal opposition therapy and traditional minimal pair intervention is the use of nonhomonymous word pairs. In a traditional approach, word pairs are typically selected based upon a child’s error pattern. For example, if a child demonstrates a /t/ for /s/ substitution, word pairs are selected representing the contrast between initial /s/ and initial /t/. The words “sea” and “tea” might be chosen. The child would predictably produce these two words the same, as homonyms, saying “tea” for both “sea” and “tea.” The inclusion of homonyms as targets is a central component of traditional minimal pair therapy. The child’s realization of the semantic consequences produced by his misproductions is hypothesized to be a motivating factor in changing speech behavior. Maximal oppositions therapy calls into question this presupposition. Because sound pairs are chosen on the basis of feature differences, rather than a child’s error patterns, the resulting target words are often not those produced by the child as homonyms. For a child demonstrating stopping of /s/, the clinician would still choose /s/ as a target. However, now the opposing sound would be chosen based upon the three variables described above. The sound /l/ might be chosen as it meets all three criteria: /l/ differs from /s/ across many features, it includes a major class difference, and /l/ is not in the child’s pretreatment inventory (i.e., it is an unknown sound). The resulting word pair might be “sea” and “Lee.” Given our knowledge of the child’s system, we could predict that these words would be produced as “tea” and “we.” Clearly, these are not homonyms for the child. The opportunity to confront the child with the communication breakdowns created by homonymy is not as readily present as in traditional minimal pair therapy. Yet reduction of homonymy is the goal of all minimal pair therapy. Gierut (1990) explains that reducing homonymous word forms is both the means and the end of traditional minimal pair therapy. It is still also the goal of maximal opposition therapy — even if homonyms are not directly used as the method to achieve this objective. The presentation of phoneme pairs that are maximally opposed allows children to gain knowledge regarding linguistic contrasts across multiple features simultaneously. The result will be an increase in the use of contrastive features and a decrease in the occurrence of homonymy, as children produce words to communicate in their daily lives.

5 Maximal Opposition Therapy

131

Research Supporting Maximal Opposition Therapy A series of studies was conducted to systematically manipulate variables relative to the selection of minimal pair targets. The first study (Gierut, 1989) found that sound pairs chosen that included maximal distinctions rather than minimal distinctions were more effective in promoting system-wide change within a child’s disordered phonological system. The second study (Gierut, 1990), indicated that the use of sound pairs which included a major class distinction were more effective than those that did not. The focus of experimentation in the third and fourth studies (Gierut, 1991; Gierut & Neumann, 1992) shifted away from the number and nature of feature differences between target phonemes. Instead, the researchers focused on using nonhomonymous word pairs containing two sounds not previously known to a child. This new approach was termed Treatment of the Empty Set and was determined to be more effective than the more traditional method of targeting only one new phoneme at a time. These studies have been summarized in Table 5–4. Gierut’s fourth study (1992) systematically manipulated the three variables previously considered: (1) number of distinctive feature differences; (2) inclusion of major versus nonmajor distinctions; and (3) use of known versus unknown sounds. The combined results of these studies led to the development of a hierarchy of effectiveness for selection of minimal pair therapy targets (Box 5–1). This hierarchy allows clinicians to predict the amount of change likely to occur within a child’s phonological system.

Box 5–1.  Hierarchy of Minimal Pairs (from Most to Least Effective) 1. Empty Set (with Maximal Oppositions): Two New Phonemes; Major Class Distinction; Maximal Opposition 2. Empty Set (with Maximal Oppositions): Two New Phonemes; Nonmajor Class Distinction; Maximal Opposition 2. Maximal Oppositions: One New Phoneme; Major Class Distinction; Maximal Opposition 3. Maximal Oppositions: One New Phoneme; Nonmajor Class Distinction; Maximal Opposition 4. Traditional Minimal Pair: One New Phoneme; Nonmajor Class Distinction; Minimal Opposition Source:  Adapted from Barlow & Gierut, 2002 and Gierut, 1992.

Table 5–4.  Examples of Research Supporting Maximal Opposition Approach and Treatment of the Empty Set

Authors, Date

Number of Subjects

Age Range

Gierut, 1989

1 child

4;7

Severity Omission of initial consonants

Results Approach: Maximal oppositions in which pairs of phonemes were selected based on number of feature differences (maximal rather than minimal) Results:  Use of word pairs incorporating maximal oppositions proved effective in teaching 16 initial consonants and promoting system-wide change.

Gierut, 1990

3 children

Age 4

Exclusion of at least 6 sounds from child’s phonetic and phonemic inventories

Approach: Maximal oppositions treatment comparing minimal opposition word pairs to maximal oppositions Results:  Maximal opposition targets found to be more effective in improving target phonemes and generalizing to untreated phonemes. Pairs that differed across major class distinctions proved more effective than nonmajor class distinctions.

Gierut, 1991

3 children

2 Boys, Age 5 1 Girl, Age 4

Exclusion of at least 6 sounds from child’s phonetic and phonemic inventories

Approach:  Treatment of the empty set that compared traditional minimal pair therapy (error & substituted phoneme) with treatment of the empty set (two phonemes in error not found in child’s inventory) Former incorporated homonymy; latter did not. Minimal feature differences, including nonmajor class distinctions, were used for both approaches. Nonsense words were utilized for all pairs. Results:  The non-homonymous approach (empty set) resulted in more accuracy of target phonemes and more generalization to untreated phonemes.

132

Table 5–4.  continued

Authors, Date Gierut & Neumann, 1992

Number of Subjects

Age Range

1 child

4;8

Severity Exclusion of at least 6 sounds from child’s phonetic and phonemic inventories

Results Approach:  Treatment of the empty set This was a replication of the Gierut (1991) study, utilizing the same sounds for the empty set treatment approach as were used for the minimal pair therapy approach in the 1991 study (voiced and voiceless “th”). Results:  Confirmed findings of previous study, treatment of the non-homonymous empty set proved more effective in facilitating phonological change.

Gierut, 1992

4 children

Mean Age: 3;10

Exclusion of at least 6 sounds from child’s phonetic and phonemic inventories

Approach: Maximal oppositions & treatment of the empty set in which experimental manipulation of three variables relative to selection of minimal pairs— (1) number of distinctive feature differences (minimal vs. maximal); (2) type of feature differences (nonmajor vs. major class distinctions); (3) exclusion or inclusion in child’s pretreatment inventory (unknown vs. known sounds) Results:  Greatest amount of phonological change occurred when two unknown targets were selected that represented maximal number of feature differences and major class distinctions.

133

134

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

According to this hierarchy, the most effective option for promoting phonological change within a child’s disordered system is to select pairs that include two sounds that are unknown to the child, and which contain maximal feature differences, and at least one major class distinction. The next two options were deemed equally effective: two unknown sounds with maximal oppositions, but no major class distinctions; or one unknown sound and one known sound with maximal oppositions, including at least one major class distinction. Phoneme pairs that included maximal oppositions, but only one unknown sound and no major class distinctions, were found to be less effective than the first three options described. The least effective method for creating positive change was to choose minimal pairs according to the traditional protocol — that is, to select only one new phoneme and to have only minimal distinctions between the new and the old sound. Thus, traditional minimal pair therapy was determined to be the least effective option for promoting phonological change within a disordered system. Research on maximal oppositions therapy has primarily focused on intervention for singleton consonants in the initial position of words (Gierut, 1989, 1990, 1991, 1992; Gierut & Neumann, 1992). Intervention targets represented inventory constraints — that is, sounds that were completely excluded from children’s pretreatment phonemic repertoires. Thus, the treatment methods of maximal oppositions and treatment of the empty set have not been experimentally applied to phonotactic constraints, such as omitting final consonants or reducing consonant clusters.

What Is the Age and Severity of Children Who Could Most Benefit from Maximal Opposition Therapy? Research studies that support maximal oppositions and treatment of the empty set included children ranging in age from 3 years, 6 months to 5 years, 4 months, who came from monolingual English speaking families. The subjects exhibited normal hearing, as well as typical oral and speech motor abilities. Most of the studies required that at least six sounds be excluded from the subjects’ phonetic and phonemic inventories. Gierut (1989) suggests that children who display “extensive gaps” in their phonological systems and/or inconsistent error patterns may benefit from maximal oppositions or treatment of the empty set, whereas traditional minimal pair therapy may be more appropriate for children with only one or two error patterns. For children with multiple speech sound errors across a variety of sound classes, these newer variations of minimal pair therapy will likely lead to greater generalization and faster progress toward normalized speech. Because maximal oppositions therapy specifically addresses the loss of contrast between singleton consonants in the initial position of words, it may be ideally suited for children who demonstrate these particular deficits. In fact, the first study conducted by Gierut (1989) involved a child with extensive initial consonant deletion. The subject was lacking 16 singleton consonants in the prevocalic word position. Maximal opposition intervention, with its focus on teaching initial consonant contrasts, was ideally suited to meet this child’s needs and did so successfully.

5 Maximal Opposition Therapy

135

Initial Data Collection and Selecting Targets Identification of Unknown Sounds Target selection for the maximal oppositions approach begins with identification of the child’s unknown sounds — that is, sounds that the child does not use as phonemes when producing words. This information may be obtained from the results of a standardized test, spontaneous speech sample analysis, or use of probe lists (see Chapter 2). Table 5–5 presents a sound production summary for Ethan, age 5 years, 2 months, whose data we will use as an illustration. Clinicians will need a list of consonant phonemes produced by the child (i.e., known sounds), as well as a list of consonant phonemes not used by the child (i.e., unknown sounds). If a phoneme is used at least one time appropriately in any position of a word, include that sound in the list of known sounds. We may use the information presented

Table 5–5.  Sound Productions for Ethan, Age 5;2 Early Sounds

I

M

F

Stops

Later Sounds

I

M

F

Fricatives

p

3

3

3

f

3

3

3

b

3

3

3

v

b

b

b

t

3

3

3

s

t

t

t

d

3

3

3

z

d

d

d

k

t

t

t

θ

t

t

t

g

d

d

d

ð

d

d

d

ʃ

s

t

t

Nasals m

3

3

3

Liquids

n

3

3

3

ɹ

w

w

ə

ŋ

----

3

3

l

w

w

ə

Glides

Affricates

w

3

3

----

tS

t

t

ts

j

3

3

----

dZ

d

d

d

3

3

----

Fricative h

136

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

in Table 5–5 to compile the following lists of Ethan’s known and unknown consonant phonemes: Ethan’s known consonant phonemes:  /p, b, t, d, m, n, ŋ, w, j, h, f/ Ethan’s unknown consonant phonemes:  /k, g, v, s, z, ʃ, θ, ð, ɹ, l, tS, dZ/ A sound probe should be administered prior to the initiation of maximal opposition intervention to secure baseline data. A list that assesses each sound in five different words should yield sufficient information for intervention planning. These words that comprise baseline and monitoring data will not receive direct instruction during therapy, and so should not be included as intervention stimuli. The probe words will also be periodically re-administered to monitor progress and assess generalization. Table 5–6 provides an example of a sound probe completed for Ethan. In this sound probe, all productions are recorded as correct (+) or incorrect (−). Clinicians may choose to include additional information regarding the nature of errored productions, for example, by notating sound substitutions. Table 5–6 allows clinicians to categorize each sound as “known” or “unknown” (remember, a sound will be considered “known” if it is produced correctly even one time). Thus, this baseline sound probe may also be used to confirm our list of known and unknown sounds that was generated for Ethan, based upon more formal assessment measures.

Sound Pair Selection Options All minimal pair variations start with identification of an unknown sound to begin intervention planning. Let’s use /k/ as an example target for Ethan. We will look first at target selection within the context of traditional minimal pair therapy, in order to highlight the differences between a traditional approach and the new options for identifying potential targets offered by the maximal opposition approach and treatment of the empty set. Research on maximal oppositions therapy has been restricted to intervention for singleton consonant sounds in the initial position of words, so we will use initial consonants as targets in our examples. Historically, production targets were identified based upon developmental norms and stimulability. For Ethan, /k/ might have been chosen as an appropriate initial target for intervention. Within a conventional minimal pair therapy framework, a second phoneme would be chosen to contrast with /k/ based upon Ethan’s error patterns. Because Ethan substitutes /t/ for /k/, these two phonemes would be chosen to create minimal pair word sets, such as “key” and “tea.” If we look more closely at our chosen sound pair (/k/ vs. /t/), we will notice that the two sounds are not very different from one another. The differences, or contrasts, between the sound pairs may be analyzed using place-manner-voicing descriptive features (see Table 5–1). When analyzed utilizing a descriptive features framework, we will see that /k/ and /t/ differ only according to the feature of place. Both sounds are stops (manner) and voiceless. In addition, this sound pair does not represent a major class feature difference. The targets /k/ and /t/ share in common all three major class features

Table 5–6.  Sound Probe for Ethan’s Baseline Data

+/−

+/−

+/−

+/−

+/−

Total

Percentage

Known/ Unknown

p

paw +

pig +

pea +

pen +

poke +

5/5

100%

Known

b

big +

bite +

bean +

ball +

boat +

5/5

100%

Known

t

toe +

two +

tan +

tell +

ten +

5/5

100%

Known

d

do +

day −

deep +

dime −

down +

3/5

60%

Known

k

key −

comb −

cow −

cup −

can −

0/5

0%

Unknown

g

go −

game −

get −

good −

gate −

0/5

0%

Unknown

m

mad +

man +

map +

me +

mean +

5/5

100%

Known

n

no +

now +

night +

knot +

need +

5/5

100%

Known

w

we +

wet +

week +

wall +

wide +

5/5

100%

Known

j

you +

yell +

yolk +

young −

your +

4/5

80%

Known

h

high +

have +

hill +

him +

home +

5/5

100%

Known

f

fall +

fan +

fed +

fig −

fight +

4/5

80%

Known

v

van −

vote −

volt −

vault −

veil −

0/5

0%

Unknown

s

sew −

sing −

sock −

soup −

some −

0/5

0%

Unknown

z

zoom −

zoo −

zip −

zap −

zing −

0/5

0%

Unknown

θ

thank −

think −

thick −

thin −

thing −

0/5

0%

Unknown

ð

that −

the −

they −

then −

them −

0/5

0%

Unknown

ʃ

shake −

shoe −

show −

shut −

ship −

0/5

0%

Unknown

Sounds Stops

Nasals

Glides

Fricatives

continues

137

138

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 5–6.  continued

+/−

+/−

+/−

+/−

+/−

Total

Percentage

Known/ Unknown

ɹ

row −

rain −

road −

ring −

rock −

0/5

0%

Unknown

l

low −

lamb −

like −

late −

lap −

0/5

0%

Unknown

tS

chain −

cheap −

chop −

chin −

chick −

0/5

0%

Unknown

dZ

gym −

jack −

jam −

jeep −

jet −

0/5

0%

Unknown

Sounds Liquids

Affricates

(see Table 5–3). Both are categorized as consonantal, and both are not categorized as sonorant or vocalic. This conventional minimal pair target choice — /k/ versus /t/ — is a good example of a minimally contrastive sound pair. Table 5–7 presents a summary of this conventional minimal pair target selection example. Now let’s try a new method for our minimal pair target selection. Rather than using the error sound (/t/) to contrast with our target sound (/k/), we will choose a sound from Ethan’s inventory that is maximally different from /k/ to create our word pairs. Because we are no longer restricted to using the one phoneme that is the child’s replacement sound, we have many options to pick from. If we choose /m/ to contrast with /k/, we achieve contrast across all three descriptive features of place (velar vs. bilabial), manner (stop vs. nasal), and voice (voiceless vs. voiced). Even better, the two sounds contrast by a major class distinction: /m/ is a sonorant and /k/ is not. For Ethan, the sound pair /k/ versus /m/ is a good choice to begin maximal opposition therapy. Table 5–8 presents a summary of this maximal opposition minimal pair target selection example. If we review our initial lists of Ethan’s known and unknown sounds, we will see that the choice of /k/ and /m/ includes only one unknown sound, as /m/ is included in Ethan’s current phonemic inventory. According to research on the treatment of the empty set (Gierut, 1991; Gierut & Neumann, 1992), we might achieve even better results by selecting a second target to contrast with /k/ that is not in the child’s current phonemic inventory. If both target sounds are unknown to the child, greater phonological change may be predicted. Consonantal /ɹ/ is a possible selection that would meet the criteria of both the empty set and maximal oppositions. The phonemes /k/ and /ɹ/ are unknown sounds for Ethan, and the two sounds are maximally distinct from one another. The two proposed targets differ in regards to place (velar vs. palatal), manner (stop vs. liquid), and voicing (voiceless vs. voiced), and they include two major class distinctions (sonorant and vocalic). Table 5–9 summarizes this target selection example that combines treatment of the empty set with maximal opposition guidelines.

5 Maximal Opposition Therapy

139

Table 5–7.  Ethan: Conventional Minimal Pair Target Selection Example

Unknown Sound

Error Sound

/k/

/t/

Descriptive Feature Difference(s)

Major Class Feature Difference(s)

1) PLACE velar vs. alveolar

NO

Sample Pairs: Homonymous key & tea cape & tape coat & tote

Table 5–8.  Ethan: Maximal Opposition Minimal Pair Target Selection Example

Unknown Sound

Known Sound

/k/

/m/

Descriptive Feature Difference(s)

Major Class Feature Difference(s)

Sample Pairs: Nonhomonymous

1) PLACE: velar versus bilabial 2) MANNER: stop versus nasal 3) VOICE: voiceless versus voiced

YES −sonorant vs. +sonorant

key & me can & man kite & might

Table 5–9.  Ethan: Empty Set with Maximal Opposition Minimal Pair Target Selection Example

Unknown Sound

Unknown Sound

/k/

/ɹ/

Descriptive Feature Difference(s)

Distinctive Feature Difference(s)

1) PLACE: velar versus palatal 2) MANNER: stop versus liquid 3) VOICE: voiceless versus voiced

YES −sonorant versus +sonorant −vocalic versus +vocalic

Sample Pairs: Nonhomonymous can & ran cat & rat kite & right

Within the hierarchy of effectiveness for minimal pair target selection developed by Gierut and illustrated in Box 5–1, our final selection (/k/ vs. /ɹ/) would be predicted to achieve the most phonological change within Ethan’s developing phonemic system. We have now identified one sound pair (/k/ vs. /ɹ/) that is a good potential target for Ethan’s maximal opposition therapy program. Are there other options? As described above, an unknown initial phoneme (e.g., /k/) is identified as a treatment target and then a second phoneme is selected which will maximally contrast with the first one

140

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

(e.g., /k/ vs. /ɹ/). In order to facilitate the selection process, Table 5–10 provides a list of potential sound pairs that include contrasts across place, manner, and voice, as well as at least one major class feature difference. For example, if /k/ is chosen as a target

Table 5–10.  Maximal Opposition Sound Pairs Phoneme

Maximal Opposition Sound Pairs*

p

/p/ versus /n/  /p/ versus /j/  /p/ versus /l/  /p/ versus /ɹ/

b

/b/ versus /h/

t

/t/ versus /m/  /t/ versus /w/  /t/ versus /j/  /t/ versus /ɹ/

d

/d/ versus /h/

k

/k/ versus /m/  /k/ versus /w/  /k/ versus /j/  /k/ versus /n/  /k/ versus /l/ /k/ versus /ɹ/

g

/g/ versus /h/

m

/m/ versus /t/  /m/ versus /k/  /m/ versus /h/  /m/ versus /s/  /m/ versus /ʃ/  /m/ versus /tS/

n

/n/ versus /p/  /n/ versus /k/  /n/ versus /h/ 

w

/w/ versus /t/  /w/ versus /k/  /w/ versus /h/  /w/ versus /s/  /w/ versus /ʃ/ /w/ versus /tS/

j

/j/ versus /p/  /j/ versus /t/  /j/ versus /k/  /j/ versus /h/  /j/ versus /f/   /j/ versus /s/  /j/ versus /θ/

h

/h/ versus /b/  /h/ versus /d/  /h/ versus /g/  /h/ versus /m/  /h/ versus /n/ /h/ versus /w/  /h/ versus /j/  /h/ versus /v/  /h/ versus /z/  /h/ versus /ð/ /h/ versus /ɹ/  /h/ versus /l/  /h/ versus /dZ/

f

/f/ versus /n/  /f/ versus /j/  /f/ versus /l/  /f/ versus /ɹ/

v

/v/ versus /h/

s

/s/ versus /m/  /s/ versus /w/  /s/ versus /j/  /s/ versus /ɹ/

z

/z/ versus /h/

θ

/θ/ versus /m/  /θ/ versus /j/ /θ/ versus /ɹ/

ð

/ð/ versus /h/

ʃ

/ʃ/ versus /m/  /ʃ/ versus /w/

ɹ

/ɹ/ versus /p/  /ɹ/ versus /k/  /ɹ/ versus /θ/

l

/l/ versus /p/  /l/ versus /k/  /l/ versus /h/  /l/ versus /f/

tS

/tS/ versus /m/  /tS/ versus /w/

dZ

/dZ/ versus /h/

/n/ versus /f/

/ɹ/ versus /h/  /ɹ/ versus /f/   /ɹ/ versus /s/ 

*Sound pairs contrast across all three dimensions of place, manner, and voice, and include at least one major class feature distinction.

5 Maximal Opposition Therapy

141

for intervention, the following pairs are listed in Table 5–10 as potential starting points for therapy because they each represent a contrast across all three dimensions of place, manner, and voicing, as well as a major class distinction: /k/ versus /m/; /k/ versus /w/; /k/ versus /n/; /k/ versus /l/; /k/ versus /ɹ/. All of these sound pairs would qualify as an appropriate target for maximal opposition therapy for Ethan. In addition, two of these sound pairs represent maximal oppositions that include phonemes that are unknown to Ethan: /k/ versus /l/ and /k/ versus /ɹ/. Either of these choices would be an excellent starting point for implementation of treatment of the empty set (combined with maximal oppositions). Keep in mind that although Table 5–10 includes the most optimal maximal contrast oppositions for each sound, clinicians may choose a sound pair that is less maximally opposed, with the understanding that it may yield less effective therapeutic results. Research indicates that we might expect greater system-wide change if we select targets that are maximally opposed, such as those pairs listed in Table 5–10, and perhaps even better results if we narrow our choices to include only sounds that are unknown to the child (i.e., not in the pretreatment inventory). However, it is a clinical decision that should be made on an individual basis. Factors such as a child’s tolerance for difficult tasks or the desire to include a variety of feature contrasts in a therapy program may justifiably influence target selection. Target selection for maximal oppositions and/or treatment of the empty set, as described in this chapter, often yields targets that may be considered more complex (Gierut, 2001). The choice of sounds that are unknown or represent maximal distinctions often results in the use of later developing or less stimulable sounds. In fact, the intentional selection of more complex structures when implementing these minimal pair variations may result in better clinical outcomes (Gierut, 2001). Chapter 6 addresses complexity approaches in detail.

Establishing Treatment Goals We might choose to begin Ethan’s maximal opposition program with the /k/ versus /ɹ/ sound pair. Table 5–11 gives examples of how these intervention targets could be presented as treatment goals. Each treatment goal must have corresponding baseline data, in order for progress to be accurately determined. As may be seen in the examples above, the nature of the baseline data will be guided by the content of the goal. The baseline probe administered to Ethan (see Table 5–6) will provide appropriate baseline data for the first option presented in his goals. For the second option, a different probe must be developed that includes lists of minimal pair words containing the target sound pair.

Beginning Therapy Maximal opposition therapy and treatment of the empty set describe target selection guidelines. The focus is on what to target, rather than how to target it (Gierut, 2001). That said, treatment protocols are described in the research literature which may serve as guidelines for our own implementation of this approach.

142

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 5–11.  Example Goals for Ethan; Maximal Opposition Therapy Goal Area #1:  /k/ versus /ɹ/ Option #1:  Write the intervention goal to target the specific sounds to be taught as the vehicle for increasing the use of contrastive features. Baseline #1:  Ethan produced /k/ and /ɹ/ in the initial position of single words with 0% accuracy, in the context of spontaneously naming pictures during administration of a baseline probe. Goal #1:  Ethan will produce /k/ and /ɹ/ in the initial position of single words with 70% accuracy, in the context of spontaneously naming pictures. Option #2:  Write the intervention goal to target the production of contrasting sounds in the context of minimal pair words. Baseline #1:  Ethan produced minimal pair words containing the contrast between initial /k/ and initial /ɹ/ with 0% accuracy, in the context of spontaneously naming pictures during administration of a baseline probe. Goal #1:  Ethan will produce minimal pair words containing the contrast between initial /k/ and initial /ɹ/ with 70% accuracy, in the context of spontaneously naming pictures.

Target Word Selection After a sound pair has been chosen to begin intervention (e.g., /k/ vs. /ɹ/), rhyming word pairs are generated with the target sounds in the initial position of the words (e.g., kite/ right). These rhyming dyads constitute the minimal pairs that form the focus of our therapy. Some descriptions of maximal oppositions include the use of real words. One early study (Gierut, 1989) incorporated five real word pairs for each sound target duo. Stimuli were selected that could be represented in picture form in order to create therapy materials. Appendix 5–1 includes sample lists of real words that incorporate maximally contrastive sound pairs. For Ethan, we might select the following five word pairs for the sound contrast /k/ versus /ɹ/ (obtained from Appendix 5–1): caw/raw, cat/rat, king/ring, cane/rain, code/ road. It is advisable to select targets that incorporate a variety of vowel sounds in relatively simple word shapes. In addition, attention should be paid to other consonant sounds in the words. Our choices for Ethan include five different vowel sounds in consonant-vowel (CV) or consonant-vowel-consonant (CVC) word shapes. The final sound of each of the four CVC words has been selected from his known inventory sound list. This will ensure that attention may be focused on production of the initial sounds and that difficulty producing the final sounds will not be a compounding factor in intervention. Such attention to the phonemic and phonotactic construction of target words is not a requirement of maximal opposition therapy, but is prudent when implementing any phonological intervention approach. Later experiments (Gierut, 1990; Gierut, 1991; Gierut, 1992; Gierut & Neumann, 1992) utilized nonsense words (NSWs) to create eight sets of stimuli for each sound pair. Lexical meaning was assigned to the NSWs in the context of stories that included

5 Maximal Opposition Therapy

143

imaginary creatures and actions (Gierut, 1990). The made-up words were developed in a manner that ensured consistency relative to factors such as syllable shape and phonemic composition. Stories containing the NSWs were presented to children before initiation of treatment, and individual cards were created to visually represent each nonsense word. These picture cards were utilized during therapy in sorting and matching games, and activities were sent home for additional practice (Gierut, 1990). Please see Chapter 6 for more ideas on the use of NSWs in therapy. The use of NSWs affords specific advantages in the context of research, since variability of stimuli may be controlled when comparing two treatment protocols (Gierut, 1991). However, practicing clinicians may find that the use of real words also carries advantages, including easier location of appropriate pictures and therapy materials. Caroline Bowen’s website (http://www.speech-language-pathology.com) includes excellent real word lists and worksheets for use with maximal opposition minimal pairs. Please also see Appendix 5–1 for sample word pairs representing maximally opposed sound contrasts.

Phases of Treatment The experimental studies included an imitation phase and a spontaneous production phase. During the imitation phase, the child is required to name each picture immediately following an adult model; practice occurs primarily within the context of drill work. Once the child demonstrates 75% accuracy during two consecutive sessions on all target words, or after seven consecutive sessions (whichever comes first), the second phase is initiated. The same set of word pairs are presented, but now the child is required to name the pictures spontaneously, with no adult model provided. Practice becomes more variable, including matching games and sorting words by initial sound. After the child demonstrates 90% accuracy during three consecutive sessions, or after 12 sessions have elapsed, intervention on the first target sound pair is completed (Gierut, 1990, 1991, 1992; Gierut & Neumann, 1992). Verbal praise for correct productions is provided continuously during the imitation phase, and then faded to intermittent reinforcement during the spontaneous phase. If a child produces a target word incorrectly, the clinician provides a verbal model but no subsequent feedback for additional attempts (Gierut, 1992). Upon completion of the two phases of intervention targeting the first sound pair, the probe is readministered. It is important to engage in an ongoing process of monitoring system-wide change, as the hope is that each sound pair target will result in some generalization. Sounds that were in error at the beginning of the intervention period may have been corrected, even if they were not direct targets of treatment. Analysis of the post-treatment sound probe will allow for selection of the next target sound (or sounds), and the process described above is then repeated for the new word set. There are not specific guidelines for the order of target selection, such as following a developmental progression. Rather, targets should be selected that will facilitate learning of a range of features and contrasts. If we select /k/ and /ɹ/ as our initial targets for Ethan, he will gain knowledge about the features represented by these two sounds, as well as the contrasts that exist between them (i.e., voiced versus unvoiced; velar versus

144

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

palatal; stop versus liquid). It would be preferable to select sounds for the second set of targets that will introduce new phonological knowledge. Let’s suppose that the posttreatment probe, administered after Ethan has mastered /k/ and /ɹ/ with 90% accuracy, indicates that Ethan continues to have difficulty with all untreated consonant sounds with the exception of /g/. Thus, Ethan’s unknown list now includes: /v, s, z, ʃ, θ, ð, ɹ, l, tS, dZ/. We might select /tS/ and /w/ as our next sound pair, in order to introduce the new contrasts of postalveolar versus bilabial and affricate versus glide. This particular pair (/tS/ vs. /w/) represents an unknown sound contrasted with a known sound for Ethan, and is therefore an example of maximal oppositions, but not treatment of the empty set. As the number of sounds included in a child’s unknown repertoire decreases, the options for treatment of the empty set will correspondingly diminish. Descriptions of maximal oppositions and treatment of the empty set do not include direct work on speech discrimination or perception, although these skills may be indirectly impacted. However, because maximal oppositions and treatment of the empty set describe minimal pair variations, it may be appropriate to consider implementation of these target selection approaches within the framework of conventional minimal pair therapy (Velleman, 2016). As described in Chapter 2, minimal pair intervention may include a familiarization phase, as well as work on perceptual discrimination of sound contrasts. Clinicians are encouraged to consider inclusion of discrimination practice, if it is suspected that the child is having difficulty in this area.

Group Therapy Maximal oppositions therapy may be incorporated into a group therapy format in a variety of ways. However, it is important to keep in mind that targets for maximal opposition therapy should be selected individually, based upon each child’s unique phonemic inventory. It is likely, therefore, that each child will have his own sets of stimulus pictures. Clinicians may want to color code the cards to more easily track targets during sessions with multiple children.

Imitation Phase Therapy activities must ensure that all students obtain sufficient practice on the identified target words, and also allow each participant to proceed from the imitation phase to the spontaneous phase when ready (i.e., when mastery criteria has been met). Simple reinforcers, such as tokens, chips, or stickers are a basic, yet effective way to keep students engaged during rapid drill work.

Spontaneous Phase Practice during the spontaneous phase becomes more variable, although the emphasis continues to be on maximizing sound production practice of the target words. Interactive games, such as matching and sorting words by their initial sound, were utilized in

5 Maximal Opposition Therapy

145

research studies with individual children (Gierut, 1990) and may be effectively incorporated into sessions with multiple children. In fact, the spontaneous phase lends itself to group therapy work, as it is possible for multiple children to engage in sound production practice simultaneously in activities such as the following: 1. What’s missing? Lay out several cards on the table (one from each participating child’s word set). Ask the students to study the cards for a moment and then close their eyes. Remove a card, ask the students to open their eyes, and ask each student, in turn, “Which card is missing?” Of course, for each student, you will want to remove the card that is from their own word set. 2. What doesn’t belong? Lay out several cards representing a category of items, such as food, clothing, or vehicles, and add one target card that does not belong. Ask each child, in turn, to identify the picture that does not belong to the category. Again, you will want to select a target card specific to each student’s needs. 3. What am I hiding? Ask each student, in turn, to select a card from another child’s target word pile and to keep the picture hidden. The child who is now missing a card will ask questions until able to guess the mystery word, or the child who is hiding the card may give clues until the word is discovered. During all of these activities, it is imperative to elicit multiple spontaneous productions of each target word so that adequate sound production practice may occur.

Case Study Initial Data Collection and Identification of Targets Identification of Unknown Sounds Table 5–12 presents a summary of sound production, obtained from the results of an articulation test and spontaneous speech sample, for Irwin, age 4;3. We may use the information presented in Table 5–12 to compile the following lists of Irwin’s known and unknown consonant phonemes: Irwin’s known consonant phonemes: /b, d, m, n, w, j, h/ Irwin’s unknown consonant phonemes: /p, t, k, g, f, v, s, z, θ, ð, ʃ, ɹ, l, tʃ, dZ/ Prior to the initiation of treatment, it is best practice to administer a sound probe in order to establish baseline performance. Each sound is sampled in the initial position of five words. A sampling of all initial consonant sounds will allow us to later measure progress, as well as generalization of skills. Table 5–13 shows Irwin’s initial probe results.

146

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 5–12.  Sound Productions for Irwin; Age 4;3 Early Sounds

I

M

F

Stops

Later Sounds

I

M

F

Fricatives

p

b

b

b

f

Ø

Ø

p

b

3

3

3

v

Ø

Ø

b

t

d

d

d

s

d

Ø

t

d

3

3

3

z

Ø

Ø

d

k

Ø

Ø

t

θ

d

t

t

g

Ø

Ø

d

ð

d

d

d

ʃ

d

t

t

Nasals m

3

3

3

Liquids

n

3

3

3

ɹ

w

w

ə

ŋ

----

n

n

l

w

w

@

Glides

Affricates

w

3

3

----

tS

Ø

t

t

j

3

3

----

dZ

Ø

d

d

3

3

----

Fricative h

As predicted, all phonemes that we listed earlier as “unknown” /p, t, k, g, f, v, s, z, θ, ð, ʃ, ɹ, r, l, tʃ, dZ/) were produced with 0% accuracy during administration of the pretreatment probe. Thus, this informal assessment measure allows us to confirm the results of a previous formal assessment, which yielded our initial phonemic inventory. If Irwin had produced any of the sounds correctly (even once), we would recategorize that sound as a “known” sound.

Identification of Initial Sound Pair Target Next, we must select an initial phoneme pair target to begin intervention. This will include at least one unknown sound from Irwin’s list: /p, t, k, g, f, v, s, z, θ, ð, ʃ, ɹ, r, l, tʃ, dZ/. In our previous example, we selected /k/ as the first unknown sound target for Ethan, according to a developmental model wherein the earliest developing sounds are targeted first. If we follow this same model for Irwin, we might begin with /p/ and then choose one of the following sound pairs that represent maximal contrasts (obtained from Table 5–10): /p/ versus /n/, /p/ versus /j/, /p/ versus /l/, /p/ versus /ɹ/. The pair /p/ versus /n/ or /p/ versus /j/ could be our initial choice because both of these duos include

Table 5–13.  Sound Probe for Irwin Baseline

+/−

+/−

+/−

+/−

+/−

Total

Percentage

Known/ Unknown

p

paw −

pig −

pea −

pen −

poke −

0/5

0%

Unknown

b

big +

bite +

bean +

ball +

boat +

5/5

100%

Known

t

toe −

two −

tan −

tell −

ten −

0/5

0%

Unknown

d

do +

day −

deep +

dime −

down +

3/5

60%

Known

k

key −

comb −

cow −

cup −

can −

0/5

0%

Unknown

g

go −

game −

get −

good −

gate −

0/5

0%

Unknown

m

mad +

man +

map +

me −

mean +

4/5

80%

Known

n

no +

now −

night −

knot +

need +

3/5

60%

Known

w

we +

wet +

week +

wall +

wide +

5/5

100%

Known

j

you +

yell −

yolk −

young −

your +

2/5

40%

Known

h

high +

have +

hill +

him +

home +

5/5

100%

Known

f

fall −

fan −

fed −

fig −

fight −

0/5

0%

Unknown

v

van −

vote −

volt −

vault −

veil −

0/5

0%

Unknown

s

sew −

sing −

sock −

soup −

some −

0/5

0%

Unknown

z

zoom −

zoo −

zip −

zap −

zing −

0/5

0%

Unknown

θ

thank −

think −

thick −

thin −

thing −

0/5

0%

Unknown

ð

that −

the −

they −

then −

them −

0/5

0%

Unknown

ʃ

shake −

shoe −

show −

shut −

ship −

0/5

0%

Unknown

Sounds Stops

Nasals

Glides

Fricatives

continues

147

148

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 5–13.  continued

+/−

+/−

+/−

+/−

+/−

Total

Percentage

Known/ Unknown

ɹ

row −

rain −

road −

ring −

rock −

0/5

0%

Unknown

l

low −

lamb −

like −

late −

lap −

0/5

0%

Unknown

tS

chain −

cheap −

chop −

chin −

chick −

0/5

0%

Unknown

dZ

gym −

jack −

jam −

jeep −

jet −

0/5

0%

Unknown

Sounds Liquids

Affricates

one unknown sound (/p/) and one known sound (/n/ or /j/). Alternately, if we select /p/ versus /l/ or /p/ versus /ɹ/, then treatment of the empty set will begin as these pairs each include two unknown sounds. An alternative way to choose phoneme pairs would consist of choosing later developing sounds. The following sound pairs include /θ/ maximally contrasted with a known sound: /θ/ versus /m/ and /θ/ versus /j/ (see Table 5–10). There is one possible combination that represents maximally opposed sounds that are both unknown to Irwin: /θ/ versus /ɹ/. This would be a choice for Irwin that would incorporate later developing sounds and constitute treatment of the empty set.

Selection of Word Pairs Now that we have identified a potential sound pair for Irwin (/θ/ vs. /ɹ/), we will need to select target words in order to begin intervention. We may choose to pick real words to create our minimal pairs (see Appendix 5–1), such as: thought/rot, think/rink, thigh/ rye, thug/rug, thank/rank. Alternately, we might decide to create nonsense words that represent the contrast between initial /θ/ and initial /ɹ/. In either case, we will use at least five word pairs to implement the therapy phases described earlier in this chapter.

References Barlow, J., & Gierut, J. (2002). Minimal pair approaches to phonological remediation. Seminars in Speech and Language, 23(1), 57–67. Chomsky, N., & Halle, M. (1968). The sound pattern of English. New York, NY: Harper and Row. Elbert, M., & Gierut, J. (1986). Handbook of clinical phonology: Approaches to assessment and treatment. San Diego, CA: College-Hill Press. Gierut, J. A. (1989). Maximal opposition approach to phonological treatment. Journal of Speech and Hearing Disorders, 54, 9–19.

5 Maximal Opposition Therapy

149

Gierut, J. A. (1990). Differential learning of phonological oppositions. Journal of Speech and Hearing Research, 33, 540–549. Gierut, J. A. (1991). Homonoymy in phonological change. Clinical Linguistics and Phonetics, 5, 119–137. Gierut, J. A. (1992). The conditions and course of clinically induced phonological change. Journal of Speech and Hearing Research, 35, 1049–1063. Gierut, J. A. (2001). Complexity in phonological treatment: Clinical factors. Language, Speech, and Hearing Services in Schools 32, 229–241. Gierut, J. A., & Neumann, H. J. (1992). Teaching and learning /th/: A non-confound. Clinical Linguistics and Phonetics, 6(3), 191–200. Velleman, S. L. (2016). Speech sound disorders. Philadelphia, PA: Wolters Kluwer.

150

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 5–1 Minimal Word Pairs of Maximal Oppositions Maximal Contrast Minimal Pair Word Lists for /p/ /p/ versus /n/

/p/ versus /j/

/p/ versus /l/

/p/ versus /ɹ/

pea

knee

pay

yay

pie

lie

pay

ray

pet

net

pot

yacht

pay

lay

pod

rod

pat

gnat

Pam

yam

peep

leap

pipe

ripe

pine

nine

pour

your

pine

line

pot

rot

pot

knot

pep

yep

pad

lad

pig

rig

Maximal Contrast Minimal Pair Word Lists for /b/ /b/ versus /h/ bye

hi

bow

hoe

bat

hat

bed

head

bike

hike

Maximal Contrast Minimal Pair Word Lists for /t/ /t/ versus /m/

/t/ versus /w/

/t/ versus /j/

/t/ versus /ɹ/

tie

my

tea

we

tam

yam

toe

row

toe

mow

type

wipe

tot

yacht

tan

ran

top

mop

tight

white

tip

yip

tub

rub

tug

mug

ton

won

tell

yell

tip

rip

tall

mall

take

wake

tore

your

take

rake

151

5 Maximal Opposition Therapy

Maximal Contrast Minimal Pair Word Lists for /d/ /d/ versus /h/ day

hay

die

hi

dot

hot

did

hid

dog

hog

Maximal Contrast Minimal Pair Word Lists for /k/ /k/ versus /m/

/k/ versus /w/

/k/ versus /j/

/k/ versus /n/

key

me

“K”

way

“K”

yay

key

knee

cat

mat

cut

what

coo

you

cat

gnat

can

man

kite

white

Kip

yip

caught

knot

coat

moat

cake

wake

coke

yolk

cow

now

cap

map

call

wall

card

yard

cut

nut

/k/ versus /l/

/k/ versus /ɹ/

“K”

lay

caw

raw

cane

lane

cat

rat

Kate

late

king

ring

cap

lap

cane

rain

cook

look

code

road

Maximal Contrast Minimal Pair Word Lists for /g/ /g/ versus /h/ guy

hi

go

hoe

gum

hum

got

hot

gold

hold continues

152

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 5–1.  continued

Maximal Contrast Minimal Pair Word Lists for /m/ /m/ versus /t/

/m/ versus /k/

/m/ versus /h/

/m/ versus /s/

my

tie

me

key

my

hi

mow

sew

mow

tow

mat

cat

meat

heat

me

sea

mop

top

man

can

mop

hop

mitt

sit

mug

tug

moat

coat

men

hen

mad

sad

mall

tall

map

cap

mouse

house

mole

sole

/m/ versus /ʃ/

/m/ versus /tS/

me

she

moo

chew

my

shy

mop

chop

mutt

shut

mane

chain

make

shake

mat

chat

mare

share

more

chore

Maximal Contrast Minimal Pair Word Lists for /n/ /n/ versus /p/

/n/ versus /k/

/n/ versus /h/

/n/ versus /f/

knee

pea

knee

key

no

hoe

no

foe

net

pet

gnat

cat

now

how

night

fight

gnat

pat

knot

caught

knee

he

gnat

fat

nine

pine

now

cow

neat

heat

neat

feet

knot

pot

nut

cut

gnat

hat

nine

fine

Maximal Contrast Minimal Pair Word Lists for /w/ /w/ versus /t/

/w/ versus /k/

/w/ versus /h/

/w/ versus /s/

we

tea

way

“K”

way

hay

we

sea

wipe

type

what

cut

we

he

why

sigh

white

tight

white

kite

whip

hip

won

sun

won

ton

wake

cake

wheat

heat

wet

set

wake

take

wall

call

when

hen

walk

sock

153

5 Maximal Opposition Therapy

/w/ versus /ʃ/

/w/ versus /tS/

we

she

wow

chow

what

shut

weep

cheap

weep

sheep

whip

chip

wheat

sheet

week

cheek

win

shin

will

chill

Maximal Contrast Minimal Pair Word Lists for /j/ /j/ versus /p/

/j/ versus /t/

/j/ versus /k/

/j/ versus /h/

yay

pay

yam

tam

yay

“K”

yay

hay

yacht

pot

yacht

tot

you

coo

you

who

yam

Pam

yip

tip

yip

Kip

yam

ham

your

pour

yell

tell

yolk

coke

yacht

hot

yep

pep

your

tore

yard

card

young

hung

/j/ versus /f/

/j/ versus /s/

/j/ versus /θ/

yay

Faye

yay

say

yum

thumb

yacht

fought

you

Sue

yin

thin

yolk

folk

yet

set

yacht

thought

yell

fell

yolk

soak

ying

thing

your

four

your

sore

your

Thor

Maximal Contrast Minimal Pair Word Lists for /h/ /h/ versus /b/

/h/ versus /d/

/h/ versus /g/

/h/ versus /m/

hi

bye

hay

day

hi

guy

hi

my

hoe

bow

hi

die

hoe

go

heat

meat

hat

bat

hot

dot

hum

gum

hop

mop

head

bed

hid

did

hot

got

hen

men

hike

bike

hog

dog

hold

gold

house

mouse

continues

154

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 5–1.  continued

/h/ versus /n/

/h/ versus /w/

/h/ versus /j/

/h/ versus /v/

hoe

no

hay

way

hay

yay

he

“V”

how

now

he

we

who

you

how

vow

he

knee

hip

whip

ham

yam

hat

vat

heat

neat

heat

wheat

hot

yacht

hail

veil

hat

gnat

hen

when

hung

young

hairy

very

/h/ versus /z/

/h/ versus /ð/

/h/ versus /ɹ/

/h/ versus /l/

he

“Z”

hay

they

hoe

row

hoe

low

who

zoo

hen

then

hat

rat

hit

lit

hip

zip

hat

that

home

roam

ham

lamb

whom

zoom

hiss

this

heap

reap

heap

leap

hero

zero

hose

those

hot

rot

hook

look

/h/ versus /dZ/ hoe

Joe

hay

jay

ham

jam

heap

jeep

hear

jeer

Maximal Contrast Minimal Pair Word Lists for /f/ /f/ versus /n/

/f/ versus /j/

/f/ versus /l/

/f/ versus /ɹ/

foe

no

Faye

yay

foe

low

foe

row

fight

night

fought

yacht

fee

Lee

fat

rat

fat

gnat

folk

yolk

fate

late

fight

right

feet

neat

fell

yell

fight

light

fig

rig

fine

nine

four

your

fast

last

foam

roam

155

5 Maximal Opposition Therapy

Maximal Contrast Minimal Pair Word Lists for /v/ /v/ versus /h/ “V”

he

vow

how

vat

hat

veil

hail

very

hairy

Maximal Contrast Minimal Pair Word Lists for /s/ /s/ versus /m/

/s/ versus /w/

/s/ versus /j/

/s/ versus /ɹ/

sew

mow

sea

we

yay

say

sew

row

sea

me

sigh

why

you

Sue

sun

run

sit

mitt

sun

won

yet

set

sat

rat

sad

mad

set

wet

soak

yolk

soap

rope

sole

mole

sock

walk

sore

your

sing

ring

Maximal Contrast Minimal Pair Word Lists for /z/ /z/ versus /h/ “Z”

he

zoo

who

zip

hip

zoom

whom

zero

hero

Maximal Contrast Minimal Pair Word Lists for /θ/ /θ/ versus /m/

/θ/ versus /j/

/θ/ versus /ɹ/

thigh

my

thumb

yum

thigh

rye

thin

Min

thin

yin

thought

rot

think

mink

thought

yacht

think

rink

thick

Mick

thing

ying

thing

ring

Thor

more

Thor

your

Thor

roar continues

156

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 5–1.  continued

Maximal Contrast Minimal Pair Word Lists for /ð/ /ð/ versus /h/ they

hay

then

hen

that

hat

this

hiss

those

hose

Maximal Contrast Minimal Pair Word Lists for /ʃ/ /ʃ/ versus /m/

/ʃ/ versus /w/

she

me

she

we

shy

my

shut

what

shut

mutt

sheep

weep

shake

make

sheet

wheat

share

mare

shin

win

Maximal Contrast Minimal Pair Word Lists for /ɹ/ /ɹ/ versus /p/

/ɹ/ versus /h/

/ɹ/ versus /f/

/ɹ/ versus /k/

ray

pay

row

hoe

row

foe

raw

caw

rod

pod

rat

hat

rat

fat

rat

cat

ripe

pipe

roam

home

right

fight

ring

king

rot

pot

reap

heap

rig

fig

rain

cane

rig

pig

rot

hot

roam

foam

road

code

/ɹ/ versus /s/

/ɹ/ versus /θ/

row

sew

rye

thigh

run

sun

rot

thought

rat

sat

rink

think

rope

soap

ring

thing

ring

sing

roar

Thor

5 Maximal Opposition Therapy

Maximal Contrast Minimal Pair Word Lists for /l/ /l/ versus /p/

/l/ versus /h/

/l/ versus /f/

/l/ versus /k/

lie

pie

low

hoe

low

foe

lay

“K”

lay

pay

lit

hit

Lee

fee

lane

cane

leap

peep

lamb

ham

late

fate

late

Kate

line

pine

leap

heap

light

fight

lap

cap

lad

pad

look

hook

last

fast

look

cook

Maximal Contrast Minimal Pair Word Lists for /tS/ /tS/ versus /m/

/tS/ versus /w/

chew

moo

chow

wow

chop

mop

cheap

weep

chain

mane

chip

whip

chat

mat

cheek

week

chore

more

chill

will

Maximal Contrast Minimal Pair Word Lists for /dZ/ /dZ/ versus /h/ Joe

hoe

jay

hay

jam

ham

jeep

heap

jeer

hear

157

6 Complexity Approaches Treatment Complexity Approach

system that could use wide-spread change

Diagnostic Information Needed

Age Range

n Targeting

n Which

Key Features

more complex structures and later sounds to acquire earlier sounds and system-wide change seems warranted.

n  This is a

n Advantageous

procedure for target selection addressing a phonological

for children with a limited phonetic/ phonemic inventory

3 to 7 years old

Severity of Phonological Disorder Moderate to severe

phonemes are not in inventory, not stimulable, examine implicational universals (with singletons) and sonority (with consonant clusters)

Basic Structure of Therapy

pairs with two individual phonemes using nonwords. n Consonant clusters: Create minimal pairs with one cluster using nonwords. n  Two phases of treatment: Imitation and spontaneous production

n Singletons:

Create minimal

159

160

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Speech-language clinicians are always confronted with the task of determining where to begin in therapy. We typically follow a developmental model that looks at what a child can and cannot produce, beginning with a structure that is not present in the child’s skill set, and that is the developmentally earliest. In relationship to speech sounds, if a child has multiple sounds in error, then a clinician would typically begin with earlier sounds. One underlying assumption is that earlier sounds are easier for the child. Thus, typically [k] and [g] would be targeted before [s] and [z]. Challenging the longstanding clinical belief that treatment should begin with simple structures, evidence points toward the positive effects of using more complex structures for target selection, and as a starting point for treatment. This has been documented in, to mention just a few, syntactic deficits (e.g., Thompson & Shapiro, 2007), lexicalsemantic impairments in individuals with aphasia (e.g., Kiran, 2007), and children with phonological disorders (e.g., Gierut, 2007). Results have shown that training children to produce more complex structures, such as affricates, results in learning simpler, but related structures (e.g., fricatives). Similarly, training clusters results in generalization to singletons. These findings indicate that children with phonological delays, like normally developing children, are able to utilize complex input to advance their phonological systems — a general premise of language learnability theory. As research studies have continued to document the clinical advantages of using complex targets in treatment, (e.g., Gierut, 1999, 2001; Gierut & Champion, 2001), the approach has grown and diversified. Maximal oppositions therapy was one step along the way in the evolution of the principles of complexity (see Chapter 5). The complex target selection in maximal oppositions incorporated maximal distinctions and the empty set. Further development of the concept of complexity led to different ways to select targets. These principles are collectively discussed in this chapter on complexity approaches. How do we select complex input in relationship to speech sounds, and specifically, phonology? Conventional approaches to phonological treatment often recommend that targets are based on stimulability, in other words what the child can produce with some degree of accuracy (Hodson & Edwards, 1997). In addition to stimulability, targets are selected according to a developmental sequence, from earlier to later. However, these recommendations result in only limited generalization (Rvachew & Nowak, 2001). Complex input consists of targeting sounds that are not stimulable, and later developing. When this occurs, improvement of both treated and untreated sounds occur (Gierut, Morrisette, Hughes, & Rowland, 1996). When looking at the phonological system as a whole, complex targets create a more system-wide change. For example, affricates are typically later in acquisition than most fricatives, and fricatives are later than stops. When applying this principle, referred to as the implicational principle, targeting affricates will possibly generalize to earlier target sounds, such as fricatives and stops. However, training earlier sounds does not show change in more complex ones. Thus, if stops are targeted, such as [k] and [g], then change in fricatives and affricates will not occur (e.g., Dinnsen, Chin, & Elbert, 1992; Dinnsen & Elbert, 1984; Schmidt & Myers, 1995). Thus, by targeting specific sounds or sound classes, phonological change can be implied, and it can be predicted. Second, complexity can refer to consonant clusters versus singletons. Based on the complexity concept, training consonant clusters can result in generalization to single-

6 Complexity Approaches

161

tons (e.g., Gierut, 1999, 2001; Tyler & Figurski, 1994). A hierarchy can be established that will provide guidance in determining which consonant clusters would be the most conducive to promoting overall phonological change. In addition, using implicational universals to predict phonological change to target singletons is important for target selection. These principles are outlined in the following paragraphs.

Predicting Phonological Change and Targeting Singletons Gierut and Hulse (2010) state there are four factors that can predict phonological change and influence generalization: (1) the error patterns the child demonstrates, (2) stimulability, (3) implicational universals, and (4) developmental norms. Error patterns refer to those sounds that are present in the child’s inventory and those that are not. Specifically, sounds that are excluded from the child’s phonemic inventory may be optimal treatment targets. Stimulability is also important for this concept. Evidence supports the treatment of nonstimulable sounds (the child cannot imitate following a model) as bringing greater change in the phonological system (Powell, Elbert, & Dinnsen, 1991; Miccio & Elbert, 1996). Gierut (2001, 2004, 2007) states that the child should show 0% accuracy with a model, whereas Barlow, Taps, and Storkel (2010), suggest that nonstimulable is less than 30% accuracy. It seems that clinicians can use their discretion somewhat with the boundary they use for being stimulable versus nonstimulable. Implicational universals describe sound properties in which one property is predictive of another. The presence of certain sound classes is predictive of another sound class. This predictive property could be stated as the presence of “X” implies” Y”, however, the existence of Y does not imply X. As an example of this, fricatives within a phonological system imply there are stops (X implies Y), but the presence of stops does not mean that fricatives are within the system (Y does not imply X). The implicational universals for the sound classes of American English consonants are listed in Table 6–1. Developmental norms are based on those sounds that occur earlier in children’s acquisition, and those that are later. The complexity approach suggests that the treatment

Table 6–1.  Implicational Universals Predicting Property

Property Predicted to Change

Pairs /s, θ/, /z, ð/ or /l, ɹ/

➞

All manner categories

Affricates /tS, dZ/

➞

Fricatives /f, v, s, z, θ, ð, ʃ, Z/

Fricatives /f, v, s, z, θ, ð, ʃ, Z/

➞

Stops /p, b, t, d, k, g/

Liquids /l, ɹ/

➞

Nasals /m, n, ŋ/

Voiced obstruents /b, d, g, v, z, ð, Z, and dZ/

➞

Voiceless obstruents /p, t, k, f, s, θ, ʃ, tS/

Note.  With this matrix, the pairs /l, ɹ/, /s, θ/, /z, ð/ are each given priority in the selection of treatment targets. Source:  Based on works, such as Gierut and Hulse (2010), for example.

162

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

of later-acquired sounds may effect changes in other portions of the phonological system. According to Barlow, Taps Richard, and Storkel (2011), the following guidelines are given when using the complexity approach: Early-acquired sounds [p, b, t, d, k, g, f, v, m, n, ŋ, w, j, h], later-acquired sounds [s, z, ʃ, Z, θ, ð, l, ɹ, tʃ, dZ]. To summarize, sounds (1) excluded from the child’s inventory, (2) that are not stimulable, (3) implying production of other classes of sounds, and (4) that are later developing, have priority as targets. Each of the four principles is discussed in more detail in the sections on Initial Data Collection and Selecting a Target. The following section examines how consonant clusters are targeted. Both twoelement and three-element clusters can be targets. Although some of the variables are the same when selecting clusters versus singletons, certain linguistic principles, such as sonority and markedness, are also utilized. These are discussed in the following section.

Targeting Consonant Clusters Consonant clusters represent complex structures. Two-element clusters, such as [tɹ], are more complex than a singleton, such as [t]. Three-element clusters, such as [stɹ], are more complex than those with two consonants. The basic premise is that targeting consonant clusters will bring change in those singletons within the cluster, as well as other untreated sounds. However, which consonant clusters should one target? As this concept evolved, three factors guided the selection of consonant clusters: sonority, stimulability, and markedness (e.g., Gierut, 1999; Gierut & Champion, 2001; Powell, Elbert, & Dinnsen, 1991). It should be noted that the use of consonant clusters has only been examined in onset clusters. Although many different final consonant clusters exist, these have not been used in the treatment protocols. The following are important concepts when examining consonant clusters (see for example Gierut (1999) or Gierut and Champion (2001). Sonority refers to a specific sound’s relative loudness in respect to other sounds with the same length, stress and pitch (e.g., Ladefoged & Johnson, 2014). Vowels have more sonority than consonants, and certain consonants are relatively louder than others. Liquids and nasals are more sonorous (relatively louder) than voiceless stops. The sonority rankings for each of the consonants are contained in Table 6–2. According to Steriade (1990) the vowels have a sonority ranking of 0. Note that the higher the number, the less sonority that consonant contains. This is rather counterintuitive, as one might expect that increasing the number increases the sonority. However, this is not the case and should be remembered as this principle is exemplified. In addition, the affricates do not have a separate category in the sonority scale and /h/ is not separately assigned a number, one assumes that it has been incorporated into the fricatives. In American English, syllables, including consonant clusters within a syllable, are structured specifically to support this principle of sonority. This is known as the Sonority Sequencing Principle (Clements, 1990). When looking at the sonority of a syllable, the vowel has the greatest sonority, declining in loudness with the onset and coda. In other words, the sonority falls toward both edges of the syllable. Definitions and examples are contained in Box 6–1. For the most part, consonant clusters also follow the pattern of increasing sonority as the sound gets closer to the vowel.

6 Complexity Approaches

163

Table 6–2.  Sonority Rankings of Consonants (Steriade, 1990) Sonority Ranking More Sonority

0

Vowels Consonants

Less Sonority

1

Glides /w, j/

2

Liquids /l, ɹ/

3

Nasals /m, n, ŋ/

4

Voiced fricatives /v, z, ð, Z/

5

Voiceless fricatives /f, s, θ, ʃ/

6

Voiced stops /b, d, g/

7

Voiceless stops /p, t, k/

Box 6–1.  Sonority Sequencing Principle The peak of the syllable, aptly named, is the loudest element of the syllable, the portion with the most sonority, whereas other elements of the syllable have less sonority. Syllables in American English, including most of the consonant clusters, are structured in this manner, with the peak of the syllable, the vowel, being the element with the most sonority. The individual elements of the cluster become more sonorous the closer the sound is to the vowel. The cluster /kɹ/ in “crowd” demonstrates this, the /k/ has less sonority (higher number) than /ɹ/ (lower number). The following is an example of the sonority hierarchy of two words:

More Sonority Less Sonority

“shop” /ʃ/ /ɑ/ /p/ “crowd” /kɹ/ /aU/ /d/ Source:  Adapted from Clements (1990).

The sonority value difference between the consonants of a cluster are used to target specific consonant clusters. The sonority rank of each of the individual consonants within the consonant cluster is subtracted from one another and, in general, the

164

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

consonant cluster with the least sonority difference, the smallest difference, is selected as a target. There are some variables that play a role, and those details are addressed in the section Initial Data Collection and Selecting a Target. Those clusters with the least sonority (smaller differences) are considered to be the most complex, and those with greater differences are considered less complex. See Box 6–2 on how to figure the sonority value difference. There are exceptions, as not all /s/ clusters follow the principle of increasing sonority as the individual elements of the cluster approach the vowel. Thus, specific /s/ clusters are not considered “true” clusters; they have been referred to as “adjunct” clusters (Gierut, 1999). The adjunct /s/ clusters are /sp/, /st/, and /sk/. As adjunct clusters, they do not seem to follow the same treatment generalization and are typically not considered as targets. These are discussed in more detail in Initial Data Collection and Selecting Targets. Stimulability, and its use in target selection were described in the previous section. With consonant clusters as well, cluster choice would be based on nonstimulability. Therefore,

Box 6–2.  How to Calculate the Sonority Difference Values for the Two-Element Consonant Clusters 1. Find any clusters that the child uses. They do not have to be “correct” clusters. 2. Use the table below to subtract the second consonant value from the first. 3. This is the sonority difference value.

Less Sonority

More Sonority

Sonority Ranking

Consonants

1

Glides /w, j/

2

Liquids /l, ɹ/

3

Nasals /m, n, ŋ/

4

Voiced fricatives /v, z, ð, Z/

5

Voiceless fricatives /f, s, θ, ʃ/

6

Voiced stops /b, d, g/

7

Voiceless stops /p, t, k/

Examples: 1) The child says [fw] for [fr] and [fl]. Sonority value of [f] = 5, sonority value of [w] = 1, the sonority difference value (5 − 1) is 4. 2) The child says [tj] for [tr]. Sonority value of [t] = 7, sonority value of [  j ] = 1, the sonority value difference vale (7 − 1) is 6. Based on the sonority principle, the [fw] would be more complex than [tj], [fw] has a smaller sonority value difference.

6 Complexity Approaches

165

the cluster does not appear in the inventory of the child. At least one of the cluster consonants should not be stimulable, and those consonants with smaller sonority differences would have priority (Taps Richard, Barlow, & Combiths, 2017). Targeting three-element clusters, have demonstrated generalization to other clusters and to singletons (Gierut & Champion, 2001). However, this selection process proceeds in a different way, and is discussed in more detail in the section Initial Data Collection and Selecting Targets. Markedness refers to those sounds that are relatively more difficult to produce, and are found less frequently in languages. Marked consonants are more complex, whereas unmarked consonants are considered less complex. Using the complexity approach, marked consonant clusters have priority over unmarked ones. Thus, those clusters with the least sonority differences are considered to be marked (more complex), whereas clusters demonstrating larger sonority differences are considered unmarked (less complex). These principles are further elaborated in Initial Data Collection and Selecting a Target.

Research Supporting Speech Sound Disorders and the Complexity Approach There are many different research articles that support the use of complexity when establishing treatment targets. These investigations date back at least 30 years and document several different aspects of complexity in treating children with phonological disorders. Table 6–3 lists a few of the more recent articles and their details. This table lists only those articles in which children were tested relating to each of the parameters. Not all research supports those parameters outlined in the complexity approach. For example, research published by Rvachew (2005), Rvachew and Bernhardt (2010), and Rvachew and Nowak (2001), found differing results. One illustration: In the Rvachew and Nowak (2001) study, 48 children with moderate to severe delays in their phonological acquisition were treated in accordance with traditional (early developing, less complex phonemes), or nontraditional (later developing, more complex phonemes) as targets. Results demonstrated that those treated with earlier developing phonemes made more progress toward the targeted sounds than treatment with later developing phonemes. When generalization was examined, there was no significant difference between the two groups. Therefore, the group of children treated with more complex phonemes did not show a significantly greater generalization to other sounds within the child’s phoneme system. Although there are opposing views in relationship to the complexity approach, it appears that it is a therapeutic option showing evidence of treatment efficacy. This treatment approach is worth considering for specific children with phonological disorders.

What Is the Age and Severity of Children Who Could Most Benefit from the Complexity Approach? Based on the published research, it appears that the complexity approach can be used for a wide age range of children. Children as young as 2;8 have benefited from treatment with this method; however, generally, the ages ranged from 3 to 7 years old.

Table 6–3.  Examples of Research Supporting the Complexity Approach

Author(s), Date

Number of Subjects

Age Range

Severity

Results

Targeting Sounds Not in the Child’s Inventory:  Nonstimulable Sounds Gierut, Elbert, & Dinnsen, 1987

6 children

3;7 to 4;6

Minimum of 6 sounds in error in at least two sound classes

Approach:  Treated sounds were (a) in inventory, (b) those not in inventory, (c) stimulable, and (d) not stimulable. Results:  Treatment that began with sounds not in inventory and not stimulable seemed to result in a greater reorganization of a child’s overall sound system.

Powell, Elbert, & Dinnsen, 1991

6 children

Miccio, Elbert, & Forrest, 1999

4 children

4;11 to 5;6

3;6 to 4;1

Below 5th percentile on Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015), Four to six sounds missing in inventory

Approach:  Subjects were trained using sounds not in inventory but stimulable.

Six sounds in error across three manner categories, below 5th percentile on the Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015).

Approach:  Subjects were trained on one fricative not in inventory.

Restricted phonological inventories

Approach:  One child was treated with more complex sounds, and one child with less complex sounds.

Results:  The degree of generalization to other sounds varied widely across subjects.

Results:  Generalization to other sounds occurred.

Targeting Later-Developing Sounds Tyler & Figurski, 1994

2 children

2;8 and 2;10

Results:  The child that was treated with more complex distinctions added less complex sounds without direct treatment.

166

Table 6–3.  continued

Author(s), Date

Number of Subjects

Age Range

Severity

Results

Targeting Later-Developing Sounds  continued Gierut, Morrisette, Hughes, & Rowland, 1996

9 children

3;5 to 5;6

At least 5 sounds in error, below the 5th percentile on the Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015).

Approach: Targeted later- developing versus earlier- developing sounds Results:  Relatively greater production accuracy occurred in conjunction with the treatment of later acquired sounds.

Implicational Universals:  One Property Is Predictive of Another Dinnsen, Chin, Elbert, & Powell, 1990

40 children

3;4 to 6;8

Six sounds in error across three manner categories

Approach: Examined implicational universals in respect to the inventories presented by the children. Results:  These universals did seem to exist in the children’s inventories.

Dinnsen, Chin, & Elbert, 1992

34 children

3;4 to 6;8

Six sounds in error across three manner categories

Approach:  Looked at implicational universals in children’s speech. Results:  Affricates implied fricatives, specifically /tʃ/ and /dZ/ implied /ʃ/ and /Z/.

Miccio & Ingrisano, 2000

1 child

5;3

Below the 1st percentile on Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015). No fricatives or affricates occurred in any context.

Approach:  Two fricatives were targeted. Results:  Fricative and affricate sound classes were present after treatment.

continues

167

Table 6–3.  continued

Author(s), Date

Number of Subjects

Age Range

Severity

Results

Targeting Consonant Clusters Gierut, 1998

Gierut, 1999 Experiment 1

Gierut, 1999 Experiment 2

Gierut & Champion, 2001

168

6 children

6 children

5 children

8 children

3;2 to 7;3

3;2 to 7;8

3;5 to 4;8

3;4 to 6;3

Below the 20th percentile on Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015). A minimum of 30 errors on singleton and cluster production.

Approach: Targeted consonant clusters were alternated with treating consonants.

Below the 7th percentile on Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015). A minimum of seven sounds excluded from inventory.

Approach:  Targeted marked, later-developing initial clusters

Below the 7th percentile on Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015). A minimum of seven sounds excluded from inventory.

Approach:  True clusters and adjunct clusters [sp, st, sk] were used to evaluate generalization.

Below the 6th percentile on Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015). A minimum of five sounds excluded from inventory.

Approach: Three-cluster consonants were taught.

Results:  More generalization was noted when children were treated with consonant clusters, or treatment alternated between treating consonants and treating consonant clusters.

Results:  Treatment of marked clusters (later in development) implied unmarked, earlier clusters.

Results:  Varied somewhat across children; however, treatment of marked true clusters implied unmarked clusters. Teaching adjunct clusters [sp, st, sk] did not seem to generalize and, thus, it appears that these clusters are unmarked (less complex) in acquisition.

Results: Posttreatment generalization did not occur to other three-element clusters. However, there was generalization to untreated singletons including affricates. The child’s original inventory plays a role in consonant clusters acquired.

6 Complexity Approaches

169

Thus, the complexity approach can be used for preschool as well as grade-school children. In addition, all children appeared to have a moderate to severe phonological disorder. This was defined as scores below the 7th percentile on the Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015), and typically having between 5 and 7 sounds that were excluded from their inventories across three manner categories (see Table 6–3). These were children with limited phonemic inventories. However, children who had phonemic structure difficulties (problems with stress, sequencing of sounds and/or producing multiple syllables) were not considered to be good candidates for the complexity approach. Therefore, the complexity approach may not be the best possibility for those children with mild speech sound disorders, or for those children with phonemic structure difficulties. These two factors — a moderate to severe phonological disorder and a limited phonemic inventory — are used as the determining variables for several of these phonological treatment protocols. Why would a clinician choose the complexity approach and not, for example, the cycles approach, or multiple oppositions? The complexity approach is based on generalization principles and would be especially valuable for children who have a very limited phonemic inventory. The concept is that by specifically picking a target based on these principles, widespread change in the entire phonological system will occur (See, for example, Taps Richard, Barlow, & Combiths, 2017.) A trial period may be implemented, using probes to determine if change is occurring in other sounds. Probes for the various sounds are contained in Appendix 6–1. Depending on how often therapy occurs per week, treatment emphasis should change if no generalization is observed after 3 to 5 months.

Initial Data Collection and Selecting Targets Singleton Targets This section discusses the selection of individual sounds as stimuli for treatment. It should be kept in mind that based on research findings, treating consonant clusters may be more effective than treating singletons (e.g., Gierut, 1999; Gierut & Champion, 2001). One study has shown that alternating the treatment of consonant clusters with singletons, each for a given time frame, also produced changes in untreated sounds (Gierut, 1998). According to Barlow, Taps, Richard, and Storkel (2010) when no clusters could be found that fit the criteria, then singletons were targeted. However, as late as 2010, Gierut and Hulse projected a matrix to target singletons. This matrix was based on selecting individual sounds and not consonant clusters. Therefore, with this ambiguity, the authors have chosen to present both target selection methods, selecting singleton targets and consonant clusters. As stated earlier, the concept of predicting phonological change uses four factors when selecting singletons: (1) error patterns, (2) stimulability, (3) implicational universals, and (4) developmental norms. Let’s go through each of these using the data that you have collected from a standardized speech assessment you have given the child. If you do not have information from a standardized assessment, the same basic guidelines can be followed. Refer to Appendix 6–1 as a source of words that could be used as probes.

170

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

When going through the steps for the complexity approach with singletons, two sounds are targeted. Based on research results using the complexity approach, it has been demonstrated that the use of two targets that are not in the child’s inventory, and are not stimulable, are the most effective (Gierut, 1992). Therefore, as we move through this section, with an introductory case study, options will be given for selecting two unknown targets. 1. Error patterns:  First, make a list of those sounds present in the child’s inventory. More importantly, a list should be made of those sounds that the child does not produce. If a child produces a sound in some contexts but not in others, put that sound in the category of sounds the child can produce. This list serves as the foundation for the other procedures. Table 6–4 shows an example of data from Christina, age 5;6. As can be seen from Table 6–4, Christina has the following sounds: [p, b, t, d, k, g], [m, n, ŋ], [w, j, h], [f, s], [l]. Sounds are included in this inventory if they only appeared once during testing. Sounds not in her inventory are: [v, z, θ, ð, ʃ ], [ɹ], [tʃ, dZ]. 2. Stimulability:  Sounds that are not stimulable have priority in the complexity approach. Based on the results of your standardized speech assessement, probe error sounds for stimulability. Stimulability testing usually refers to the child’s ability to produce a misarticulated sound in an appropriate manner when “stimulated” by the clinician. Although there are variations in the procedure, the clinician commonly asks the child to “watch, listen, and do what I do.” From stimulability testing, sounds are targeted that are not stimulable. Ideally, that represents 0% accuracy in production. For Christina, the following sounds were stimulable: [d, g, v, s, z]. The following sounds were not stimulable: [ʃ, ɹ, θ, ð, tʃ, dZ]. 3. Implicational universals:  According to this principle, the sound pairs [s, θ], [z, ð], [l, ɹ] have priority as targets (Gierut & Hulse, 2010) (see Table 6–1). Implicational universals suggest that the presence of affricates predicts the presence of fricatives, fricatives the presence of stops, and liquids the presence of nasals. For Christina, one sound of each of the sound pairs [s, θ], [z, ð], [ l, ɹ] is either in her inventory or is stimulable. For example, [s] is stimulable. The same is true for [z]. For the last sound pair [ l, ɹ], [  l  ] is in Christina’s inventory. Using implicational universals, affricates would be a good selection for Christina. The presence of affricates should promote the production of fricatives. 4. Developmental norms:  Later-developing sounds have priority over earlierdeveloping ones. Later-acquired sounds are [s, z, ʃ, Z, θ, ð, l, ɹ, tʃ, dZ].

6 Complexity Approaches

171

Table 6–4.  Summary of Inventory and Stimulability for Christina, Age 5;6 Correct production:  3   Substitution:  Write in the sound   Deletion: Ø Early Sounds

I

M

F

Stimulable

Stops

Later Sounds

I

M

F

Stimulable

Fricatives

p

3

3

3

f

3

3

3

b

3

3

3

v

b

b

b

Yes

t

3

3

3

s

3

t

t

Yes

d

3

Ø

3

z

d

d

d

Yes

k

3

3

3

θ

t

Ø

t

No

g

3

Ø

d

ð

d

d

d

No

ʃ

t

t

t

No

No

Nasals

Yes

Yes

m

3

3

3

Liquids

n

3

3

3

ɹ

w

w

ə

ŋ

3

3

3

l

3

3

3

Glides

Affricates

w

3

3

3

tʃ

t

t

t

No

j

3

3

3

dZ

d

d

d

No

3

3

3

Fricative h

List any consonant clusters that are noted, including substitutions or deletions of any sounds within the cluster. Consonant Clusters:  [gw] (for [gr] and [gl]), [kw] (for [kr]), [tw] (correct in probe words used as a substitute for [tr]). All other clusters demonstrated reduction to one segment. Sonority Difference:  [gw] 6 − 1 = 5, [kw] 7 − 1 = 6, [tw] 7 − 1 = 6

For Christina, the affricates [tʃ] or [dZ] would, again, be considered good targets. They are among the later-developing sounds and their acquisition would imply the presence of fricatives. Targeting two fricatives [ʃ, θ, or ð] would not be as effective. The presence of fricatives implies the presence of stops, however, Christina’s inventory includes the stop consonants, or they are stimulable. The [ɹ], although being a later sound, is not a good choice due to implicational universals. The only sound class that liquids imply is nasals. Nasals are already present in Christina’s inventory.

172

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

To summarize, the following are data for target selection: 1. Sounds not in the child’s inventory, 2. Sounds that are not stimulable, 3. Consider using one of the pairs [s, θ], [z, ð] or [l, ɹ] if both of the pair are not in the child’s inventory and are not stimulable. Otherwise consider using the affricates as targets, 4. Target later developing sounds

Two-Element Consonant Cluster Targets Most standardized speech assessments examine a few consonant clusters, although none of them examine all of the initial consonant clusters of American English. Therefore, each of the consonant clusters that might be used as a target is listed in Table 6–5. Please note that not all initial consonant clusters in American English are included in Table 6–5, only those that could be used as targets. In addition, the sonority differences are noted for each of the clusters. Recall that priority is given to clusters that have the least sonority difference between the consonants of the clusters. If the assessment you are using does not provide enough information on consonant clusters, then words are listed for each of the clusters that could be used as probes. To collect data you would need to determine: (1) the clusters not in the child’s inventory, (2) the smallest sonority difference demonstrated, and (3) the individual consonants that are not in the child’s inventory. Also gather data on the clusters that the child produced that are not clusters of American English, but were possibly used as substitutions for other consonant clusters, such as [fw] (as a substitute for [fɹ], for example), or [θw] as a substitute for [  ʃɹ]. The sonority difference for these clusters will also be necessary information. Please refer back to Box 6–2, page 164. Unlike the target selection for singletons, only one consonant cluster is selected for treatment.

Which Clusters Are Not Targeted for Therapy As noted previously, the clusters [sp], [st], and [sk] do not follow the Sonority Sequencing Principle. As such they are not treated like true clusters. As adjunct clusters they are considered to be unmarked, thus, easier for the child. These clusters are not targeted for treatment. Also the clusters with [j] ([pj, kj, fj, bj, and mj] are not considered treatment targets. However, if your child does produce one of these clusters the sonority difference should be noted (Box 6–2, page 164) and used as a reference for the selection process. The clusters [sm], [sn], and [sl] while being more marked clusters (less sonority difference) are also frequently not used as targets. If these clusters are produced in a similar way as the other s-clusters, they are excluded from target possibilities. For example, if they all show a reduction to one element such as [m], [n], and [l] this would eliminate them from being a target. Or if they all show reduction to [t], for example, they would not be considered good targets. As a general statement, the authors would suggest that these clusters are not targeted or only as an exception.

Table 6–5.  Initial Consonant Clusters That Could Be Targets: Ranked From Least (Smallest) to Most (Largest) Sonority Consonant Clusters

Sonority Calculation

Sonority Difference

sm

5–3=2

2

small, smell, smart, smoke, smile

sn

5–3=2

2

snow, snack, snail, snake, snap

sl

5–2=3

3

slow, sleep, slap, sled, slide

fl

5–2=3

3

fly, flow, flat, flag, flap

fr

5–2=3

3

fry, free, fruit, frog, from

thr [θɹ]

5–2=3

3

throw, three, threw, thread, threat

shr [ʃɹ]

5–2=3

3

shred, shrink, shrimp, shrub, shrug

bl

6–2=4

4

blue, blow, black, block, blink

br

6–2=4

4

brain, brake, brick, bright, broom

dr

6–2=4

4

dry, draw, drum, dream, drink

gl

6–2=4

4

glue, glow, glad, glove, glide

gr

6–2=4

4

grow, gray, green, grape, group

pl

7–2=5

5

play, plow, plum, plant, plane

pr

7–2=5

5

prune, print, proof, proud, prove

tr

7–2=5

5

true, train, trick, trail, trap

cl [kl]

7–2=5

5

clay, clue, clap, clean, clock

cr [kɹ]

7–2=5

5

crow, cry, crab, crumb, crawl

dw

6–1=5

5

dwell, dwarf, dwelling

tw

7–1=6

6

twig, tweet, twin, twelve, twice

qu [kw]

7–1=6

6

quack, quick, quit, queen, quite

Least

Most

Word Examples

Note.  [sp, st, sk] clusters are not on this chart as they are considered adjunct clusters and are unmarked (easier for the child). They are not typically targeted for treatment. The clusters [mj, pj, bj, kj, and fj] are also not included in Table 6–5, as they are clusters that would usually not be targeted.

173

174

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Christina produced the clusters [gw] (for [gɹ]), [kw] (for [kw] and [kɹ]), [tw] (for [tɹ]). If we look at the sonority differences the following are demonstrated: [gw] 6 − 1 = 5, [kw] 7 − 1 = 6, [tw] 7 − 1 = 6. Christina’s smallest sonority difference is 5. Therefore, if possible a cluster with a sonority difference of 4 or even 3 should be targeted. (One typically tries to find a cluster that is smaller by at least one sonority difference). If we look at Table 6–5 we see that the clusters with a sonority of 4 must be eliminated as one of the consonants in the cluster is either stimulable or present in Christina’s inventory. For example, [fl] and [fɹ] can be eliminated due to the correct production of [f] and [ l ]. However, [θɹ] or [ ʃɹ] would possibly both be good targets. In both cases, Christina cannot produce the individual elements of the cluster and the sonority difference equals 3.

Three-Element Consonant Cluster Targets Three-element clusters consist of [skw], [spɹ], [stɹ], [skɹ], and [spl]. Using a three-element cluster in treatment was more effective if the child produced the second and third member of the cluster accurately but not the initial [s] (Morissette, Farris, & Gierut, 2006). Therefore, for [spl] the child should be able to produce [p] and [l]; however, [s] as a singleton should not be a part of the child’s inventory. At least one of the threeelement cluster sounds should be in the child’s inventory, ideally the second and third sound of the three-element cluster should be in the inventory or stimulable (Gierut & Champion, 2010). Using a consonant cluster as a treatment target consists of the following information: 1. Include a cluster not in the child’s inventory, also target clusters if the child lacks clusters altogether (Taps Richard, Barlow, & Combiths, 2017); 2. Look at sonority differences between clusters that do occur in the child’s inventory even if they are produced inaccurately; 3. Select a consonant cluster with a smaller sonority difference than evident in the child’s inventory; 4. Two-element clusters:  select one cluster as a target in which at least one, preferably two of consonants are not present in the inventory and possibly not stimulable; 5. Three-element clusters:  Only one cluster is chosen. Ideally, the child should have the second and third sound of the three-element cluster in the inventory. However, at least one of the three elements should not be in the inventory and not stimulable. To summarize for Christina: Christina has two possibilities for two-element clusters [θɹ] or [ʃɹ]. As she does have [s] in her inventory, then we would need a three-element cluster with at least one sound that is not in her inventory. Christina has [s, t, p, and k] in her inventory, but not [ɹ]. Three-element possibilities are [spɹ], [stɹ], [skɹ].

6 Complexity Approaches

175

Beginning Therapy:  Targeting Singletons The complexity approach is a target selection procedure. It is not a specific treatment option such as cycles, for example. However, research on the complexity approach has systematically outlined the therapy process (e.g., Gierut, 1992, 2001). Nonwords were used in all the studies. Nonwords have been effective in the treatment of phonological disorders (Gierut, Morrisette, & Ziemer, 2010; Leonard, 1973). For example, Gierut et al. (2010) found that nonwords, as opposed to real words, induced greater, more rapid generalization in the phonological system as a function of treatment. In addition, children exposed to nonwords sustained levels of performance even after treatment was discontinued. Nonwords also provide a means of controlling several factors in specific contexts and across children (Gierut, 1999). As demonstrated in the research studies (e.g., Gierut, 1999; Gierut & Champion, 2001) eight word pairs, thus, 16 individual items, with the sound in the initial position were used as stimuli. The stimuli used a variety of vowels and some were actually two syllables. The only other consonants used in the nonwords were [m, n, b, d]. This protocol was used in the sections that follow. As noted in previous information on singletons (see page 171), possible targets for Christina included [tʃ ] and [dZ]. If a clinician thinks the minimal pairs [tʃ ] − [dZ] are too close in perception and production, then another possibility would be [dZ] and [ð], as voiced obstruents predict voiceless ones. Therefore, these two should promote voiceless fricatives. The following examples are minimal pairs using these two possibilities ([tʃ ] and [dZ] or [dZ] and [ð]) and a variety of vowels: [dZup] [tʃup] or [ðup]

[dZi] [tʃi] or [ði]

[dZaU] [tʃaU] or [ðaU] [dZm] [tʃm] or [ðm] [dZɑbi] [tʃɑbi] or [ðɑbi] [dZæd] [tʃæd] or [ðæd] [dZEmeI] [tʃEmeI] or [ðEmeI] [dZInə] [tʃInə] or [ðInə] Therapy consists of two phases, an imitation, and a spontaneous production stage. The first phase, imitation, was based on the clinician providing a model. During this phase, the clinician uses whatever feedback is necessary to achieve the sound. For example, sound placement, cueing, or successive approximations could be used. This phase continues until 75% accuracy is reached over at least two sessions. As many as seven sessions can be spent on this imitation phase. It is important to note that therapy within these research projects was individual therapy, 3 times per week for 60 minutes each session. Most clinicians do not have the luxury of seeing a child that often or that long. The concept of seven sessions could be adjusted accordingly. Activities during this phase could consist of storytelling (the nonsense words could be the characters or some of the actions), sorting, matching, worksheets with the characters, and so forth. Figure 6–1 provides some character and story possibilities.

Figure 6-1. Characters and Nonwords for [dʒ] and [tʃ]

176

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Jobee has on his chobee. This is Jupe and Chupe.

Chee is trying to walk up the jee.

Jinna has a very pointed chinna.

Chemay is listening to his favorite song “Jemay Love”.

Jad says what is chad?

Figure 6–1.  Characters and nonwords for [dZ] and [tʃ ]. Source: Adapted from Bunch of Monsters CC BY by Bruneberg, Digital Art/Drawing & Painting/Fantasy, https://bruneburg.deviantart​ Bunch of Monsters CC BY .com/art/Bunch-of-Monsters-CC-BY-262299101 by Bruneberg

Digital Art/Drawing & Painting/Fantasy

The second phase, spontaneous production, was conducted until an accuracy level of 90% over 3 sessions was realized or maximally 12 sessions. Again adjustments could be made for the time element. Unlike traditional articulation therapy, the treatment goal is not to move systematically through sound production to words, then sentences, and finally conversational speech. The concept of the complexity approach is to create a circumstance whereby change in the phonological system begins to occur. To that end, probes should be used at periodic time intervals. Probes are not being targeted in therapy but contain the sound or sounds that you are trying to achieve as well as other sounds that were not found in the child’s inventory. At regular intervals, the child is asked to repeat the probes and

6 Complexity Approaches

177

the sound in question is judged as being correct or incorrect. The same words are given to the child each time so that a percentage correct can be calculated. Baseline data consist of using probes before and during therapy to measure if systemwide change is occurring. Thus, probes are important for measuring therapy outcome. First, probes were used in the research designs to measure the degree of generalization that occurred in the child’s sound system. Although the targeted sounds may not show progress, other expansions of the phonological system might be occurring. Second, probes could indicate that modifications may need to be made in therapy. If you find that the child is consistently not achieving any accuracy with the probes, this would signal a modification in therapy is necessary. Perhaps the complexity approach is not a good therapy method for this child or the targeted sounds need to be adjusted. Third, probes can be used as a means to document therapy progress in reports that are sent home to parents during the school year and for yearly evaluations. Administering probes is not time consuming. In less than 5 minutes, several probes can be administered and scored. Appendix 6–1 is an example of probes that could be used for the various sounds. As shown in Appendix 6–1, the measure contains five words with the target sound in the initial-word position and five in the final-word position. You can use a set of probes for the sounds targeted in therapy and/or for those sounds originally not in the child’s inventory. If the child is missing many sounds from the inventory, probe some of the sounds one time, and others the next. Probes should be used every two to three months.

Beginning Therapy: Consonant Clusters An analysis of the existing clusters in the child’s inventory is used as the basis for selection. If the child demonstrates clusters with a minimal sonority difference of 5, then sonority differences of 4 or less should be chosen. The individual consonants in the cluster should not be in the child’s inventory. The criteria for selection of three-element clusters are somewhat different. Ideally, the child should not have [s] in her/his inventory, but the other two consonants could be present. This section follows the procedure from several research projects in which at least two of the three sounds of the cluster were in the child’s inventory (Gierut & Champion, 2001). Based on the selection of one consonant cluster for treatment, fifteen nonwords were established. The nonwords contained a variety of vowels and had the following structures: CCVC = 7 of the nonwords, CCVCV = 4 of the nonwords, and CCVCVC = 4 nonwords. The consonants within the nonword (with the exception of the initial cluster) were limited to [m, n, b, d] (Gierut, 1999). Similar to the therapy for singletons, there were two phases of therapy, imitation, and spontaneous production. Imitation, continues until 75% accuracy is reached over two sessions or maximally seven sessions. For two-element clusters the child repeats the nonwords after the clinician’s model. Due to the complex nature of three-element clusters, the model was somewhat different. First, the entire three-element cluster was modeled. If errors were noted, then the following model emphasized the three-element cluster

178

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

[str-pause- i], for example. If still incorrect, then the model proceeded to emphasize the first two sounds, thus, [st ri]. Finally, the model emphasized the last two elements of the cluster [s tri]. Clinicians may have their own way of attempting three-element clusters, but this is one possibility. Gierut (1999) notes that the production of the cluster should be maintained, and clinicians should not train the individual sounds first and then put them together, so to speak. The spontaneous portion of therapy continues until the child can produce the cluster with 90% accuracy over three consecutive sessions or until maximally 12 sessions have been completed. For Christina, there were several options for therapy using consonant clusters. Twoelement targets included [θɹ] or [ʃɹ], while three-element clusters [spɹ], [stɹ] or [skɹ] were possibilities. For this example, let’s choose [stɹ]. The following nonwords could be generated: [stɹɑb] [stɹim] [stɹun] [stɹod] [stɹæd] [stɹm] [stɹen] [stɹumo] [stɹEbi] [stɹedə] [stɹænu] [stɹInod] [stɹobun] [stɹɑməm] [stɹaUbɑd]

Establishing Treatment Goals The complexity approach is designed to create a larger change in a child’s phonological system. There are several hypothetical goals that could be generated based on the treated singleton or cluster production. The change could be documented using intelligibility measures, such as proportion of whole word correctness, or percentage of consonants correct. The process is similar if you have selected consonant clusters as your therapy goal. A baseline statement would be made, and then a stated goal would be made. Because consonant clusters are supposed to generalize to possibly other clusters but also individual sounds, the probes should contain both individual sounds that were not in the child’s inventory and possibly consonant clusters as well. Again, Appendix 6–1 could be used to probe individual sounds and the words in Table 6–5 to probe consonant clusters. The goal would be similar to the one for individual consonants. In referring back to Christina’s case study, it is noted that her singleton choices include [tS - dZ]. Targeting clusters, a three-element cluster [stɹ] was seen as a possible target. Her goals would include a baseline measure and either a goal for the singleton pair [tS - dZ], or the three-element cluster [stɹ]. Possible goals could reflect an increase in speech intelligibility, as measured by the Proportion of Whole Words Correct: PWWC (Ingram & Ingram, 2001), or the Percentage of Consonants Correct: PCC (Shriberg, et al., 1997). See Chapter 2, pages 48–50 for use of these metrics. Table 6–6 contains several possibilities that could be used as goals for Christina. One of these goals could be chosen as an option. The complexity approach is somewhat different than other therapeutic methods when noting and reporting progress. The goal of this treatment is to promote change in

6 Complexity Approaches

179

Table 6–6.  Goal Options for Christina Overall Baseline:  Prior to treatment Christina’s phonetic inventory included the following: [p, b, t, d, k, g], [m, n, ŋ], [w, j], [f, s], [l]. She has not yet acquired these sounds: [v, z, θ, ð, ʃ], [ɹ], [tʃ, dZ]. All unknown sounds will be monitored throughout treatment in addition to the treatment goals. Research suggests that teaching later-developing, nonstimulable sounds will create more change in a child’s phonemic system. Option #1:  Write the intervention goal to expand singleton knowledge. Goal Area #1:  /tS/ and /dZ/ Baseline #1:  Christina produced /tS/ and /dZ/ in probe words with 0% accuracy in the context of imitating words during administration of the baseline probe. Goal #1:  Christina will independently produce six new singleton sounds, including [tS ] and [dZ] in single words with 70% accuracy as measured by an independent probe administered four times per year. Option #2:  Write the intervention goal to expand consonant cluster knowledge. Goal Area #2:  /stɹ/ Baseline #2:  Christina produced /stɹ/ in probe words with 0% accuracy in the context of imitating words during administration of the baseline probe. Goal #2:  Christina will independently produce six new singletons and/or consonant clusters in single words as measured by an independent probe administered four times per year. Option #3:  Write the intervention goal to increase intelligibility of the sound system (using Proportion of Whole Word Correctness). Goal Area #3:  Increased Intelligibility of Sound System (using Proportion of Whole Word Correct Measures or Percentage of Consonants Correct) Baseline #3:  Prior to treatment Christina’s proportion of whole word correctness was 40% for singleton consonants. Goal #3:  Christina will increase her proportion of whole word correctness to 70% in single words as measured by probes administered four times per year. Baseline #4:  Prior to treatment Christina’s overall intelligibility in single words was 40% as measured by Percentage of Consonants Correct. Goal #4:  Christina will increase her intelligibility level to 70% in single words as measured by Percentage of Consonants Correct. Source:  Adapted from Taps, J. (2017) Sample complex phonological goals. Personal correspondence.

the overall phonological system. And as such, when reporting progress, this approach places less emphasis on reaching stated treatment goals and more on noting change. If the child does not reach a high accuracy level of the targeted sound or cluster, then that might not be extremely important if other sounds within the child’s system are emerging. Therefore, using an intelligibility measure for baseline and goals might be the best

180

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

option. Also, keep in mind that this approach uses a set percentage or a set number of sessions to move from the modeling phase to the spontaneous one. The original design, based on one-hour therapy sessions, was 75% or 7 sessions before the child moved to the spontaneous phase. As a clinician you can adjust this accordingly. However, it would be wise to preset a certain number of sessions (or a certain time frame) and then move to the spontaneous phase. Do not attempt the modeling phase for the entire school year.

Group Therapy Ideally, all children in a group are working on the same sound. That would mean that you are using the complexity approach with all children in your group, and that the same sounds were chosen for all. This is probably rarely going to be the case. You will have children doing different sounds and possibly at different skill levels. How does this work with a group? Using individual folders for each child will help get the group running smoothly. Inside the folder are handouts that the child can use for a particular therapy session. Folders might contain the sound that is your target, a word list that will be used for treatment, or sentences that could be used to target certain sounds, for example. This necessitates some type of previous work for the clinician, but would be beneficial and time saving in the long run. The sheets could be structured so that they include work for several therapy sessions. The clinician could have an additional sheet for each child for monitoring progress and documenting the percentage of correct/incorrect for that session. Or, the sheets in the child’s folder could possibly be sent home to the parents as a means of keeping them informed, or as homework. The authors found that a “treasure chest” would be good incentive with 2nd graders through 4th graders. By sending home worksheets to be completed and brought back to therapy (along with the parent’s signature), the child who demonstrates practice would earn points toward an item in the treasure chest. There are always children who do not return the “homework,” but many are motivated by an additional trip to the chest for a prize. Items from The Dollar Store, or 99 Cents Only Store provide a myriad of items that are really motivating for that age group. Do not fill your treasure chests with school items (pencils, pens, small rulers) but rather with items that you know girls or boys would really like. For example, small cars, Silly Putty, stuffed animals (yes, even the boys like stuffed toys) and lip gloss. Some children will want to get a prize for a sibling, so prizes for different age groups are also helpful. Because the complexity approach uses nonwords, even if children do have different sounds, a similar group therapy protocol could be used. First, the child needs to be familiar with the nonwords. Because they are given meaning through stories, pictures, or matching, for example, the clinician will need to go over these during the modeling phase. Other children in the group find it fun to guess the nonwords, or to remember the nonwords of other children. However, what happens to the group when each child has a different goal and/or approach? Perhaps one child is using the complexity approach, one a traditional soundby-sound approach, and one the cycles approach. Working with the concept that each

6 Complexity Approaches

181

child has a folder with therapy stimuli creates more possibilities. Let’s assume you are using a story to establish the nonwords. The child using the complexity approach would model or say the specific nonword spontaneously. Now the child using the traditional approach fills in that part of the story with the next word on her/his handout. If it doesn’t make sense, if it is silly, that is typically fine. Children think it is funny to say nonsense “crazy” things. The child with the cycles approach could also do the same thing with one of the words they are using as targets. Group therapy ideas are as variable as your imagination. The “Dixie Cup” (Rayburn, 2015) is one group activity that could be used with diverse groups such as those consisting of children with language goals, and those children with articulation goals. Words, sounds (possibly words to be put in a sentence for language skills) are written on the bottom of the Dixie cup. Each child gets their group of cups. The final goal is to build a tower with the cups, bottom side up, beginning with a bottom row of “ X” cups, decreasing the row by one cup for each level. Each child repeats their task, (e.g., 10 times) with articulation goals, and then he and she can stack a cup. Each child in the group contributes to the tower until it is done or it falls over. Another game possibility is where each child gets to take a turn at the game board after doing their skill for “X” number of times. Clinicians may have their own way of establishing group ideas, and these are just a couple that could be used.

Case Study Matthew This is the second year of speech therapy for Matthew, age 6;1, who is still fairly unintelligible. During the previous year of therapy, he worked on [k] and [g]. He is also presently struggling with beginning reading and spelling skills. His parents and his teacher are concerned. Table 6–7 summarizes the results of an articulation test for Matthew. Matthew has the following sounds in his inventory: [p, b, t, d, m, n, ŋ (in final position), w, j, h, and ʃ (in initial and final position)]. Matthew does not have the following sounds in his inventory: [k, g, f, v, s, z, θ, ð, ɹ, l, tʃ, dZ]. Matthew’s errors include [k] → [t], [g] → [d], [f] → [p], or is deleted, [v] → [b], [s] → [θ], [z] → [ð or θ], [θ] → [f ], [ð] → [d], [ɹ, l] → [w] ,or [@] in word-final position, [tʃ ] → [ ʃ ] and [dZ] → [tʃ ]. Initial probes indicate that Matthew has 0% accuracy, and is not stimulable for [g, f, v, s, z, θ, ð, ɹ, and l]. Targeting singletons:  The pairs that have priority include [s, θ], [z, ð] and [l, ɹ]. Matthew is not stimulable for the sounds in all of those pairs. According to the implicational universals (see Table 6–1), fricatives imply stops. By targeting a fricative pair, such as [s, θ] or [z, ð], then other fricatives ([f, v]) and stops (such as [g]) could emerge. This is not the case if the liquids [l, ɹ] are targeted. Also, voiced obstruents imply voiceless ones. Based on this rationale, the best target pair for Matthew would be [z, ð].

Table 6–7.  Sound Productions for Matthew, Age 6;1 Early Sounds

I

M

F

Stimulable

Stops

Later Sounds

I

M

F

Stimulable

Fricatives

p

3

3

3

f

p

Ø

p

No

b

3

3

3

v

b

b

b

No

t

3

3

3

s

θ

θ

θ

No

d

3

3

3

z

ð

ð

θ

No

k

t

t

t

Yes

θ

f

f

f

No

g

d

d

d

No

ð

d

d

d

No

ʃ

3

3

3

Nasals m

3

3

3

Liquids

n

3

3

3

ɹ

w

Ø

@

No

ŋ

----

n

3

l

w

w

@

No

Yes

Glides

Affricates ----

tʃ

ʃ

ʃ

ʃ

Yes

3

----

dZ

tʃ

tʃ

tʃ

Yes

3

----

w

3

j

3

Fricative h

List any consonant clusters that are seen in the speech sample. Consonant Clusters: Initial Medial Final [tɹ] → [tw] [dɹ] → [dw]



[pɹ] → [pw]



[gɹ] → [dw]



[bl] → [bw]













All other consonant clusters are reduced to one consonant. Sonority differences:  [tw], [pw] 7 − 1 = 6, [bw], [dw] 6 − 1 = 5 The smallest sonority difference is 5. One should try to target a consonant cluster with a sonority difference of 4 or 3.

182

6 Complexity Approaches

183

Nonword pairs for singletons:  Use a variety of vowels in one- to two-syllable nonwords. Other consonants are limited to [m, n, b, d]. [zi] [ði]

[zubi] [ðubi]

[zɑb] [ðɑb] [zaUbo] [ðaUbo] [zæd] [ðæd]

[zIn] [ðIn]

[z] [ð] [zEmi] [ðEmi] Targeting consonant clusters:  Matthew’s cluster inventory includes [tw, dw, pw, bw]. All other consonant clusters are reduced to one consonant. The sonority differences are as follows: [tw], [pw] 7 − 1 = 6, [bw], [dw] 6 − 1 = 5. Therefore, one should try to target a consonant cluster with a sonority difference of 4 or 3. If we look at Table 6–5, we find that there are several options for Matthew. If we choose a cluster in which both elements are not stimulable and unknown, the following are possibilities: [gɹ, gl, θɹ, fɹ, sl]. The s-clusters, [sn], [sm], and [sl] are usually not good targets and, in this case, they also are not. If a child uses a similar pattern across all s-clusters, (which Matthew does, he reduces them to one singleton consonant), then they are not acceptable targets. If we look at the consonant clusters with a sonority difference of 3 or 4, and also consider nonstimulability, we find that the following clusters could be potential targets: Sonority difference of 4, [gɹ, gl], sonority difference of 3, [fɹ, fl, θɹ]. Targeting three-element clusters: In order to use three-element clusters, the child should be able to produce the second and/or third element of the cluster. Matthew can do this with all of the three-element clusters: [skw, spɹ, stɹ, skɹ, spl]. If we narrow this down to three-element clusters, in which both the second and third element are stimulable, which is the ideal target possibility, we arrive at [skw]. Therefore, a clinician could work on three-element clusters. The nonwords could be structured accordingly: [skwɑ] [skweI] [skwɔI] [skwoUd] [skwim]

[skwEb]

[skwIm] [skwun] [skwIpi] [skwaUdoU] [skwaIbə] [skwme] [skwæbu] [skwɑmEn] [skwɔIbId] [skwæd@d] [skwibɑd] [skwumən]

Case Study: Irwin Irwin is the case study that was developed in Chapter 5: Maximal Oppositions (pages 145– 148). Because of some of the similarities between maximal oppositions and the complexity approach, this case study is used to demonstrate the differences that might, or might not, occur in target selection for singletons. The following was noted in Irwin’s inventory: Irwin’s known consonant phonemes (in his inventory):  /b, d, m, n, w, j, h/ Irwin’s unknown consonant phonemes (not in his inventory):  /p, t, k, g, f, v, s, z, θ, ð, ʃ, ɹ, l, tʃ, dZ/.

184

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

All unknown sounds were produced with 0% accuracy. Irwin has a very restricted phonemic inventory. If we apply the principles of the complexity approach for singleton consonants, the following pairs would be given priority: [s, θ], [z, ð] and [l, ɹ]. Irwin is not stimulable for all of the mentioned sounds. If we examine implicational principles for targets, fricatives imply stops and voiced sounds imply voiceless ones. Therefore, selecting [z, ð] could possibly impact [p, t, k, g], other fricatives and/or voiceless consonants. This would be the pair chosen if the complexity approach was being implemented. Using maximal oppositions, the pair [θ] versus [ɹ] was chosen. The different processes for target selection for the complexity approach versus maximal oppositions may indeed yield different results.

References Barlow, J., Taps Richard, J., & Storkel, H. (2010). Phonological assessment and treatment target selection. Seminar presented at the national convention of the American Speech-LanguageHearing Association, San Diego, CA. Clements, G. N. (1990). The role of the sonority cycle in core syllabification. In J. Kingston and M. E. Beckman (Eds.) Papers in laboratory phonology I: Between the grammar and the physics of speech (pp. 283–333). Cambridge, UK: Cambridge University Press. Dinnsen, D. A., Chin, S. B., & Elbert, M.(1992). On the lawfulness of change in phonetic inventories. Lingua, 86, 207–222. Dinnsen, D. A., Chin, S. B., Elbert, M., & Powell, T. W. (1990). Some constraints on functioanly disordered phonologies: Phonetic inventories and phonotactics. Journal of Speech and Hearing Research, 33, 28–37. Dinnsen, D. A., & Elbert, M. (1984). On the relationship between phonology and learning. In M. Elbert, D. A. Dinnsen, & G. Weismer, (Eds.), Phonological theory and the misarticulating child, ASHA Monographs No. 22, (pp. 55–68), Rockville, MD: ASHA. Gierut, J. A. (1991). Homonymy in phonological change. Clinical Linguistics and Phonetics, 5, 119–137. Gierut, J. A. (1992). The conditions and course of clinically-induced phonological change. Journal of Speech and Hearing Research, 35, 1049–1063. Gierut, J. A. (1998). Natural domains of cyclicity in phonological acquisition. Journal of Clinical Linguistics and Phonetics, 12, 481–499. Gierut, J. A. (1999). Syllable onsets: Clusters and adjuncts in acquisition. Journal of Speech, Language, and Hearing Research, 42, 708–726. Gierut, J. A. (2001). Complexity in phonological treatment: Clinical factors. Language, Speech, and Hearing Services in Schools, 32, 229–241. Gierut, J. A. (2004). Clinical application of phonological complexity. CSHA Magazine, 34, 6–7. Gierut, J. A. (2007). Phonological complexity and language learnability. American Journal of Speech-Language Pathology, 16, 6–17. Gierut, J. A., & Champion, A. H. (2001). Syllable onsets II: Three-element clusters in phonological treatment. Journal of Speech, Language, and Hearing Research, 44, 886–904. Gierut, J. A., Elbert, M., & Dinnsen, D. A. (1987). A functional analysis of phonological knowledge and generalization learning in misarticulating children. Journal of Speech and Hearing Research, 30, 462–479.

6 Complexity Approaches

185

Gierut, J. A., & Hulse, L. E. (2010). Evidence-based practice: A matrix for predicting phonological generalization. Clinical Linguistics and Phonetics, 24, 323–334. Gierut, J. A., Morrisette, M. L., Hughes, M. T., & Rowland, S. (1996). Phonological treatment efficacy and developmental norms. Language, Speech, and Hearing Services in Schools, 27, 215–230. Gierut, J. A., Morrisette, M. L., & Ziemer, S. M. (2010). Nonwords and generalization in children with phonological disorders. American Journal of Speech-Language Pathology, 19, 167–177. Goldman, R., & Fristoe, M. (2015). Goldman-Fristoe Test of Articulation (3rd ed.). Boston, MA: NCIS Pearson. Hodson, B. W., & Edwards, M. L. (1997). Perspectives in applied phonology. Gathersburg, MD: Aspen. Ingram, D., & Ingram, K. D. (2001). A whole-word approach to phonological analysis and intervention. Language, Speech, and Hearing Services in Schools, 32, 271–283. Kiran S. (2007). Complexity in the treatment of naming deficits. American Journal of Speech Language Pathology, 16, 18–29. Ladefoged, P., & Johnson, K. (2014). A course in phonetics (7th ed.). Boston, MA: Cengage Learning. Leonard, L. B. (1973). Referential effects on articulatory learning. Language and Speech, 16, 45–56. Miccio, A. W., & Elbert, M. (1996). Enhancing stimulability: A treatment program. Journal of Communication Disorders, 29, 335–351. Miccio, A. W., Elbert, J., & Forrest, K. (1999). The relationship between stimulability and phonological acquisition in children with normally developing and disordered phonologies. American Journal of Speech-Language Pathology, 8, 347–363. Miccio, A. W., & Ingrisano, D. R. (2000). The acquisition of fricatives and affricates: Evidence from a disordered phonological system. American Journal of Speech-Language Pathology, 9, 214–229. Powell, T. W., Elbert, M., & Dinnsen, D. A. (1991). Stimulability as a factor in the phonologic generalization of misarticulating preschool children. Journal of Speech and Hearing Research, 34, 1318–1328. Rayburn, J. (2015, July, 29). Articulation therapy: Five activities to elicit 100+ trials. Retrieved from thespeechroomnews.com/2015/07/articulation-therapy-5-activities-to-elicit-100-trials.html/ Rvachew, S. (2005). Stimulability and treatment success. Topics in Language Disorders, 3, 207–219. Rvachew, S., & Bernhardt, B. M. (2010). Clinical implications of dynamic systems theory for phonological development. American Journal of Speech-Language Pathology, 19, 34–50. Rvachew, S., & Nowak, M. (2001). The effect of target-selection strategy on phonological learning. Journal of Speech, Language, and Hearing Research, 44, 610–623. Schmidt, A. M., & Meyers, K. A. (1995). Traditional and phonological treatment for teaching English fricatives and affricates to Koreans. Journal of Speech and Hearing Research, 38, 828–838. Shriberg, L. D., Austin, D., Lewis, B. A., McSweeney, J. L., & Wilson, D. L. (1997). The percentage of consonants correct (PCC) metric: Extensions and reliability data. Journal of Speech, Language, and Hearing Research, 40, 708–722. Taps Richard, J. (2012). Applying complexity principles. Retrieved from http://slpath.com/applying​ complexityprinciples.html Taps Richard, J. (2017) Sample complex phonological goals. Personal correspondence, August 4, 2017. Taps Richard, J. Barlow, J. A., & Combiths, P. N. (2017). Applying phonological complexity in the schools: Insights from 32 case studies. Paper presented at the national convention of the American Speech-Language-Hearing Association, Los Angeles, CA. Thompson C. K., & Shapiro L. P. (2007). Complexity in treatment of syntactic deficits. American Journal of Speech-Language Pathology, 16, 30–42. Tyler, A. A., & Figurski, G. R. (1994). Phonetic inventory changes after treating distinctions along an implicational hierarchy. Clinical Linguistics & Phonetics, 8, 91–107.

186

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 6–1 Probes for Collecting Baseline Data and Monitoring Progress [p]

[b]

[t]

[d]

[k]

[g]

paw

big

toe

do

key

go

pig

bite

two

day

comb

game

pea

bean

tan

deep

cow

get

pen

ball

tail

dime

cup

good

poke

boat

ten

down

can

gate

up

tub

bat

bad

back

bag

lip

cab

boot

bed

book

hog

top

knob

hot

hide

duck

dog

tape

mob

out

wood

look

dig

mop

web

nut

need

pack

leg

[m]

[n]

[ŋ]

mad

no

man

now

map

night

me

knot

mean

need

came

pin

wing

dime

bone

long

name

down

hang

time

one

king

gum

pain

bang

[f]

[v]

[s]

[z]

[ ʃ ]

[Z]

fall

van

sew

zoom

shake

Jacques

fan

vote

sing

zoo

shoe

fed

volt

sock

zip

show

fig

vault

soup

zap

shut

6 Complexity Approaches

187

[Z]

[f]

[v]

[s]

[z]

[ ʃ ]

fight

veil

some

zing

ship

beef

cave

bus

bees

cash

beige

cough

dive

face

boys

leash

rouge

half

five

gas

does

brush

leaf

give

mess

is

wash

tough

have

yes

noise

wish

[θ]

[ð]

[tʃ ]

[dZ]

thank

that

chain

gym

think

the

cheap

jack

thick

they

chop

jam

thin

then

chin

jeep

thing

them

chick

jet

bath

smooth

beach

age

math

bathe

catch

cage

moth

breathe

peach

edge

tooth

clothe

match

huge

teeth

loathe

witch

fudge

[ l ]

[ɹ]

[ j]

[w]

[h]

low

row

you

we

high

lamb

rain

yell

wet

have

like

road

yolk

week

hill

late

ring

young

wall

him

lap

rock

your

wide

home

ball

bear

bull

four

pail

her

will

more

fall

deer

7 Phonotactic Therapy Treatment Phonotactic Therapy

Age Range Not specified

Severity of Phonological Disorder Moderate, severe, or profound Demonstrate restricted word shape inventories or whole word error patterns

Key Features n Simultaneously

expands a child’s

word shape inventory, sound inventory, and vocabulary n  Targets new phonemes within the context of established word shapes and new word shapes utilizing established phonemes

Diagnostic Information Needed Independent analyses of phonotactic inventory (syllable

and word shapes) and phonemic (sound) inventory Relational analysis of error patterns, if possible

Target Selection Factors for target selection include consideration of developmental norms, impact on intelligibility, potential to maximize change, emerging skills, and individual communication needs.

Basic Structure of Therapy n Intervention focuses on expanding a child’s word shape and sound inventories n Strategies involve imitating the learning patterns of typically developing children or using a child’s current skills to facilitate production of a new skill.

Least

189

190

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Phonotactic therapy, as described by Velleman (2002, 2016), emphasizes word shapes as important structures for creating contrast in a child’s developing phonological system. This approach to target selection allows clinicians to simultaneously expand a child’s word shape inventory, sound inventory, and vocabulary. Phonotactic intervention expands both the syllable and word structure possibilities of the child while also introducing new phonemes within these structures. It represents a practical option for children who demonstrate restricted word shape inventories or who exhibit whole word error patterns (e.g., reduplication or harmony). We must have a solid understanding of three principles that are fundamental to this approach before we consider the details of implementation: 1. The vital role of phonotactics in children’s sound systems 2. Words as the fundamental learning unit of phonological intervention 3. The developmental trajectory of phonotactic learning Let’s look at each of these concepts in turn.

Phonotactics:  What is phonotactics? Why is it important? Traditionally, articulation therapy focused on the physical production of individual sounds. Yet sounds in isolation are of little help when trying to communicate in daily life. Children must be able to combine sounds in the context of spoken words and sentences. Phonotactics describes the rules that govern these combinations. Sounds combine to create syllables, and syllables, in turn, are the building blocks of words. Every language has its own rules for how syllables and words may be constructed. For example, English allows the use of /s/ blends as onsets in words such as “school” and “star.” Spanish, on the other hand, has a prohibition against /s/ blends as onsets, yet allows their presence in the first syllable of words if preceded by /E/, resulting in words such as “escuela” (school) and “estrella” (star). In addition to dictating where clusters may appear in words, phonotactics also includes rules about how consonants may combine to form clusters, how consonants and vowels may combine to form words, and how syllables may join together in words, including their stress patterns. The term “syllable shape” refers to the particular combination of consonants and vowels included in a syllable. The letter “V” is used to signify a vowel sound and the letter “C” to indicate a consonant sound, when representing a child’s productions. Syllable shapes range from very simple (e.g., V for [aI], “I,” or CV for [noU], “no”) to complex (e.g., CCCVCCC for [stɹEŋθs], “strengths”). A “word shape” includes the sounds and syllables that construct a given word, and range from short, simple shapes (e.g., V for [aI]) to long, multisyllabic constructions (e.g., CVCVCVCVCVC for [hI.poU.pɑ[email protected]@s], “hippopotamus”). For one syllable words, such as “no” or “cat,” the syllable and word shape will be synonymous (e.g., CV or CVC, respectively). Multisyllabic words, on the other hand, may contain more than one syllable of the same shape or more than one syllable of differing shapes. For example, the word “mama” contains two CV syllables

7  Phonotactic Therapy

191

in the word shape CVCV and the word “upon” contains one V syllable and one CVC syllable in the word shape VCVC. Why is attention to these syllable and word shapes important for children with speech sound disorders? Mastering the phonotactic rules of one’s native language is just as necessary for effective communication as learning consonant and vowel sounds, because all have an essential role in creating contrast (Velleman, 2016). When attempting to build contrast in a child’s disordered sound system, we often begin with teaching phonemic contrast — that is, the distinctions between sounds. For example, we might focus on helping a child to produce words that begin with /t/ and /s/ in word pairs such as “tea” and “see” or “tap” and “sap.” Contrasting sounds in words is essential for reducing homonymy and building an intelligible vocabulary. Yet, we must not overlook the vital role of phonotactics in producing contrast as well. A variety of word shapes are the necessary building blocks for creating an expanding lexicon. As a child masters more word shapes over time (moving from, say, CV to CVC to CCVC to CCVCVC), he or she is able to say an increasing number of words. A restricted phonotactic inventory may impair intelligibility as much, if not more, than a limited sound repertoire.

Words:  How are words the fundamental unit of phonological learning? Just as children do not communicate in isolated sounds, there is reason to believe that they do not learn sounds as entities separate from words during the acquisition of language. Instead, as children attempt to imitate the speech of adults, they gain phonological knowledge through their attempts to say meaningful words. Velleman and Vihman (2002) refer to this as lexically based learning, and suggest that therapy should mimic natural development by focusing on phonological learning at the word level rather than the sound level. Phonotactic therapy offers an option for doing just that. By identifying word shapes as a focus of intervention, clinicians are invited to consider the relationship between phonological and lexical development. Research indicates that there is a strong relationship between vocabulary and phonological development, with word learning and acquisition of sounds mutually influencing one another. Expansion of vocabulary in young children has been shown to occur simultaneously with growth in sound inventories (Storkel & Morrisette, 2002). In addition, a child’s phonetic repertoire may influence lexical development. It has been documented that words containing sounds that are easier for children to produce are more likely to be included in their early lexicons (Leonard, Schwartz, Folger, Newhoff, & Wilcox, 1979; Stoel-Gammon & Cooper, 1984). In fact, children tend to attempt new words if they begin with familiar sounds and syllable shapes (Lederer, 2002). Work completed by Gierut, Morrisette, and Champion (1999) indicates that there remains a bidirectional relationship between lexical and phonological learning beyond the 50-word stage. This research suggests that selection of frequently occurring words during sound production therapy facilitates sound learning and results in greater generalization to untreated words and contexts. These research findings highlight the interactive and dynamic relationship between phonotactic, phonemic, and lexical development in children’s developing sound systems. We will see how phonotactic therapy is uniquely suited to address these three domains of development simultaneously.

192

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Development:  What is the developmental trajectory of phonological learning? An important feature of phonotactic therapy is its focus on the child’s developing skills. As such, it provides a potential intervention framework for children with very limited sound production skills. When confronted with a child who is able to produce only a few sounds and word shapes, we may wonder, where do I begin? For clinicians who are working with children who have so much to learn, it can be overwhelming to choose a starting place. Phonotactic therapy advises us to begin at the production level that the child demonstrates and build their communication skills in a developmental progression. This is a path that is familiar to clinicians and parents alike. As a child gains more skills, a different phonological intervention approach may become more appropriate. In the meantime, phonotactic intervention may be the ideal method for beginning the journey to intelligible speech. As we will see later in this chapter, typical development patterns are not the only criteria used for target selection when implementing a phonotactic approach. However, because they are an important factor, we will want to reference information regarding acquisition norms. There is much individual variation, yet researchers have established some guidelines for phonotactic development (e.g., Velleman, 2003). Table 7–1 includes expected ages of acquisition of early word shapes.

Table 7–1.  Phonotactic Inventory Development Word Shape

Expected Age*

Example

CV

12 months

“pa”

VC

12 months

“up”

C1V1C1V1

8 to12 months

“mama”

C1V1C1V2

8 to 12 months

“baby”

V1CV2

12 months

“Abby”

C1V1C2V2

12 to 18 months

“bunny”

C1VC1

18 to 24 months

“pop”

C1VC2

24 to 30 months

“moon”

CCV

30 to 36 months

“pway” (for “play”)

CCVC

30 to 36 months

“twain” (for “train”)

VCC and CVCC

30 to 48 months

“ant” and “hand”

*Ages are approximate. Source: Compiled from Chin & Dinnsen (1992); Grunwell (1985); McLeod, van Doorn, & Reed (2001); Smit, Hand, Freilinger, Bernthal, & Bird (1990); Stoel-Gammon (1987); Velleman (2003); and Watson & Scukanec (1997).

7  Phonotactic Therapy

193

As seen in Table 7–1, the earliest words produced by children are typically comprised of one syllable, consonant-vowel combinations (e.g., “no” or “bye”) or two-syllable, reduplicated sequences (e.g., “mama” or “bye-bye”; Velleman, 2003). Over time, children also begin to produce closed syllables, including VC (e.g., “up”) and CVC (e.g., “pop”). The complexity of individual syllables increases as children develop, as does the number of syllables that children are able to produce in a single word. By 30 months, most children begin to produce simple cluster sequences in the initial and/or final positions of words, although their productions may not be accurate (e.g., [pweI] for “play”; McLeod, van Doorn, & Reed, 2001). These skills will continue to grow over time until a child is able to produce syllable shapes like CCCV for “straw” and combine such shapes into complex multisyllabic words like “strawberry.” At the same time as children are acquiring these foundational phonotactic skills, their sound inventory is expanding as well (Velleman, 2003). Table 7–2 provides a visual representation of this dual developmental process. A child’s individual developmental trajectory also plays an important role in phonotactic intervention. Although reference to typical norms is important, it is even more vital to understand how each child’s individual phonological system is developing. Phonotactic intervention begins with what the child already knows, and uses this knowledge to form the foundation for acquisition of new skills. To this end, researchers have

Table 7–2.  Phonotactic and Sound Inventory Development Age

Word Shapes

1;6 to 2;0

C1V1“no” V1C1“up” C1V1C1V1“mama” C1V1C1V2“baby”

m b p w

n

C1V1C2V2“happy” C1VC1“pop”

m b p w

n d t

m b p w f

n d t

m b p w v f

n d t

2;0 to 2;6

2;6 to 3;0

3;0 to 3;6

C1VC2“moon”

CCV“pway”   (stop + glide) CCVC“twain”   (stop + glide)

Segments

h ŋ g k h ŋ g k j h

s

ŋ g k j ʃ tS

h

l Source:  Compiled from Grunwell (1985); Bleile (1995); Velleman (1998, 2003).

194

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

recommended targeting new word shapes utilizing sounds that are already in a child’s inventory (Ingram & Ingram, 2001; Velleman, 2016). In addition, Velleman (2016) advocates for using emerging abilities to foster continued growth. For example, if a child is beginning to demonstrate occasional use of the word shape CVC, this may become an appropriate target for intervention. As the name implies, the acquisition of new phonotactic skills is the primary focus of phonotactic therapy. Yet as we saw illustrated in Table 7–2, development of a child’s word shape inventory occurs simultaneously with the expansion of the sound inventory. Therefore, intervention may include the concomitant targeting of sound acquisition. In this chapter, we include both phonotactic and segmental goals, following recommendations to target new phonemes within the context of established word shapes, while conversely, targeting new word shapes utilizing established phonemes (Ingram & Ingram, 2001; Velleman, 2016).

Research Supporting Phonotactic Therapy Phonotactic therapy enjoys a strong theoretical foundation. However, a body of research to provide evidence for its effectiveness does not yet exist. Velleman (2002) describes a case study of a preschool age boy who made significant progress toward phonotactic and phonetic goals when simultaneously addressed within a phonotactic therapy framework. Hopefully, researchers will be inspired to conduct experimental investigations of phonotactic therapy to provide clinicians with appropriate practice guidelines in the future.

What Is the Age and Severity of Children Who Could Most Benefit from Phonotactic Therapy? Children who demonstrate difficulties with the production or contrastive use of syllable and word shapes will benefit from phonotactic therapy. These children typically present with moderate, severe, or even profound speech sound disorders, and reduced speech intelligibility. Although there is not available research supporting a specific age that is optimal for using phonotactic therapy, one would assume that this therapy would be appropriate for young children. In Velleman’s (2002) article on phonotactic therapy, the case study was a child age 3;4. Weiner (1981) successfully utilized a similar method with two 4-year-old children. It appears that this therapy might be suitable for a wider age range if the children meet the following criteria. Appropriate candidates for phonotactic intervention will: 1. Demonstrate restricted syllable and word shape inventories (per independent analysis), and/or 2. Exhibit error patterns that reflect syllable and word level difficulties (per relational analysis).

7  Phonotactic Therapy

195

Error patterns that are appropriate targets for phonotactic intervention are presented in Table 7–3, utilizing the terminology of phonological processes. Table 7–3 includes common phonological processes that describe whole word error patterns. Most fall under the aptly named category of syllable structure processes, including initial consonant deletion, final consonant deletion, syllable deletion, cluster reduction, and reduplication. Harmony is not usually considered a syllable structure process, and yet assimilation errors occur at the level of the whole word rather than at the level of individual segments. Children who demonstrate assimilation errors have difficulty changing the vowels or consonants within a word. They might say [pɑp] for “pot” or [gɑg] for “dog.” These errors cannot be accounted for simply as sound substitutions. Instead, the composition of the syllable and words in question impact the child’s productions. The list presented in Table 7–3 is not exhaustive. Any difficulties using different vowels, different consonants, or different syllables within the same word may also be appropriate targets, as well as stress production errors that occur at the word level (Velleman, 2016). It is also important to note that the examples included in Table 7–3 represent relatively simple word shapes. These same errors might be demonstrated in more complex words. For example, a child might exhibit initial or final consonant deletion in multisyllabic words (e.g., [æ[email protected]] for “cantelope” or [[email protected]] for “telephone”). Children who present with Childhood Apraxia of Speech (CAS) may also benefit from a phonotactic approach to intervention, given that these children often demonstrate difficulty producing word shapes correctly. Velleman (2016) proposes that targeting established sounds in the context of established word shapes may improve automaticity for children with CAS. In this chapter, however, phonotactic intervention principles are presented for children whose primary deficit is expressive phonological skills, rather than motor planning.

Table 7–3.  Potential Targets for Phonotactic Therapy

Examples

Syllable Structure Change

Initial Consonant Deletion

[oU] for [boU] “bow” [oUt] for [boUt] “boat”

V for CV VC for C1V1C2

Final Consonant Deletion

[] for [p] “up” [k] for [kp] “cup”

V for VC CV for C1V1C2

Cluster Reduction

[teI] for [steI] “stay” [teIn] for [steIn] “stain”

CV for C1C2V1 C1V1C2 for C1C2V1C3

Syllable Deletion

[næ] for [næ.n@] “nana” [næ.n@] for [[email protected]æ.n@] “banana”

CV for C1V1C1V2 C1V1C1V2 for C1V1C2V2C2V1

Reduplication

[pɑ.pɑ] for [poU.ni] “pony”

C1V1C1V1 for C1V1C2V2

Harmony (i.e., Assimilation)

[pɑp] for [pɑt] “pot”

C1V1C1 for C1V1C2

196

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Initial Data Collection and Selecting Targets Comprehensive evaluation of a child’s expressive phonology includes relational and independent analyses. Both will reveal important information for remediation planning. When identifying candidates for a phonotactic intervention approach, a comparison of the child’s productions to the adult model (i.e., relational analysis) will help us to identify error patterns that occur at the syllable and word level. That said, independent analysis forms the foundation for target selection when developing phonotactic intervention. We must find out what the child is able to do independent of the adult system. By documenting each child’s unique inventory of word shapes and sounds, we will discover what pieces of the adult phonological system are already in place and what remains to be learned. Therefore, we will begin our data collection with an independent analysis.

Step 1: Establish Word Shape and Sound inventories Tables 7–4 and 7–5 provide the results of an independent analysis of a spontaneous speech sample obtained for Esther, age 3;1. The phonotactic inventory form (see Table 7–4) documents the number of syllables, word shapes, and types of syllables produced by Esther. The sound inventory form (see Table 7–5) includes information regarding Esther’s production of vowel sounds, singleton consonant sounds, and clusters. It is vital to include vowels in our analysis, as they are essential to early phonological development. Esther spontaneously produced all vowel sounds with the exception of /U/, /ɔI/, //, /2/, and the rhotic diphthongs. Further probing revealed that she was able to produce /U/ and /ɔI/ in imitation of an adult model. In Table 7–5, the adult form of the words has been intentionally omitted. One of the advantages of independent analysis for highly unintelligible children is that we do not need to know what words they intended to produce. As long as the child’s production represents an attempted word (and is not part of a jargon or babbling sequence), we may include it in our independent analysis of word shapes. Only a few of Esther’s spontaneous word productions were identifiable to the listener. In addition, Esther was unwilling to name pictures or objects despite repeated attempts. Thus, she was an ideal candidate for independent analysis. In this chapter, we refer to all word shapes and sounds produced by the child as “known.” Table 7–6 provides a summary of Esther’s “known” word shapes and sounds.

Table 7–4.  Phonotactic Inventory Form for Esther; Age 3;1

Utterance

Number of Syllables

Word Shape

Vowel as Nucleus

Open versus Closed

Consonant Cluster

[dI]

1

CV

Yes

Open

No

[bɑ]

1

CV

Yes

Open

No

[b]

1

VC

Yes

Closed

No

[maU]

1

CV

Yes

Open

No

[m]

1

VC

Yes

Closed

No

[tu]

1

CV

Yes

Open

No

[di]

1

CV

Yes

Open

No

[æ.d@]

2

VCV

Yes

Open + Open

No

[tI]

1

CV

Yes

Open

No

[dɑ.dɑ]

2

C1V1C1V1

Yes

Open + Open

No

[toU]

1

CV

Yes

Open

No

[.bi]

2

VCV

Yes

Open + Open

No

[doU]

1

CV

Yes

Open

No

[doU]

1

CV

Yes

Open

No

[bu]

1

CV

Yes

Open

No

[dE]

1

CV

Yes

Open

No

[di]

1

CV

Yes

Open

No

[moU]

1

CV

Yes

Open

No

[tI]

1

CV

Yes

Open

No

[b]

1

VC

Yes

Closed

No

[tu]

1

CV

Yes

Open

No

[id]

1

VC

Yes

Closed

No

[I.t@]

2

VCV

Yes

Open + Open

No

[beI.beI]

2

C1V1C1V1

Yes

Open + Open

No

[mɑ.mɑ]

2

C1V1C1V1

Yes

Open + Open

No

[æ.di]

2

VCV

Yes

Open + Open

No

[oUt]

1

VC

Yes

Closed

No continues

197

Table 7–4.  continued

Utterance

Number of Syllables

Word Shape

Vowel as Nucleus

Open versus Closed

Consonant Cluster

[toU]

1

CV

Yes

Open

No

[m]

1

VC

Yes

Closed

No

[E.doU]

2

VCV

Yes

Open

No

[haU]

1

CV

Yes

Open

No

[taU]

1

CV

Yes

Open

No

[æ.tu]

2

VCV

Yes

Open + Open

No

[dæ]

1

CV

Yes

Open

No

[toU]

1

CV

Yes

Open

No

[du]

1

CV

Yes

Open

No

[tæ]

1

CV

Yes

Open

No

[.ti]

2

VCV

Yes

Open + Open

No

[maI]

1

CV

Yes

Open

No

Most frequent type? CV

Yes % = 100

Open% = 87.5%

List: None

1 = 29 2 = 10 3+ = 0 Average = 1.26

198

Also: VC, VCV, C1V1C1V1

Closed% = 12.5%

Table 7–5.  Independent Analysis for Esther, Age 3;1 Early Sounds

I

M

Later Sounds

F

Stops

Fricatives

p

f

b

+++

++

++

v

t

++++ ++++

+++

+

s

d

++++ ++++ +

++++

+

z

k

θ

g

ð

Nasals

ʃ

m

+++

+

++

I

M

F

Liquids

n

ɹ

ŋ

l

Glides

Affricates

w

tʃ

j

dZ

Fricative h

+

Consonant Clusters: Initial Medial Final none noted none noted none noted







Vowels:  Put a + below the vowel sound if present in the inventory: i

I

eI

E



u

U

oU

ɑ/ɔ

++++++

++++

++

++

+++++

+++++

produced after a model

++++++ ++

+++++

/@

/2

aI

aU

ɔI

++++++++

did not produce

+

+++

produced after a model

199

200

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 7–6.  Step 1: Esther’s Known Word Shapes and Sounds Step 1 Known Word Shapes CV VC VCV C1V1C1V1 Known Sounds All vowels (except rhotic) Initial:  /h, b, t, d, m/ Medial:  /b, t, d, m/ Final:  /b, t, d, m/

Step 2:  Selecting Word Shape and Sound Targets Now that we have established a catalog of Esther’s known word shapes and sounds, the next step is to identify potential targets for growth. Velleman (1998, 2016) recommends that we identify the following factors: n Syllable and/or word level constraints that have the greatest negative impact; n Emerging word shapes that may be promoted; n New word shapes that would create the most dramatic change across the system; n Word shapes that would be next developmentally; n Individual communication needs of the child.

Let’s look at each factor relative to Esther’s current phonotactic system. n Syllable and/or word level constraints that have the greatest negative impact:

Esther uses primarily single syllable words and open syllables. She has very limited use of final consonants, is unable to use more than one consonant within the same word, and uses no consonant clusters. n Emerging word shapes that may be promoted:  Esther did not demonstrate

emerging use of any new word shapes. n New word shapes that would create the most dramatic change across the

system:  For Esther, the ability to vary consonants within the same word, in word shapes such as C1V1C2V2 and C1VC2 would create system wide change, as well as the ability to use CVC word shapes, which occur frequently in English.

7  Phonotactic Therapy

201

n Word shapes that would be next developmentally:  According to Table 7–1, the

following word shapes would be next developmentally for Esther: C1V1C1V2, C1V1C2V2, C1VC1, and C1VC2. n Individual communication needs of the child:  Esther has a sister named

Maddie, a dog named Buddy, and a cat named Tabby; therefore, acquisition of the C1V1C2V2 word shape would be highly functional for her and presumably motivating. There are no absolutes in regards to target selection for phonotactic therapy. Reviewing the list of factors presented, we might decide to target C1V1C2V2, C1VC1, and C1VC2. These three word shapes satisfy many of the criteria proposed by Velleman (1998, 2016): they will address significant constraints restricting the use of a variation of consonants within the same word and the use of final consonants; they will promote fast, system-wide change; they are developmentally appropriate; and the word shape C1V1C2V will assist Esther in communicating in her daily life. Why not include C1V1C1V2 (e.g., “baby”)? We could, and if producing words such as “mommy” and “daddy” (rather than “mama” or “dada”) was important to Esther or her family, we would certainly want to consider inclusion of this word shape. However, this structure does not address the variation of consonants or the use of closed syllables (i.e., final consonants) that we identified as priorities. In addition, it is possible that this word shape will emerge on its own as Esther’s phonotactic skills develop. If not, direct remediation would be warranted. Also, what about consonant clusters? English includes many consonant clusters and Esther demonstrates an absolute constraint against them. The ability to produce word shapes with clusters would result in a reduction of homonymy. For these reasons, consonant clusters would be an important consideration as a later target for Esther. We will select sounds as intervention targets as well. Using a developmental approach, potential targets might include /p, w, n/, as these segments are missing from Esther’s inventory and are expected for a child of her chronological age (see Table 7–2). Potential targets for both phonotactic and segmental growth are listed in Table 7–7. You may have noticed that the order has been reversed for listing word shapes and sounds in the second column. The utility of this will be evident as we progress through Step 3.

Step 3:  Selecting Target Words The final phase of preparation involves choosing words that contain our target word shapes and sounds. As described previously, we will use words that represent known word shapes combined with new, target sounds and, alternately, known sounds in new word shapes. In this manner, we utilize the child’s previous knowledge (e.g., known sounds) as the foundation for developing new skills (e.g., target word shapes). In addition, selection of vocabulary words involves consideration of the child’s individual communication needs, as well as age-appropriateness. We want to include words that Esther will understand and have opportunities to use in her daily life. Table 7–8 presents examples of target words that will address Esther’s phonological goals, while increasing her functional communication.

202

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 7–7.  Step 2: Esther’s Target Word Shapes and Sounds Step 1

Step 2

Known Word Shapes

Target Sounds

CV VC VCV C1V1C1V1

/p/ /w/ /n/

Known Sounds All vowels (except rhotic) Initial:  /h, b, t, d, m/ Medial:  /b, t, d, m/ Final:  /b, t, d, m/

Target Word Shapes C1V1C2V2 C1VC1 C1VC2

Table 7–8.  Step 3: Esther’s Target Words Step 1

Step 2

Step 3

Known Word Shapes

Target Sounds

Target Words (Examples)

CV VC VCV C1V1C1V1 Known Sounds All vowels (except rhotic) Initial:  /h, b, t, d, m/ Medial:  /b, t, d, m/ Final:  /b, t, d, m/

/p/ /w/ /n/

paw, up, papa why, we, away no, on, no-no

Target Word Shapes C1V1C2V2 C1VC1 C1VC2

tummy, Maddie, Buddy mom, dad, bib, tot bed, hot, hat, home, boot

Once Esther is able to produce /n/, /p/, and /w/ in simpler word shapes, these sounds may be added to the list of known sounds and used in the target word shapes to build an expanding list of vocabulary (e.g., wet, done, happy).

Establishing Treatment Goals Velleman (2016) reminds us that the goal of phonotactic therapy is the production of new word shapes or other phonotactic forms, not segmental accuracy. Therefore, goals should be written to focus on the target structures regardless of the sounds produced by the child. If expansion of the sound inventory is also included in the intervention

7  Phonotactic Therapy

203

program, then these goals will be written separately from the phonotactic goals. Thus, Esther will have two goal areas, one for phonotactic production and one for sound production, as follows: 1. Phonotactic Production:  C1V1C2V2, C1VC1, C1VC2 2. Sound Production:  [p, w, n] Table 7–9 presents example goals for Esther. Table 7–9 also includes baseline data obtained from our initial independent analyses of Esther’s phonotactic and sound inventories (see Tables 7–4 and 7–5). Because we chose word shapes and sounds not yet produced by Esther, our baseline for her targets is 0% accuracy. Alternatively, we might have decided to develop a baseline probe using the target words that were identified for intervention. For Esther, examples of such words are included in Table 7–9 (e.g., papa, we, no-no, tummy, dad, home). A list such as this could be administered prior to the initiation of intervention and again periodically as a progress monitoring tool. Nontreatment words might be added as well to probe for generalization of skills.

Beginning Therapy Phonotactic therapy includes implicit learning of new structures, as well as explicit learning of word targets that have been carefully chosen to facilitate expansion of word shape and sound inventories (Velleman & Vihman, 2002). Descriptions of this approach include specific suggestions for facilitating production of new word shapes.

Table 7–9.  Example Goals for Esther; Phonotactic Therapy Goal Area #1:  Phonotactic Production Target Word Shapes:  C1V1C2V2, C1VC1, C1VC2 Baseline #1:  Esther produced the C1V1C2V2, C1VC1, and C1VC2 word shapes with 0% accuracy in spontaneous productions, in the context of play. Goal #1:  Esther will produce the C1V1C2V2, C1VC1, and C1VC2 word shapes in target words with 70% accuracy, regardless of the accuracy of sound production, in elicited or spontaneous productions, in the context of play and/or naming pictures. Goal Area #2:  Sound Production Target Sounds:  /p, w, n/ Baseline #2:  Esther produced /p, w, n/ with 0% accuracy, in spontaneous productions, in the context of play. Goal #2:  Esther will produce /p, w, n/ in target words representing known word shapes with 70% accuracy, in elicited or spontaneous productions, in the context of play and/or naming pictures.

204

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

The developmental approach we encountered during target selection is also evident in the recommendations provided by Velleman (2002; 2016) for phonotactic intervention. Many of the ideas summarized in Table 7–10 involve imitating the learning patterns of typically developing children.

Table 7–10.  Phonotactic Therapy Ideas Based on Typical Learning Patterns

Phonotactic Deficit

Because children typically . . .

Ideas for Remediation

Single Syllable Productions only

Learn reduplicated syllables first for two syllable words

Target reduplicated syllables (e.g., “mama” and “bye-bye”)

Initial Consonant Deletion

Produce /h/ early

Target /h/ in CV and CVCV word shapes (e.g., “hi” and “haha”)

Final Consonant Deletion

Prefer voiceless stops, velars, nasals, and fricatives in final position

Target final voiceless stops, velars, nasals or fricatives (e.g., “up,” “pick,” “on,” “bus”)

Final Consonant Deletion

Produce final consonants after lax (short) vowels more often than after tense (long) vowels

Target words with lax (short) vowels (e.g., “big,” “bed,” “foot”)

Final Consonant Deletion

Learn CVC with consonant harmony first

Target words with consonant harmony (e.g., “pop” and “mom”) before words with different consonants

Final Consonant Deletion

Produce final consonants in words that begin with common CV combinations more often than words that begin with rare CV combinations

Target CVC words that begin with common CV combinations, such as /mɑ/ (e.g., “mop” and “mom”)

Reduplication of CVCV words — vowel differentiation absent

Produce /i/ as second vowel in CVCV words

Target CVCV words with /i/ as second vowel — preferably with alveolar sounds paired with the high, front /i/ (e.g., “daddy” and “nanny”)

Reduplication; Consonant Harmony

Prefer front-back order of sounds (if they have a place of order preference)

Target two consonants that move from front to back in terms of place of production, such as bilabial to alveolar (e.g., “bunny” or “mat”)

Cluster Reduction

Learn some final clusters early, such as -nt, -mp, -ŋk

Target final -nt, -mp, -ŋk (e.g., “don’t, jump, thank”)

Source: Compiled from Kehoe & Stoel-Gammon (2001); Velleman (2002, 2016); Zamuner, Gerken, & Hammond (2004).

7  Phonotactic Therapy

205

Other ideas for remediation center on using a child’s current skills to facilitate production of a new skill. Table 7–11 provides a list of these suggestions.

Case Study Initial Data Collection and Selecting Targets Relational analysis, including administration of a standardized test, revealed that Nathan exhibited frequent syllable structure processes and assimilation (i.e., harmony) errors. His most common phonological processes were reduplication, final consonant deletion, cluster reduction, and assimilations (including alveolar, nasal, and bilabial). Because Nathan’s primary difficulties involved errors at the word level, further assessment of phonotactics was conducted through analysis of a spontaneous speech sample. In addition, an independent analysis of his sound inventory was documented.

Step 1: Establish Word Shape and Sound Inventories The following presents results of an independent analysis for Nathan, age 4;2, including a phonotactic inventory (Table 7–12) and a sound inventory (Table 7–13). Table 7–11.  Phonotactic Therapy Ideas Based on Child’s Current Skills Phonological Deficit

If a child is able to . . .

Ideas for Remediation

Initial Consonant Deletion

Produce VC syllables

Target is repeated VC, VC, VC until the C moves from the coda to the onset (e.g., “up, up, up” becomes “pup, pup, pup”

Final Consonant Deletion

Produce CVCV word shapes

Target production of CVC by fading the second vowel (e.g., “bunny” → “bun”)

Final Consonant Deletion

Produce C in onsets

Target production of two word phrases that contain the same consonant in the coda of the first word and onset of the second (e.g., “pick cake”)

Reduplication

Produce C1V1C1V1 word shapes

Target variations in vowels only in C1V1C1V2 (e.g., “baby”), or variations in consonants only in C1V1C2V1 (e.g., “wheatie”) before variations in both in C1V1C2V2 (e.g., “doggy”)

Cluster Reduction

Produce singleton C in onsets and codas

Target phrases with final /s/ followed by an initial consonant that will produce a blend when combined (e.g., “ice tart” → “I start”)

Source:  Compiled from Berrnhardt (1994); Velleman (2002, 2016).

Table 7–12.  Phonotactic Inventory Form for Nathan; Age 4;2

Utterance

Number of Syllables

Word Shape

Vowel as Nucleus

Open versus Closed

Consonant Cluster

[mi]

1

CV

Yes

Open

No

[bni]

2

C1V1C2V1

Yes

Open + Open

No

[p]

1

VC

Yes

Closed

No

[noU]

1

CV

Yes

Open

No

[w.w]

2

C1V1C1V1

Yes

Open + Open

No

[du]

1

CV

Yes

Open

No

[si.si]

2

C1V1C1V1

Yes

Open + Open

No

[jæ.jæ]

2

C1V1C1V1

Yes

Open + Open

No

[paU]

1

CV

Yes

Open

No

[dɑ.dɑ]

2

C1V1C1V1

Yes

Open + Open

No

[ti]

1

CV

Yes

Open

No

[hæ.pi]

2

C1V1C2V2

Yes

Open+ Open

No

[boU]

1

CV

Yes

Open

No

[mɑm]

1

C1V1C1

Yes

Closed

No

[su]

1

CV

Yes

Open

No

[dE.bE]

2

C1V1C2V1

Yes

Open + Open

No

[ti]

1

CV

Yes

Open

No

[maI]

1

CV

Yes

Open

No

[haI]

1

CV

Yes

Open

No

[mɑ.mi]

2

C1V1C1V2

Yes

Open + Open

No

[tu]

1

CV

Yes

Open

No

[In]

1

VC

Yes

Closed

No

[p]

1

CV

Yes

Open

No

[beI.ti]

2

C1V1C2V2

Yes

Open + Open

No

[mɑ.mɑ]

2

C1V1C1V1

Yes

Open + Open

No

206

7  Phonotactic Therapy

207

Table 7–12.  continued

Utterance

Number of Syllables

Word Shape

Vowel as Nucleus

Open versus Closed

Consonant Cluster

[did]

1

C1V1C1

Yes

Closed

No

[oUt]

1

VC

Yes

Closed

No

[woU]

1

CV

Yes

Open

No

[Ip]

1

VC

Yes

Closed

No

[dɔI.dɔI]

2

C1V1C1V1

Yes

Open + Open

No

[haI]

1

CV

Yes

Open

No

[n]

1

CV

Yes

Open

No

[bæb]

1

C1V1C1

Yes

Closed

No

[toU.t@]

2

C1V1C1V2

Yes

Open + Open

No

[moU]

1

CV

Yes

Open

No

[pu]

1

CV

Yes

Open

No

[dæ.dæ]

2

C1V1C1V1

Yes

Open + Open

No

[sUs]

1

C1V1C1

Yes

Closed

No

[nE.n@]

2

C1V1C1V2

Yes

Open + Open

No

[baI.baI]

2

C1V1C1V1

Yes

Open + Open

1 = 25 2 = 15 3+ = 0

Most frequent type?

Yes % = 100

Open% = 85%

Average = 1.375

CV and C1V1C1V1

List: None

Closed% = 15%

By referencing Tables 7–12 and 7–13, we are able to gather Nathan’s word shape and sound inventories. These are summarized in Table 7–14. As shown in Table 7–14, Nathan produces all vowel sounds, with the exception of central vowels with r-coloring and rhotic diphthongs. His consonant inventory includes bilabial and alveolar stops, nasals, glides and the fricatives /h/ and /s/. Nathan’s word shape productions consist primarily of open syllables (85%) and single syllable words. We will now look at potential targets for intervention.

Table 7–13.  Independent Analysis for Nathan, Age 4;2 Early Sounds

I

M

Later Sounds

F

Stops

I

M

F

+++

+

+

Fricatives

p

+++

+

++

f

b

++++ +

++

+

v

t

++++

++

+

s

d

++++ ++

+++

+

z

k

θ

g

ð

Nasals

ʃ

m

++++ ++

++

+

Liquids

n

+++

++

+

ɹ

ŋ

l

Glides

Affricates

w

++

+

tʃ

j

+

+

dZ

Fricative h

+++

Consonant Clusters: Initial Medial Final none noted none noted none noted







Vowels:  Put a + below the vowel sound if present in the inventory: i

I

eI

E



u

U

oU

ɑ/ɔ

++++++ ++++

++

+

+++

++++++

++++

+

++++++

++++++

208

/@

/2

aI

aU

ɔI

++++++ ++

did not produce

+++++

+

++

7  Phonotactic Therapy

209

Table 7–14.  Step 1: Nathan’s Known Word Shapes and Sounds Step 1 Known Word Shapes CV VC VCV C1V1C1V1 C1V1C1V2 C1V1C2V1 C1V1C2V2 C1V1C1 Known Sounds All vowels (except rhotic) Initial:  /p, b, t, d, w, j, m, n, h, s/ Medial:  /p, b, t, d, w, j, m, n, s/ Final:  /p, b, t, d, m, n, s/

Step 2:  Selecting Word Shape and Sound Targets When selecting targets for phonotactic therapy, we want to consider the factors presented previously in light of Nathan’s particular sound system. n Syllable and/or word level constraints that have the greatest negative

impact:  Nathan uses primarily single syllable words and open syllables. He demonstrates limited use of final consonants, infrequently varies the consonant within a word, and uses no consonant clusters. n Emerging word shapes that may be promoted:  Nathan demonstrates

emerging use of final consonants and emerging ability to vary the consonant within a word. n New word shapes that would create the most dramatic change across the

system:  Final consonants and consonant clusters occur frequently in English. Therefore, increasing the use of these phonotactic forms will have a significant impact on Nathan’s speech intelligibility. n Word shapes that would be next developmentally:  According to Table 7–1, the

following word shapes would be next developmentally for Nathan: C1VC2, CCV, and CCVC. n Individual communication needs of the child:  Nathan’s favorite activities

include playing with Thomas the Train, but he has difficulty producing the

210

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

word “train.” He is often misunderstood by peers and adults when talking about or requesting trains, as he produces the word “train” as [teI] or [teIn]. Reviewing these factors, the following word shapes were selected as appropriate targets for Nathan: C1VC2, CCV, and CCVC. Not only are these structures the next developmental step for Nathan, they also meet the other criteria. Nathan already demonstrates emerging ability to use final consonants in C1VC1 words and to vary consonants in C1V1C2V1 and C1V1C2V2 words, so it is reasonable to expect that he may be able to draw upon these skills during production of C1VC2 words. In addition, the ability to use final consonants and clusters will significantly increase his intelligibility. And finally, these new skills may help him to communicate more effectively about trains, his favorite activity. Sound targets were selected for Nathan according to a developmental model, including /k, g, f/. Table 7–15 presents a list of word shape and sound targets for Nathan.

Step 3:  Selecting Target Words Our final step is to create lists of words to be used in therapy that incorporate Nathan’s targets. As illustrated previously, we will use known sounds when targeting new word shapes, and known word shapes when targeting new sounds. Table 7–16 includes examples of words that are appropriate choices for Nathan’s phonotactic intervention program (including, of course, the word “train”!).

Table 7–15.  Step 2: Nathan’s Target Word Shapes and Sounds Step 1

Step 2

Known Word Shapes

Target Sounds

CV VC VCV C1V1C1V1

/k/ /g/ /f/

C1V1C1V2 C1V1C2V1 C1V1C2V2 C1V1C1 Known Sounds All vowels (except rhotic) Initial:  /p, b, t, d, w, j, m, n, h, s/ Medial:  /p, b, t, d, w, j, m, n, s/ Final:  /p, b, t, d, m, n, s/

Target Word Shapes C1VC2 C1C2V C1C2VC

7  Phonotactic Therapy

211

Table 7–16.  Step 3: Nathan’s Target Words Step 1

Step 2

Step 3

Known Word Shapes

Target Sounds

Target Words (Examples)

CV VC VCV C1V1C1V1

/k/ /g/ /f/

key, coco, cake go, egg, Gabby fat, off, taffy

C1V1C1V2 C1V1C2V1 C1V1C2V2 C1V1C1 Known Sounds All vowels (except rhotic) Initial:  /p, b, t, d, w, j, m, n, h, s/ Medial:  /p, b, t, d, w, j, m, n, s/ Final:  /p, b, t, d, m, n, s/

Target Word Shapes C1VC2 C1C2V C1C2VC

hot, wet, bed, sun, moon stay, snow, play [pweI] train [tweIn], brown [bwaUn]

References Bernhardt, B. (1994). Phonological intervention techniques for syllable and word structure development. Clinics in Communication Disorders, 4(1), 54–65. Bleile, K. M. (1995). Manual of articulation and phonological disorders: Infancy through adulthood. Clifton Park, NY: Singular. Chin, S., & Dinnsen, D. (1992). Consonant clusters in disordered speech: Constraints and correspondence patterns. Journal of Child Language, 19, 259–285. Gierut, J. A., Morrisette, M. L., & Champion, A. H. (1999). Lexical constraints in phonological acquisition. Journal of Child Language, 26, 261–294. Grunwell, P. (1985). Phonological Assessment of Child Speech (PACS). Scarborough, Ontario, Canada: Nelson Thomson Learning. Ingram, D., & Ingram, K. (2001). A whole-word approach to phonological analysis and intervention. Language, Speech, and Hearing Services in Schools, 32, 271–283. Kehoe, M. M., & Stoel-Gammon, C. (2001). Development of syllable structure in English-speaking children with particular reference to rhymes. Journal of Child Language, 28(2), 393–432. Lederer, S. H. (2002). Selecting and facilitating the first vocabulary for children with developmental language delays: A focused stimulation approach. Young Exceptional Children, 6(1), 10–17. Leonard, L., Schwartz, R., Folger, M., Newhoff, M., & Wilcox, M. (1979). Children’s imitations of lexical items. Child Development, 50, 19–27. McLeod, S., van Doorn, J., & Reed, V. (2001). Consonant cluster development in two-year-olds: General trends and individual difference. Journal of Speech, Language, and Hearing Research, 44, 1144–1171.

212

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Smit, A. B., Hand, L., Freilinger, J. J., Bernthal, J. E., & Bird, A. (1990). The Iowa articulation norms project and its Nebraska replication. Journal of Speech and Hearing Disorders, 55(4), 779–798. Stoel-Gammon, C. (1987). Phonological skills of 2-year-olds. Language, Speech, and Hearing Services in the Schools, 18, 323–329. Stoel-Gammon, C., & Cooper, J. A. (1984). Patterns of early lexical and phonological development. Journal of Child Language, 11, 247–271. Storkel, H. L., & Morrisette, M. L. (2002). The lexicon and phonology: Interactions in language acquisition. Language, Speech, and Hearing Services in Schools, 33, 24–37. Velleman, S. L. (1998). Making phonology functional: What do I do first? Boston, MA: ButterworthHeinemann. Velleman, S. L. (2002). Phonotactic therapy. In S. Velleman (Ed.), Seminars in Speech and Language, 23, 43–55. Velleman, S. L. (2003). Childhood apraxia of speech resource guide. Cliften Park, NY: Thomson Delmar Learning. Velleman, S. L. (2016). Speech sound disorders. Philadelphia, PA: Wolters Kluwer. Velleman, S. L., & Vihman, M. M. (2002). Whole-word phonology and templates: Trap, bootstrap, or some of each? Language, Speech, and Hearing Services in Schools, 33, 9–23. Watson, M. M., & Scukanec, G. P. (1997). Profiling the phonological abilities of 2-year-olds: A longitudinal investigation. Infant-Toddler Intervention, 7, 67–77. Weiner F. (1981). Treatment of phonological disability using the method of meaningful minimal contrast: Two case studies. Journal of Speech and Hearing Disorders, 46, 97–103. Zamuner, T. S., Gerken, L. A., & Hammond, M. (2004). Phonotactic probabilities in young children’s speech production. Journal of Child Language, 31, 515–536.

8 Core Vocabulary Approach Treatment

Key Features

Core Vocabulary Approach

n A therapy

Age Range Age 3 to11. The most progress is made if the child is approximately 3 years old.

Severity of Phonological Disorder Typically in the severe range. Must demonstrate an inconsistent speech disorder

to decrease inconsistent productions n  Targets wholeword productions not error patterns n Aims for the best consistent production by the child, not necessarily accuracy of the word

Diagnostic Information Needed The child must be

diagnosed with inconsistent speech disorder. Twentyfive words are said three times with an activity between each repetition. If 40% or more demonstrate inconsistency in their productions, this is considered an inconsistent speech disorder.

Basic Structure of Therapy Based on a list of at least 50 words

generated by the child, parents, and teacher that are “meaningful” and “powerful” to the child n Multiple modalities are used to create the best production possibility for the child. n  Best production is noted and used in therapy and homework sessions.

213

214

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

The Core Vocabulary Approach is an intervention developed for children with an inconsistent speech disorder. Inconsistency is characterized by the unpredictable use of a rather large number of different phones with multiple error types (Dodd, Holm, Crosbie, & McIntosh, 2010). These children produce the same words differently not only from one context to the next, but also within the same context. Children with an inconsistent speech sound disorder may indeed produce the same word differently each time they say it. Although some variation in word production is normal, children with an inconsistent speech disorder demonstrate a very high degree of articulatory variation. For example, one child who was a subject in the research produced “parrot” in the following ways: [kætoU@], [kɑdɔw@], and [koUtu@] (McIntosh & Dodd, 2008). When this occurs on a high percentage of words, it has been labeled an inconsistent speech disorder. This approach was originally developed as a response to a child who was in a research project and did not seem to improve with the strategies that were showing improvement in the other children’s speech (Dodd & Iacono, 1989). It was observed that this child had inconsistent productions of the same lexical items. Many words demonstrated various random changes in their production. Because this child was not making progress in therapy, it was hypothesized that he might require a different approach. The child needed to produce the same word in the same way each time it was said. This child received two months of weekly intervention that targeted consistent (although not necessarily accurate) production of highly functional words. After this time, consistency and accuracy increased (e.g., Crosbie, Holm, & Dodd, 2005). It is notable that there are certain similarities between a child with an inconsistent speech disorder and childhood apraxia of speech (Dodd et al., 2010). For example, both groups demonstrate an increase in errors with increased length of utterance. However, the inappropriate oral movements during speech (such as groping and silent posturing) that affect prosody, rate, and fluency are not present in the child with an inconsistent speech disorder. In addition, although those with childhood apraxia of speech show more difficulty with imitation, the child with an inconsistent speech sound disorder is better with imitation than in spontaneous productions. It has been hypothesized that children with apraxia of speech evidence poor sequencing of sounds whereas those with an inconsistent speech disorder make the wrong choice of phonemes. It has been postulated that children with an inconsistent speech disorder have an underlying deficit of generating phonological plans for word production (Dodd, Holm, Crosbie, & McIntosh, 2006).

Research Supporting Speech Sound Disorders and Core Vocabulary Intervention A subset of children diagnosed with inconsistent speech disorder served as the basis for research supporting the core vocabulary intervention. The core vocabulary approach gives the child practice stabilizing productions with a limited number of highly functional core vocabulary words. Table 8–1 is a summary of the research that has noted the effectiveness of this approach. Case studies have not been included in this table.

215

3;5 to 4;3

Three

Dodd & Bradford, 2000

Moderate to severe speech sound disorder as measured by the Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015) and Percentage of Consonants Correct (Shriberg et al., 1997). Two children showed an inconsistent speech disorder, over 40% variability in a word sample.

Approach:  The children were each treated with three different approaches:  PROMPT (Hayden, Eigen, Walker, & Olsen, 2010), core vocabulary, and phonological contrast. Results:  The children with an inconsistent speech disorder benefited most from the core vocabulary approach to establish consistency of production. However, once consistency was established, one of the two children improved with the phonological contrast method.

Approach:  All children received two 8-week interventions: one of core vocabulary and one phonological contrast therapy. Changes in consistency of production and accuracy were used to measure the effect of each intervention. Results:  Core vocabulary therapy resulted in greater change in children with an inconsistent speech disorder, whereas phonological contrast therapy resulted in more change in children with a consistent speech disorder.

4;8 to 6;5

18 children

Crosbie, Holm, & Dodd, 2005

Based on results of the Diagnostic Evaluation of Articulation and Phonology (Dodd, Hua, Crosbie, Holm, & Ozanne, 2006), children were considered severe. Ten children were classified as having an inconsistent speech disorder, whereas the other eight had consistent errors.

Approach:  18 children were treated with a core vocabulary approach, 12 children did not receive treatment. Results:  Children in the treated group made more progress than in the untreated group. However, the children with inconsistent speech disorder made less progress than other subgroups of speech sound disorders.

20 of the 30 children were classified as severe to profound on an intelligibility rating.

30% were 3 to 4, 60% were 4 to 5. The age range spanned from 3 to 11.

320 total, 30 children with an inconsistent speech disorder

Broomfield & Dodd, 2005

Results Approach:  Core vocabulary approach was used to treat all three children. Results:  All three subjects showed gains in intelligibility, accuracy, and consistency of word production. Individual differences required clinical adaptation of the method.

Severity Severe range for the Percentage of Consonants Correct (Shriberg, Austin, Lewis, McSweeney, & Wilson, 1997) All demonstrated an inconsistent speech disorder, over 40% variability in a word sample.

Age Range 3;0 to 4;2

Three

Number of Subjects

McIntosh & Dodd, 2008

Author, Date

Table 8–1.  Examples of Current Research Supporting Core Vocabulary Treatment in Children with Inconsistent Speech Sound Disorders

216

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

In addition, the core vocabulary approach has been used with other populations of children. Herman et al. (2015) evaluated whether core vocabulary intervention improved the accuracy and intelligibility of four deaf children. Children in this study were between the ages of 9 and 11, with profound sensorineural deafness (with either cochlear implants and/or digital hearing aids). Their speech was characterized by inconsistent productions, such as, different error forms for the same lexical item. After eight weeks of therapy, the children’s consistency and accuracy of single word productions improved and generalized to untreated words. Sentence intelligibility ratings also improved, and more words could be identified after therapy. In addition, Baker (2017) presents a pilot study with a child, 3;9, who had auditory brainstem implants at one year old. After five weeks of core vocabulary intervention (two times per week), increases were noted in elicited and spontaneous vocalizations, phonemic inventory size, and syllable shape inventory. To summarize, the core vocabulary intervention model for children with inconsistent speech disorders seems to be supported by research. It presents an option to be used for a relatively short time until stabilization of the word forms occurs. After that, other methods can be used to decrease the noted error patterns.

What Is the Age and Severity of Children Who Could Most Benefit from Core Vocabulary Intervention? It appears that the core vocabulary intervention can be used for a rather wide age range of children. Children in the research studies have varied in age from 3 to 11 years old. However, Broomfield and Dodd (2005) note that children who made the most progress with core vocabulary were around 3 years old. Dodd et al., (2010) state that the core vocabulary approach could be used for any age child who is diagnosed with an inconsistent speech disorder. When therapy is two times per week, stabilization of word forms occurred in approximately eight weeks (e.g., Crosbie et al., 2005).

Initial Data Collection and Selecting Targets The core vocabulary approach is a unique therapy method. It is unique in that its primary goal is to stabilize productions, not necessarily to decrease the error patterns. In addition, it is unique in that it is custom designed for the individual child. Thus, based on the child’s immediate needs and environment, a vocabulary is selected that will increase the child’s functionality. When should you evaluate a child for an inconsistent speech disorder? Further evaluation is necessary first, when there is evidence of many productional variations of the same word, and second, when the child is not responding to traditional or phonological contrast therapy. The Word Inconsistency Assessment of the Diagnostic Evaluation of Articulation and Phonology (DEAP) (Dodd, Hua, Crosbie, Holm, & Ozanne, 2006) is one measure that can be used to evaluate productions that vary. In this assessment, 25 pictures are named three times in one session. Sample words include girl, jump, bridge, kangaroo, umbrella, zebra, and helicopter. Each trial is separated by an activity. If the

8 Core Vocabulary Approach

217

word is produced identically (not necessarily accurately) on all three trials, the child receives a “0.” If the child demonstrates a variable production, this is converted to a percentage. If this word set has more than 40% variability (10 or more of the 25 words), the child is categorized as having an inconsistent speech disorder. However, what do you do if you don’t have this test? Appendix 8–1 contains a short form that is similar to the DEAP. There are 25 words, ranging from one to four syllables. Three trial columns are provided for pronunciation notes. If 10 or more of the words are produced inconsistently on at least one of the three trials, this would qualify the child as demonstrating an inconsistent speech disorder. Although the diagnosis of an inconsistent speech sound disorder is the center of using the core vocabulary approach, it should be remembered that these children should not show evidence of oral motor difficulties, such as those noted in childhood apraxia of speech. The DEAP does have an oral motor screener, and there are several available on the market. Table 8–2 is a short list of tests and protocols that could be used.

Beginning Therapy: Establishing and Working on Core Vocabulary Core vocabulary differs from other interventions as it targets whole-word consistency, not necessarily accuracy. There is a very specific process for selecting words, as well as the way they are taught and practiced. Selecting Words:  The child, parents/caregivers, and the child’s teacher generate a list of 50 to 70 functionally relevant words. Relevance is the key issue here, not the phonemes the words contain. These words should be important and meaningful to the child. So, if possible, the child should be actively involved in selecting this word list. These lists will vary from child to child, but should contain items that are useful in the child’s environment on a daily basis. For example, the words can include names (family members, other important people, pets’ names), places (school, pool, grandma’s house), functional words (please, thank you, good, sorry), and favorite things (football, ice cream, biking). It appears that lexical consistency generalizes after approximately 50 words, so this should be the minimal target. Let’s look at the steps we would go through to establish a core vocabulary for Sandy, who is 3;8 years old. Sandy was referred to the speech/language specialist due to her lack of intelligibility. Although her rate of speech was a little slow, she did not show evidence of any oral-motor difficulties when given a screening test for oral-motor skills. After giving her an articulation test and listening to her conversation, which centered around her summer activities, the clinician noted variable productions of /s/, /z/, /f/, /v/, /l/, /k/, /g/, /ʃ/,/tS/, /dZ/, /θ/, /ð/, and /ɹ/. She scored below the 1st percentile on the Goldman-Fristoe Test of Articulation (Goldman & Fristoe, 2015), and overall intelligibility was rated as poor. Errors seemed to increase with two- or more-syllable words. Sandy was given a consistency probe (See Appendix 8–1), and demonstrated inconsistent productions on 18/25 words, thus, 72% of the time. Due to this high percentage, she could be classified as having an inconsistent speech disorder. Examples of this variable production include “kangaroo” [tænhu], [tɑnu], [n@wu], “thumb” [tm], [bm], [tɑ] and “balloon” [bun], [wun], [@dun]. Sandy loved to talk but was very hard to understand.

Table 8–2.  Examples of Tests for Assessing Nonverbal Oral and Speech Motor Performance in Children

Test, Author, Date

Author, Date

Age Range

Provides Screening Measures

Abilities Assessed

Kaufman Speech Praxis Test for Children

Kaufman, 1995

2;0 to 5;11

No, entire battery takes 5–15 minutes

Items organized from simple to complex motorspeech movements.

Apraxia Profile Preschool and School-Age Versions

Hickman, 1997

3;0 to 13;11

No

Used to document a child’s oral-motor sequencing deficits and establish the level of oral movements and sequences produced successfully.

Verbal Motor Production Assessment for Children

Hayden & Square, 1999

3;0 to 12;0

No

Assesses global motor control, focal oral motor control and sequencing Two supplemental areas include connected speech and language controlspeech characteristics.

Oral Speech Mechanism Screening Examination, 3rd ed.

St. Louis & Ruscello, 2000

5;0 to 78

Yes

Uses an observational sequence for evaluation of a client’s lips, tongue, jaw, teeth, palate, pharynx, velopharyngeal mechanism, breathing, and diadochokinesis.

Screening Test for Developmental Apraxia of Speech, 2nd ed.

Blakeley, 2001

4;0 to 7;11

Yes

Examines expressive language discrepancy, prosody, articulation, and verbal sequencing.

Verbal Dyspraxia Profile

Jelm, 2001

No age range given

No, a checklist

Checklist to assist with the diagnosis of children who may demonstrate verbal dyspraxia Includes automatic and imitative oral-motor movement checklists for jaw, lips/ cheeks, and tongue.

Diagnostic Evaluation of Articulation and Phonology (DEAP). Oral motor screen.

Dodd, Hua, Crosbie, Holm, & Ozanne, 2006

3;0 to 8;11

Yes, it is a screening measure.

Provides scores for diadochokinetic, isolated, and sequenced movements.

218

8 Core Vocabulary Approach

219

Table 8–2.  continued

Test, Author, Date

Author, Date

Age Range

Dynamic Evaluation of Motor Speech Skill (DEMSS), Strand, McCauley, Weigand, Stoeckel, & Baas (2013)

Strand, McCauley, Weigand, Stoeckel, & Baas, reliability data published in 2013

3;0 to 6;7

Provides Screening Measures No

Abilities Assessed Assesses, based on imitation, the word and vowel articulation accuracy, prosody, and consistency of utterance.

She helped formulate the word list by pointing to items, holding up her doll, for example, and enlisting the aid of her parents to help figure out the words she wanted. Sandy’s preschool teacher also helped by noting specific items that Sandy routinely asked for or were needed to help her function better with the other children. The core vocabulary items are contained in Table 8–3. Before Therapy Begins:  Once you have selected the words, you and the child’s family and teacher will need to find pictures of the selected core vocabulary words. The name is written underneath the picture. Holm, Crosbie, and Dodd (2013) suggest some type of storage bag or box, in which the picture cards are stored and can be picked out for the individual therapy sessions. The number of words practiced in a session can vary, but no more than 10 words is suggested (Holm et al., 2013). Dodd et al., (2010) recommend that core vocabulary intervention be two times a week for 30-minute individual sessions. There should be home practice between sessions. The total time for intervention is usually 6 to 8 weeks. During the first session of the week the child’s best production of a subset of core vocabulary words should be established. This is done by randomly choosing words (for example 10 cards) from the larger set of core vocabulary word cards. Teaching “Best Production”:  First, pictures and letters are used for all targets. Second, it must be emphasized to the child that the primary goal is that she/he produces the word exactly the same way each time. It does not need to be an error-free production, but it must be the same production. Third, production is drilled sound-by-sound as the child breaks the word into syllables and then reassembles them. Fourth, after reassembly, the word is practiced five times before moving on to the next target word. This method is hypothesized to teach phonological assembly that is problematic for these children. If necessary, a number of production cues could be used including: an auditory model for immediate and delayed imitation, visual-phonetic cues, verbal-phonetic cues, manual tactile feedback, prosodic cueing, successive approximation or shaping, orthographic cues, metaphor use, the use of metaphonological information about the specific sound class or sound, and phonological awareness instruction (Holm, Crosbie, & Dodd, 2013).

220

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 8–3.  Core Vocabulary Items for Sandy, 3;8 Years Old Rosie (dog’s name)

hungry

pajamas

Chuck (brother’s name)

chocolate

phone

Sidney (best friend’s name)

strawberry

teddy

school

potty

go

swimming

give

house

apple

skirt

more

banana

sweater

car

cereal

nose

color

ice cream

mouth

crayons

please

hand

water

thank you

upstairs

coat

help

candy

cold

shoes

soda

bye-bye

princess (Sandy’s doll)

ouch

pizza

cookie

game

burger

that

play

French fries

camping

open

bathtub

bike

more

comb

outside

glass

brush

juice

bowl

teeth

milk

want

baby

park

toast

brother

thirsty

sandwich

eat

bed

Let’s use an example. One of Sandy’s core vocabulary words was “ice cream.” We could start out with giving a model (auditory modal for immediate imitation), and then noting that “ice cream” has two syllables, or beats (phonological awareness information about syllableness). We could even count out the number of sounds in the first word “ice” and the second one “cream” as we separate “ice cream” into its individual sound components. Visual-phonetic and verbal-phonetic cues could be used as we say “ice,” noting the positioning of the mouth and lips for “i” and then the movement to the snake sound (metaphor). We have a picture of “ice cream” with the words written underneath the picture. Our finger could follow the word “ice” as we say it. The snake sound at the end is a long sound (metaphonological information) and we could perhaps do some direct tactile feedback showing how the lips are spread for the “ce.” Again, recall that the primary goal here is consistency not accuracy. Therefore, the child’s best production

8 Core Vocabulary Approach

221

might be an acceptable developmental error. For example, if we can get Sandy to say [aIθ] for [aIs] every single time, then we have won the battle. Dodd et al. (2010) note that the dialogue during this feedback attempt should focus on phonological awareness, the child needs to learn how to phonologically plan words. During the treatment sessions, feedback should be provided when there is an error production. It is not always evident to the child what is required, or why they are in therapy. From clinical experience, it was interesting to note what it is that children perceived as their reason or goals for therapy. Very often even 3rd- or 4th-grade children who had been in therapy for years could not pinpoint what they were working on. Yes, speech, but that was very often the extent of their understanding. And consequently, it is important that the child has explicit knowledge about the nature of the error and how it might be corrected. Consistency is the ultimate goal, not accuracy. However, often, with some direct instruction, accuracy can be achieved. Holm et al. (2013) state that clinicians should avoid just asking the child to imitate the target word. For the child with an inconsistent speech sound disorder, imitation already provides the phonological plan for the word. The child needs to be able to generate their own plan of the word. The clinician’s goal is to provide information about the plan. The first session of each week focuses on the best production of approximately 10 words. The second session of the week focuses on massed practice. During this time, the clinician can still use various cueing techniques, but imitation should be kept to a minimum. Children with inconsistent speech disorders are better at imitating than spontaneous productions. The child needs to learn to plan the sequence of phonological units in the word, not just imitate them. Dodd et al. (2010) state that the amount of practice during fun activities or games should be high, the child should attempt between 150 to 170 responses in 30 minutes. During the end of the second session, the 10 practice words are tested for consistency across three repetitions. Any words that are consistent (not necessarily accurate) are removed from the list. At the end of each week, any inconsistent words are retained for further therapy. In between each session, the parents/caregivers, family members, and teachers are involved in daily practice. These helpers should focus on the consistency of the child’s productions. It must be emphasized to all individuals who are working with the child that the goal of intervention is that the child says the word in exactly the same way each time she or he attempts to say it. At this point, it is not necessarily a production without errors; it is a production without variability.

Monitoring Progress and Generalization A generalization probe is used every two weeks to monitor consistency. This probe should be a 10-item list of words that are not being used in treatment. Once the untreated words become consistent, Dodd et al. (2010) recommend using the Inconsistency Subtest of the Diagnostic Evaluation of Articulation and Phonology (Dodd et al., 2006) or a similar measure to reconfirm consistency. If generalization has occurred (less than 40% variation in production), then it appears that core vocabulary as an intervention method can be terminated. At this point, it is reported that children’s speech becomes more

222

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

consistent and accurate, and that their speech is now characterized by developmental, and not atypical, error patterns (Holm et al., 2005). A case study of an inconsistent speech sound disorder noted that after establishing a consistent vocabulary of 70 words, consistency generalized. At this time, a phonological contrast approach was reintroduced and deemed to be successful (Dodd & Iacono, 1989). The core vocabulary list for Sandy is in Table 8–3. From that list, 10 words were selected: please, swimming, want, outside, Chuck, hungry, school, apple, brother, and eat. Let’s use one of the words “brother” to see how we could achieve the best production. n “Sandy, listen to the word ‘brother.’” The clinician says the word slowly but not

so slowly that the prosodic unit is destroyed. However, a pause would occur between “bro” and “ther.” (Auditory model, prosodic cueing) n The word “brother” has two beats “bro” and “ther.” Tap out the beats on

the table or put chips in a box for each syllable, (Phonological awareness instruction). n First syllable:  Let’s look at the first part “bro.” “Do you know how many sounds

it has? It has three sounds ‘b,’ ‘r,’ ‘o.’” These are said slowly and the written letters of the word are traced as the sounds are said, (Phonological awareness instruction, orthographic cues). n The first part begins with “b.” “Watch me as I say that sound. My lips come

together. The second sound is ‘r.’ That is our racecar sound, ‘rrrrr.’ The last sound is ‘uh,’ like when we sigh, ‘uh.’” (Visual-phonetic and verbal-phonetic cueing, using metaphors). If Sandy says [bw] for [bɹ], the clinician could point out the lip rounding on [w] versus no lip rounding on [ɹ]. The clinician could push the corners of Sandy’s mouth out while she tries to say [bɹ] (Manual tactile feedback). Try to obtain a sound without lip rounding, not a perfect [ɹ] (Successive approximations). n Putting together the three sounds: Use your finger to trace the letters as you

combine [b] + [ɹ] + [] (Orthographic cue). n Second syllable:  “Now let’s look and listen to the second syllable ‘ther.’ How

many sounds do you hear?” The clinician says the second syllable slowly. This might be difficult, as “ther” could be perceived as three sounds rather than two [ð] + [2]. It could be perceived as [ð], a type of [@] and then an [ɹ]. If this occurs, put a pause between [ð] and [2], (Phonological awareness instruction, prosodic cueing). n Point out the voicing and the slight protrusion of the tongue between the teeth

for [ð] (Visual-phonetic and verbal-phonetic cues). n Go back to the race car sound “rrrr” (Metaphor). If Sandy says the final [2] as

[@], then the racecar sound will need to be re-emphasized. Attempt to obtain some sound other than [@] (Successive approximation).

8 Core Vocabulary Approach

223

n Bringing the two syllables together:  “Let’s put the first part and the second part

together ‘bro-ther.’” Put your left hand down on the table to emphasize the first syllable, then a pause, then the right hand down for the second syllable. The hands could at first be fairly far apart. As Sandy repeats the word, first with a longer pause, the hands are moved together to slowly take out the pause, (Prosodic cueing). n After going through these steps and having Sandy repeat the word five times,

the best production was [bwð@]. This was considered an acceptable response, the two “r” sounds were developmental errors. This was considered Sandy's best production and was used for “brother” in treatment. The clinician would then go through the other words using similar techniques. The goal for this session is to obtain a best production of all the words. The parents/caregiver and teacher would all need to be informed about this production to be able to practice it between sessions. The second practice session would first review best productions, and then with motivating games and activities reinforce the utterances, first as isolated words, then within a carrier phrase, and finally in a spontaneous activity. For example, with “brother,” Sandy could tell a story about “Brother Bo” (a made-up character) and have him do all sorts of antics. Pictures could be provided to help with the story.

Group Therapy The core vocabulary approach is unique, due in part, to the fact that it is tailored to each individual child. Therefore, according to the authors, this technique should not be implemented in a group. (Dodd et al., 2010; Holm et al., 2013).

Establishing Treatment Goals The goal of the core vocabulary approach is to increase intelligibility (Dodd et al., 2010). Therefore, an intelligibility goal based on a baseline measure would be a good choice. Although decreasing variability is central to the core vocabulary approach, due to the time frame of this therapy (approximately 8 weeks) a goal of decreasing variability would not be a good goal for an entire year. By decreasing variability, and increasing intelligibility, specific error patterns appear to decrease noticeably. However, the clinician cannot be certain at the onset which error patterns might be affected. Therefore, a goal that is centered on specific sounds or patterns might not be successful because there is no way to predict in what ways the child will respond to the decrease in variability. Therefore, the following goals are suggested to address intelligibility. The clinician might want to use one or the other, but probably not both. Options, baseline measurements, and goals are listed for each in Table 8–4.

224

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 8–4.  Goal Possibilities for Sandy, 3;4: Core Vocabulary Approach Option #1:  Write the intervention goal to target intelligibility. Goal Area #1:  Intelligibility Baseline #1:  Prior to treatment, a speech sample was collected from Sandy. Her intelligibility was calculated by dividing the number of intelligible words (those that were easily understood, even if not all sounds were accurate such as “tun” for “sun”) divided by the number of unintelligible words. At baseline, Sandy’s overall intelligibility was 30%. (See Appendix 2–6, page 64 for a matrix.) Goal #1:  Sandy will increase her overall intelligibility to 75% in conversation as measured by speech samples collected three times per year. Option #2:  Write the intervention goal to target overall intelligibility using the Percentage of Consonants Correct or Proportion of Whole Word Correctness (PWWC). See Chapter 2, pages 49–50 for details. Goal Area #2:  Intelligibility using PCC or PWWC Baseline #2a:  Prior to treatment, Sandy’s Percentage of Consonants Correct (PCC) for singleton consonants was 35%. For example, she said “dat” for “that,” which is 1 out of 2 consonants correct. Goal #2a:  Sandy will increase her overall accuracy to 70% in conversation as measured by Percentage of Consonants Correct (PCC) on independent probes administered three times per year. Baseline #2b:  Prior to treatment, Sandy’s Proportion of Whole Word Correctness (PWWC) for singleton consonants was 30%. For example, she said “dat” for “that,” which was not counted as correct, whereas “two” was counted as correct as all vowels and consonants were produced accurately. Goal #2b:  Sandy will increase her overall accuracy to 70% in conversation as measured by Proportion of Whole Word Correctness (PWWC), in which 70% of the overall words are produced correctly. This will be calculated from independent probes administered three times per year.

Case Study Josh was referred to the speech/language specialist due to his unintelligibility. He was 3;10 and his parents had at first thought that he might grow out of his speech difficulties. Based on recommendations by his pediatrician and day care teacher, Josh was scheduled for a speech/language evaluation. Josh had normal hearing and receptive language appeared to be in the normal range. The clinician had difficulties scoring an expressive test due to his error patterns. The clinician noted that when she asked him to repeat a word, Josh’s productions changed. Therefore, he produced multiple errors for the same word. The clinician administered a probe, and in 15/25, (60%) of the utterances, Josh varied his productions. Results from an oral-motor screener showed no difficulties with these skills. Josh was diagnosed as having an inconsistent speech disorder.

8 Core Vocabulary Approach

225

Examples of the variability included the following: giraffe

[d@bæt] [wEwæb] [d@wæp]

elephant [Ew@fIn] [Iw@p@n] [wEp@n] snake [naIt] [neIp] [deIt] crayon [t@wɑ] [teIw@n] [teIj@n] alligator [weIt@] [ɑw@teIt@] [ɑteIt@] clown [taUn] [waUn] [t@waUn] The following sounds appeared at least once in Josh’s inventory: [p, b, t, d, m, n, ŋ, w, j, v, and f ]. Josh does not have the following sounds in his inventory: [k, g, s, z, θ, ð, ɹ, l, ʃ, tʃ , dZ]. Josh uses several different substitutions for these sounds. For example, although he can produce [ f ], Josh will substitute [b] and [p] for [ f ], as well. Another example is [ j  ], correct in “crayon” one time and then a [w] for [ j  ] substitution could be noted in other productions of the same word. The core vocabulary approach was selected for Josh. Input for the core vocabulary was received from Josh, the parents, and the day care teacher. Special emphasis was placed on interaction with the children in the day care center. The other children have difficulty understanding Josh, and he often gets frustrated trying to communicate. Table 8–5 is a list of the words that were selected. Appendix 8–2 is an example of the first five words with picture cards. Let’s go through one of the five words to see how one could possibly proceed. Word: truck n Phonological awareness instruction: Includes how many syllables or beats a

word has, how many sounds are in the word “truck,” which sound(s) are at the beginning of the word, which ones at the end. These skills should be practiced with very young children, as well. n Auditory model for immediate and delayed imitation:  At first, the clinician is

using imitation quite consistently. This will be faded as soon as possible. n Visual-phonetic cues and verbal-phonetic cues:  The clinician shows and tells

the child where the tongue is, for example with [t] transitioning into [ɹ]. The clinician might draw attention to the movement from the alveolar ridge to probably a retroflexed [ɹ]. The tongue has to glide back from the alveolar ridge when voicing is initiated. n Manual tactile feedback:  This could provide an assist for the previous point.

The clinician shows the child the wide spread lips (not lip rounding) and could use the thumb and index finger to gently hold the edges of the mouth back. n Prosodic cueing:  Modeling could be done in a slower rate of speech. Do not

stretch it out so much that it becomes completely distorted.

226

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Table 8–5.  Core Vocabulary for Josh, age 3;10 blocks

Kate (sister’s name)

little

game

John (brother’s name)

big

snack

grandma

lots

football

grandpa

bubbles

truck

Auntie Lou

shirt

car

Uncle Joe

socks

crayons

hot dog

shoes

please

burger

yucky

thank you

fries

scary

pencil

bike

sleep

painting

scooter

watch

help

swimming

look

lunch

more

C’mon

sandwich

cereal

gimme

juice

story

brush

milk

all gone

bath

puzzle

open

cookie

potty

close

cracker

water

hurt

ice cream

book

all done

play

outside

throw

sorry

chips

here

blanket

apple

off

throw

banana

n Successive approximation or shaping:  The clinician is doing this to try and get

the “best” production. For example, with the [ɹ] in “truck,” the clinician might first try to get lip spreading and not lip rounding (if it is a [w] production). Second, Josh could be instructed to start at the “bumpy spot” right behind the front teeth. Third, glide the tongue back close to the palate. The clinician is hoping that each of these changes will produce an r-type sound that is closer to an accurate production. n Orthographic cues:  The clinician draws attention to the “t” + “r” on the written

form under the picture. One could also just write out the two letters, “t” and “r” and see if Josh can transition from one sound to the next when the clinician traces under the individual sounds. n Using metaphors:  Give the sound a name, such as the “s” sound is sometimes

called the snake sound. As mentioned earlier, the “r” could be the race car sound.

8 Core Vocabulary Approach

227

n Metaphonological information about the specific sound class or sound:  For this

area you are asking the child to be consciously aware of sounds. For example, [t] is a short sound, [ɹ] is a long sound. Phonological awareness instruction also targets metaphonology. After going through these steps and repeating the word, the best production noted was [twk], although the lip rounding on [w] was noticeably reduced. The goal was consistent production of this word in this exact manner in later therapy sessions.

References Baker, C. W. (2017). The effect of core vocabulary therapy on speech outcomes for a child with an auditory brainstem implant: A pilot study (Master’s thesis). Retrieved from https://repositories.lib​ .utexas.edu/bitstream/handle/2152/60437/BAKER-THESIS-2017.pdf?sequence=1 Blakeley, R. W. (2001). Screening Test for Developmental Apraxia of Speech (2nd ed.). Austin, TX: Pro-Ed. Broomfield, J., & Dodd, B. (2005). Epidemiology of speech disorders. In B. Dodd (Ed.), Differential diagnosis and treatment of children with speech disorder (2nd ed., pp. 83–99). London, UK: Whurr. Crosbie, S., Holm, A., & Dodd, B. (2005). Intervention for children with severe speech disorder: A comparison of two approaches. International Journal of Language and Communication Disorders, 40, 467–491. Dodd, B., & Bradford, A. (2000). A comparison of three different therapy methods for children with different types of developmental phonological disorder. International Journal of Language and Communication Disorders, 35, 189–209. Dodd, B., Holm, A., Crosbie, S., & McIntosh, B. (2006). A core vocabulary approach for management of inconsistent speech disorder. Advances in Speech-Language Pathology, 8, 220–230. Dodd, B., Holm, A., Crosbie, S., & McIntosh, B. (2010). Core vocabulary intervention. In A. L. Williams, S. McLeod, & R. J. McCauley (Eds.), Interventions for speech sound disorders (pp. 117– 136). Baltimore, MD: Brookes. Dodd, B., Hua, Z., Crosbie, S., Holm, A., & Ozanne, A. (2006). Diagnostic Evaluation of Articulation and Phonology (DEAP). San Antonio, TX: Psych Corp of Harcourt Assessment. Dodd, B., & Iacono, T. (1989). Phonological disorders in children: Changes in phonological process use during treatment. British Journal of Communication Disorders, 24, 333–351. Goldman, R., & Fristoe, M. (2015). Goldman-Fristoe Test of Articulation (3rd ed.). Boston, MA: NCIS Pearson. Hayden, D. A., Eigen, J., Walker, A., & Olsen, L. (2010). PROMPT: A tactually grounded model. In A. L. Williams, S. McLeod, & R. J. McCauley (Eds.), Interventions for speech sound disorders in children (pp. 453–474). Baltimore, MD: Brookes Hayden, D., & Square, P. (1999). Verbal motor production assessment for children. San Antonio, TX: Psychological Corporation. Herman, R., Ford, K., Thomas, J., Oyebade, N., Bennett, D., & Dodd, B. (2015). Evaluation of core vocabulary therapy for deaf children: Four treatment case studies. Child Language Teaching & Therapy, 31, 221–235. Hickman, L. (1997). Apraxia Profile. San Antonio, TX: The Psychological Corporation. Holm, A., Crosbie, S., & Dodd, B. (2013). Treating inconsistent speech disorders. In B. Dodd (Ed.), Differential diagnosis and treatment of children with speech disorder (3rd ed., pp. 182–201). Hoboken, NJ: John Wiley and Sons.

228

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Jelm, J. M. (2001). Verbal Dyspraxia Profile. DeKalb, IL: Janelle. Kaufman, N. (1995). Kaufman Speech Praxis Test for Children. Detroit, MI: Wayne State University Press. McIntosh, B., & Dodd, B. (2008). Evaluation of core vocabulary intervention for treatment of inconsistent phonological disorder: Three treatment case studies. Child Language and Teaching Therapy, 25, 9–30. St. Louis K. O., & Ruscello D. (2000). Oral Speech Mechanism Screening Examination (3rd ed.) Austin, TX: Pro-Ed. Shriberg, L. D., Austin, D., Lewis, B. A., McSweeney, J. L., & Wilson, D. L. (1997). The percentage of consonants correct (PCC) metric: Extensions and reliability data. Journal of Speech, Language, and Hearing Research, 40, 708–722. Strand, E. A., McCauley, R. J., Weigand, S. D., Stoeckel, R. E., & Baas, B. S. (2013). A motor speech assessment for children with severe speech disorders: Reliability and validity evidence. Journal of Speech, Language, and Hearing Research, 56, 505–520.

8 Core Vocabulary Approach

229

Appendix 8–1 Probe Words for Consistency Measure Word

Trial #1

Trial #2

Trial #3

1. kangaroo 2. thumb 3. alligator 4. slide 5. zebra 6. clown 7. umbrella 8. snake 9. crayon 10. zipper 11. bicycle 12. spider 13. spaghetti 14. shovel 15. princess 16. chair 17. helicopter 18. toothbrush 19. fish 20. balloon 21. giraffe 22. butterfly 23. dinosaur 24. elephant 25. strawberry Have the child say each of the words three times during the session with an activity between each sequence of productions. Monitor any vowel and/or consonant changes. If 10 or more of the words change during repetitions (40%), this would qualify a child for an inconsistent speech sound disorder.

230

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

Appendix 8–2 Cards for Core Vocabulary

blocks

snack

football

game

truck

9 Cycles Phonological Remediation Approach Treatment Cycles Phonological Remediation Approach

Age Range 2;9 to 7 years of age

Severity of Phonological Disorder Moderate to severe (children

are typically unintelligible)

Key Features n Cycles are established that target a pattern for a specific time interval. n Primary, secondary, and advanced targets can be selected. n Words chosen must be

stimulable so that the child can initially achieve a high level of success.

Diagnostic Information Needed Phonological process analysis: processes above 40% deviancy are used in pattern selection.

Basic Structure of Therapy n “Primary patterns” are used to establish cycles initially. n  Therapy has a specific structure including auditory enhancement, drill-practice, metaphonological tasks, and homework.

231

232

Phonological Treatment of Speech Sound Disorders in Children: A Practical Guide

The cycles phonological remediation approach is used widely as a target selection and treatment method for children with severe-to-profound expressive speech sound disorders. It was developed and refined to facilitate the acquisition of intelligible speech patterns in “cycles.” A cycle is a period of time in which specific phonological patterns are treated. The cycles phonological remediation approach has a detailed way to select targets, structure therapy, and move from cycle to cycle of deviant phonological patterns. Unlike the motor approach, cycles training does not train for mastery of a sound. Rather it structures target words in such a way that 100% accuracy is obtained in production-practice drills. Thus, it is critical that the child be “stimulable” and capable of producing the pattern with “assists” such as tactile cues or amplification, for example. The cycles approach was developed based on the following principles (Hodson, 2007, 2011; Hodson & Paden, 1991): 1. Phonological acquisition is a gradual process. Thus, the cycles approach enhances the development of appropriate phonological patterns in a sequential manner, recycling some patterns until they begin to carry over to conversational speech. 2. Children with normal hearing acquire the sound system primarily by listening. Therefore, the cycles approach does not concentrate on tongue placement (such as the motor approach), but rather provides other cues that can be effectively used by the child. 3. Children use kinesthetic and auditory skills as they acquire new patterns that aid self-monitoring skills. Thus, the cycles approach uses slight amplification during a listening activity at the beginning and end of each session to improve perceptual abilities. 4. The phonetic environment can help or hinder the correct production of a new pattern. Thus, care is taken when selecting target words. For example, the target and error sounds should not appear together in the target words until a later cycle. If a child uses a /d/ for /g/ substitution, words such as “dog” or “dig” should not be used as targets. Depending on the child’s possibilities, better words might be “dime” or “D.” 5. Children tend to generalize new speech skills to other targets. Thus specific patterns are targeted to facilitate generalization and increase intelligibility. 6. Matching the child’s ability to specific therapy goals can aid a child’s learning process. Thus, if the skill level of a child is evaluated and treatment begins one level above that, the child will be optimally challenged and successful from the beginning of treatment. 7. Children should be energetically involved in their phonological acquisition, they need to be actively engaged rather than passive imitators.

9 Cycles Phonological Remediation Approach

233

The cycles approach is an attempt to closely approximate the way phonological development normally occurs — as a gradual process. Targeted patterns in the cycle are used to stimulate the emergence of a pattern, not specifically the mastery of it (Hodson & Padden, 1991). The following sections of this chapter will outline the research that supports this approach, summarize how to collect data, select target patterns, and structure therapy.

Research Supporting the Cycles Approach Although the cycles approach has been widely published and utilized since the mid-70s, there is not a lot of research data that use controlled studies to document treatment efficacy. The research that is available often uses a modified cycles approach, and that can mean several different things. For example, one study (Conture, Louko, & Edwards, 1993) used a “modified cycles” to mean that an entire process, such as final consonant deletion, was targeted and not individual patterns or specific words. The steps that Hodson and Paden outlined (for example, Hodson & Paden, 1991) do not seem to be followed in these modified approaches. However, the cycles approach has decades of documented clinical experience treating children with unintelligible speech. In addition, children with a hearing loss (Gordon-Brannan, Hodson, & Wynne, 1992), and cleft palate (Hodson, Chin, Redmond, & Simpson, 1983), have also shown improvement with this type of therapy. Table 9–1 summarizes the major controlled studies involving the cycles approach. Case studies and those studies in which the subjects are non-native English speakers are not included. There are also several studies (for example, Ceron, Keske-Soares, Freitas, & Gubiano, 2010; Keske-Soares, Brancalioni, Marini, Pagliarin, & Ceron, 2008; Mota, Keske-Soares, Bagetti, Ceron, & Melo Filha, 2010) that document the effectiveness of the cycles approach when treating Portuguese children with speech sound disorders.

What Is the Age and Severity of Children Who Could Most Benefit from the Cycles Approach? As stated by the authors (Hodson & Paden, 1991), this approach is for highly unintelligible children. As such, the children in the studies all demonstrated a moderate-to severe-phonological disorder. The varied age range, from 2;9 to 7 years of age, makes the cycles approach a possibility for even very young children. The concept that sets this approach apart from others is that it appears children make progress in a relatively short time (e.g., Tyler, Edwards, & Saxman, 1987). This gets the children back into the classroom faster and thus, is more effective than spending a much longer period of time moving from one sound to the next, as is the case in the motor approach. As we will see in the following sections, Hodson and Paden (1991) indicate that the child should be stimulable for the sound or pattern that is targeted. This is different than, for example, the complexity approaches and maximal oppositions that typically target phonemes that are not stimulable.

Table 9–1.  Examples of Research Supporting the Cycles Phonological Remediation Approach Author, Date Rudolph & Wendt, 2014

Harbers, Paden, & Halle, 1999

Rvachew, Rafaat, & Martin, 1999

Number of Subjects 3 children

4 children

10 children in Study 1, 13 in Study 2

Age Range 4;3 to 5;3

3;5 to 4;2

Mean age 4;6 to 4;7

Severity

Results

Moderate to severe speech sound disorders

Approach:  Three phonological patterns were targeted for two cycles for each child.

Severe range based on the Hodson Assessment of Phonological PatternsRevised (Hodson, 2004)

Approach:  Both the cycles approach and phonological awareness skills were trained.

GoldmanFristoe scores ranged from