Lisa Scott Trautman is an assistant professor in the Department of Communicative Disorders and Sciences at Wichita State University. In addition to her position at WSU, she works as a professional consultant to the Stuttering Foundation of America. Her clinical background includes co-directing the fluency services at the WSU Speech and Hearing Clinic with colleague Brian Ray, five years experience as a school clinician in Nebraska serving children birth to 21, and four years on faculty in the UNL Speech and Hearing Clinic prior to joining the department at WSU. An ASHA Recognized Fluency Specialist, she has evaluated, treated, and/or consulted with over 150 school-age children who stutter, and many adults and preschool children as well. She has made presentations at district, state, and national meetings, and co-authored several publications in the area of fluency disorders.
James Panico is a first-year doctoral student at the University of Nebraska-Lincoln. James has a B.A. in Communication Sciences and Disorders from the University of Florida and an M.S. degree in Speech-Language Pathology from Florida State University. James is focusing his Ph.D. program in the area of fluency disorders.
A basic component of most stuttering therapy programs is manipulation of the length and grammatical complexity of client utterances (Healey, Norris, Scott Trautman, & Susca, 1999). For example, Ryan and Ryan (1995) reported that a fluency-shaping treatment program based on a gradual increase in utterance length and grammatical complexity was effective in establishing fluency in school-age children who stutter. Thus, manipulating the length and complexity of client utterances produced in therapy assists him/her in achieving fluent productions.
In a traditional length/complexity hierarchy, utterance length relates to the number of words or syllables produced per speaking turn. Utterance complexity, on the other hand, is associated with the syntactic difficulty of what's said. Both length and complexity can be independently manipulated to facilitate a more fluent response.
In this paper we will describe a modified approach clinicians can use when manipulating linguistic complexity in stuttering therapy. The approach we recommend is substantially different from the typical method focusing only on length and complexity through use of isolated word lists, phrases, or sentences. Instead, our approach extends these methods by changing the cognitive-linguistic and discourse demands placed on the child as well as other features of linguistic complexity that have not been addressed in previous treatment paradigms. Using Norris and Hoffman's (1993) Situational-Discourse-Semantic (SDS) model of communication, linguistic complexity is manipulated within the context of topic-centered, thematic activities. The SDS model provides a framework for structuring and systematically manipulating cognitive-linguistic, discourse, and individual utterance length/complexity demands. The use of a consistent topic or theme provides the framework for creating meaningful communicative interactions rather than relying on arbitrary linguistic tasks such as word or sentence lists.
The SDS Model
The SDS model describes a dynamic interaction of three linguistic domains inherent in any communication event: 1) Situational (i.e., cognitive-linguistic demand), 2) Discourse (i.e., structure and format of the overall idea being communicated), and 3) Semantic (i.e., meaning, length, and complexity of individual utterances). Each domain addresses its particular aspect of linguistic performance in a hierarchy of 10 levels, from very simple at the bottom of the model to very complex at the top of the model. (At this point, it might be helpful to the reader to print the SDS model. You must have Adobe's Free Acrobat Reader to access the handout. You can download it at www.adobe.com/products/acrobat/readstep.html ).
The Situational domain illustrates a continuum of cognitive-linguistic demands through hierarchical changes in contextualization. Speech is contextualized when it concerns topics in the here-and-now, supported by real objects, toys, pictures, or print. For example, telling a story while looking at pictures is one way to contextualize therapy activities. When a person talks about an experience that is displaced in time or space and is unsupported by objects, symbolic representations, pictures, or print, the topic is decontextualized. An example might be telling a story about a recent vacation. If there are no pictures, print, or object to assist in structuring the story and the child is relying on language alone to recreate the event for the listener, the story is decontextualized. In the SDS model, there are 5 contextualized and 5 decontextualized levels within the Situational context.
The Discourse domain illustrates a continuum of 10 levels of discourse organization. When a speaker connects several utterances together about a topic, he or she must rely on some form of discourse organization to structure ideas. Referring to the lower end of the model, primitive discourse structures typically observed in very young children are described. As linguistic development occurs, the child begins to use more sophisticated discourse structures such as ordered or reactive sequences, with typically developing children expected to tell stories using complete episodic structure by age 8. As discourse development continues, complex discourse structures at the upper end of the model are noted (e.g., compound, complex, and interactive episodes).
The Semantic domain also is also organized on a hierarchical continuum of 10 levels of complexity, closely associated with length and grammatical complexity of an utterance. Less complex semantic levels such as labeling or describing require shorter, less grammatically complex utterances. Conversely, more sophisticated semantic levels (e.g., interpretations, inferences, evaluations) often correspond to increased utterance length or grammatical complexity. This domain most closely resembles the traditional linguistic hierarchy used in fluency therapy that was described earlier.
The dynamic interaction of the three domains within a communication event is typically nonlinear; increased sophistication in one domain might precipitate decreased sophistication in the others. For example, as a child's ability to use fluency skills when talking about contextualized topics becomes more sophisticated in the Discourse and Semantic domains, a change to decontextualized topics (i.e., removal of contextual support) may necessitate a decrease in sophistication in the two other domains.
Steps in Applying the SDS Model to Fluency Therapy
In our opinion, the SDS model can assist clinicians in manipulating linguistic complexity by using it to structure the types of utterances produced during therapy. We have found that designing a theme around a client-initiated topic of interest helps facilitate meaningful communication within therapy sessions and also helps increase motivation for therapy. Using a theme also helps minimize cognitive-linguistic demands because the child can attach new learning, i.e., speech modification skills, to prior knowledge about the topic.
Traditional treatment approaches usually only emphasize manipulation of individual utterance length and complexity. In our approach, instead of targeting only utterance length and complexity, we also consider the importance of the Situational and Discourse domains, keeping Semantic complexity as a guide for utterance length/complexity manipulations within activities.
Step 1 in planning treatment is selecting a topic that is highly interesting to the child, or about which the child has considerable knowledge (e.g., sports, hobbies, curricular topics). This topic then provides a theme for therapy; most or all treatment activities and materials used to teach new skills will be based on this theme.
Step 2 requires the clinician to consider how much contextualization the child will need to support talking. To make this decision, the Situational domain of the SDS model is used as a guide. For example, if the clinician observes that the child is more disfluent when talking in decontextualized situations, then therapy activities will necessitate the use of with objects, toys, pictures, or print in order to decrease cognitive-linguistic demands. If the clinician is unsure whether contextualization affects the child's fluency, he or she should probe by collecting brief speech samples at various levels of the Situational domain (refer to the model). For example, speech sampling could occur by asking the child to talk while engaged in pretend play or retelling a story with pictures about getting ready for school (contextualized-symbolic), retelling their own experiences getting ready for school (decontextualized-egocentered), or retelling a scene from a favorite movie or television program (decontextualized-decentered).
Once the theme is chosen and the degree of contextualization is known, Step 3 involves developing ideas for therapy activities and collecting materials associated with the theme. We encourage children and/or their families or teachers to contribute ideas, information, and materials related to the theme so that we are sure to stay within the realm of the child's prior knowledge. Numerous materials for most themes can be obtained from toy sets, books, magazines, newspaper articles, and various websites.
Finally, at Step 4, the clinician should plan therapy activities that follow the theme and will facilitate learning of speech modification skills through systematic manipulation of various levels in the Semantic and Discourse domains. For example, if the clinician is targeting single word or phrase-level productions from the child, asking the child to label or describe (Semantic levels 3 & 4) materials or objects associated with the theme will often elicit these desired utterance lengths. If a child has just learned a new fluency-enhancing skill, the clinician might probe the child's ability to use it across several utterances by asking the child to recall three things that happened in therapy that day. If the child says, "We played a game, we talked about dinosaurs, and I got a sticker," he or she would be talking at Level 3 (descriptive list) of the Discourse domain. Later, when the child is using the fluency-enhancing skill in more complex speaking tasks, the clinician might ask the child to retell a story, thus eliciting a complete episodic structure (Discourse level 7).
As with traditional approaches using a length/complexity paradigm, we recommend that therapy be initiated at levels where the child is most fluent so that new skills can be introduced in a fluency-facilitating environment. As the child masters each skill, then systematic manipulations should be made within each of the three domains. In our clinical experience, the Situational domain tends to remain the most constant across therapy sessions because the clinician chooses to target either contextualized or decontextualized topics based on the client's need. Within a therapy session, however, the Semantic and Discourse domains can be varied across activities to produce different lengths and complexities of individual utterances as well as different structures of content.
Examples of Using the SDS Model in Fluency Therapy
To illustrate how to use the SDS model, we'll describe a sequence of therapy activities for a 7-year-old boy named Brad. Brad was enrolled in individual treatment, had no previous therapy, and was developing typically in all communication and learning areas except speech fluency.
At the evaluation, Brad was aware of his stuttering and felt negatively about it, being reluctant to talk in many situations. His negative emotions and reactions to stuttering were surprising, given that his stuttering severity was in the mild range (i.e., less than 8%). Brad's disfluencies were characterized by rapid part-word repetitions and some brief prolongations at the single word level, and he was more disfluent when talking about decontextualized topics than when talking about contextualized topics. When disfluencies occurred, a raise in vocal pitch often accompanied them. His parents described Brad as loving and extremely sensitive. Brad's father once scolded him for stuttering. Brad told us that his father didn't "like the way" he talked, thus Brad was reluctant to interact with his father. When we asked what Brad was interested in, he said he liked professional football, knowing many of the teams and players. He frequently watched football on television with his family and played in a junior football league once a week.
Based on Brad's disfluency pattern, targeted speech modification skills included easy onset of phonation, smooth transitions or continuous phonation, and pullouts. Following the steps outlined in the previous section, we made the following clinical decisions.
Step 1: Choose a theme that is interesting to the child. Football was selected as the theme for therapy because Brad was interested in and knew a great deal about that sport.
Steps 2: Consider the degree of contextualization needed. Brad exhibited better fluency when topics were contextualized, so we decided to keep initial therapy activities at Levels 3-5 (contextualized-relational, contextualized-symbolic, and contextualized-logical) of the Situational domain in the SDS model. This meant that our activities would involve using real objects (level 3), toys, books, or pictures (levels 4 & 5); and focus on playing football "games," (level 3), telling stories and recreating games with toys (level 4), or making up our own games or rules (level 5).
Step 3: Gather materials needed to support the chosen Situational context(s). We then gathered as many materials as we could to support the football theme. Some examples: We made a football field game board to use for game-based activities; we acquired a soft football and toy football players; we used colored paper to make "fans" and hung them on the wall. We either gathered or made and illustrated index cards listing players' names/positions/statistics, names of football teams, positions or plays in football, and rules to the game.
Steps 4: Plan therapy activities. Since Brad had never had fluency intervention, we decided to initially target easy onsets, smooth transitions, and pullouts at the single word and phrase levels (i.e., labeling and describing in the Semantic domain).
One activity we used to help Brad learn to use these skills, as well as begin changing his negative feelings about talking with his father, was having him brainstorm all the reasons someone might be a fan of a particular football team. As Brad listed reasons someone is a fan (e.g., liking the team colors, they win a lot, they're good sports), we wrote the words or phrases down. We then asked him to practice one of the speech modification skills by reading the list. We then asked him reasons someone might be his fan. He listed, "I'm nice, I'm helpful, I'm polite, I'm funny, I'm a hard worker." As homework, we asked Brad to "interview" his father and find out why he was a fan of Brad's. Brad was instructed to take the list home and tell his father why someone might be his fan, using the targeted speech modification skill practiced in therapy. His father was asked to add to the list of reasons and talk about them with Brad. When Brad returned to the next therapy session, he reported that his father "is one of my BIGGEST fans," and was excited to talk more about football and practice his speech modification skills.
In this brief example of an activity in therapy, we have used all three domains of the SDS model while also addressing Brad's emotional relationship with his father. Within the context of the SDS model, the activity included the:
Healey, E.C., Norris, J., Scott-Trautman, L., & Susca, M. (1998). Enhancements to integrated approaches for treating stuttering. International Stuttering Awareness Day Internet Conference, Minnesota State University, Mankato, MN.
Norris, J. and Hoffman, P. (1993). Whole language intervention for school-age children. San Diego: Singular Publishers.
Ryan, B.P. and Ryan, B.V.K. (1995). Programmed stuttering treatment for children: Comparison of two establishment programs through transfer, maintenance, and follow-up. Journal of Speech, Language, and Hearing Research, 38, 61-75.