Supporting Theory of Mind Development Considerations and Recommendations for Professionals Providing Services to Individuals with Autism Spectrum Disorder

Top Lang Disorders Vol. 28, No. 4, pp. 340–364 Copyright c 2008 Wolters Kluwer Health | Lippincott Williams & Wilkins Supporting Theory of Mind Development Considerations and Recommendations for Professionals Providing Services to Individuals With Autism Spectrum Disorder

Tiffany L. Hutchins, PhD; Patricia A. Prelock, PhD, CCC-SLP

Theory of mind (ToM) difficulties represent a core deficit underlying the social, behavioral, and communicative impairments characteristic of autism spectrum disorders (ASD). This article provides a developmental perspective on ToM that can serve as a framework for understanding and addressing ToM deficits characteristic of individuals with ASD to assist in education planning. Pop- ular methods for assessing ToM are reviewed and the use of social stories and comic strip conversations to facilitate ToM understanding is described in a case vignette, which provides preliminary evidence for the feasibility of this type of intervention for encouraging ToM development in children with ASD. Professionals who provide services to individuals with ASD are encouraged to incorporate assessment and intervention procedures that acknowledge the importance of ToM in their program plan and to consider the connections between ToM and communicative and behavioral functioning. Key words: assessment, autism, intervention, social stories, theory of mind HEORY OF MIND (ToM) may be broadly texts (Happe, 1994; Perner, Frith, Leslie, & Tconstrued as the ability to reason about Leekam, 1989; Prior, Dahlstrom, & Squires, the thoughts, feelings, and intentions of the 1990). This has led some to argue for what is self and others. Over the last 2 decades, ToM known as the ToM hypothesis of autism. From has become one of the most energized and this perspective, ToM represents a core deficit prolific areas of research in the field of autism underlying the social, behavioral, and com- spectrum disorder (ASD). As a result, there is municative impairments characteristic of ASD now broad consensus that there are autism- (Baron-Cohen, 1995; Baron-Cohen, Leslie, & specific deficits in ToM. Indeed, individuals Frith, 1985). with ASD generally perform poorly compared Professionals in health and education (in- with their age- and language-matched peers cluding speech–language pathologists, social on tests designed not only to assess their workers, case managers, special educators, ability to attribute false beliefs (Baron-Cohen, psychologists, teachers, and others) com- 1995) but also to infer others’ beliefs and emo- monly provide services to individuals with tions in a variety of social and situational con- ASD. Despite the theoretical importance of ToM, experience suggests that ToM is rarely considered in decision-making processes surrounding the health and education of in- From the Department of Communication Sciences, dividuals with ASD (e.g., when developing University of Vermont, Burlington. goals for therapy, drafting individual educa- Corresponding author: Tiffany L. Hutchins, PhD, De- tion plans [IEPs], or selecting measurable partment of Communication Sciences, University of Vermont, 406 Pomeroy Hall, 489 Main St, Burlington, outcomes). Several reasons may account for VT 05405 (e-mail: [email protected]). considerations of ToM being overlooked. 340 Supporting Theory of Mind Development 341

One reason undoubtedly involves heavy re- multifaceted construct (Astington, 2005). In liance on intensive interventions, such as a recent attempt to describe the breadth applied behavior analysis (Lovaas, 1987), of ToM, Hutchins, Bonazinga, Prelock, and TEACCH language and communication cur- Taylor (2008) offered a candidate list of re- riculum (Watson, Lord, Schaffer, & Schopler, lated constructs that may be subsumed un- 1989), “Floortime” and the Developmental der ToM or closely connected to it. These Individual Difference, Relationship-Based ap- included the ability to engage in or under- proach (Greenspan & Wieder, 1998), and Piv- stand false-beliefs, pretense, deception, de- otal Response Training (Koegel & Koegel, sire and intention, appearance–reality and 2006; Koegel, Koegel, & Carter, 1998). Each mental–physical distinctions, affect recogni- of these approaches has been evaluated as tion and the causes of emotion, the notion having some empirical support for remediat- that seeing leads to knowing, second-order ing the behavioral, communication, and so- thinking (e.g., understanding what Tiffany cial deficits associated with ASD (National Re- thinks Patty thinks), visual perspective taking, search Council, 2001; Prelock, 2006). Other empathy, and the understanding and produc- reasons that ToM-focused intervention may tion of mental state terms and speech acts. be overlooked involve difficulties surround- To this list should be added, appreciation ing the interpretation of results from tradi- of humor, knowledge of the links between tional ToM assessments and a failure to de- traits and behavior, understanding metaphor- velop interventions that support ToM devel- ical and nonliteral uses of language, the dis- opment and more optimal social interaction tinction between lies and jokes, and the ability routines among children with ASD. to engage in moral reasoning (Hutchins et al., Nevertheless, ASD-specific deficits in ToM 2008). Considering that this inventory is not continue to challenge providers, and they expected to be exhaustive, it is clear that have important implications for behavioral, ToM is a term that “refuses to be corralled” language, and social outcomes. The aim of this (Astington, 2005, p. 6). article is to provide a developmental perspective on ToM that can serve as a framework ToM development: Difficulties and for understanding and addressing the ToM considerations deficits characteristic of ASD within more Given the range of mental states and con- traditional educational programming. Popular ceptual understandings embodied in the term methods for assessing ToM are reviewed and “ToM,” charting its developmental course re- a promising approach to facilitate ToM under- mains a topic of debate. To complicate mat- standing is described in a case vignette. Rec- ters, few studies have examined longitudi- ommendations are offered for professionals nally the development of ToM of individuals who provide services to children with ASD. with ASD. Holroyd and Baron-Cohen (1993) reported no change in false-belief task performance among 17 children with ASD over THE STUDY OF ToM 7 years. Similarly, Ozonoff and McEvoy (1994) observed no change in false-belief or more ad- What is ToM? vanced ToM abilities among 17 adolescents Theory of mind has been widely used to with ASD over 3 years. describe a range of abilities inherent to the In contrast, using a more sensitive develop- development of social understanding. Despite mental battery of ToM abilities, Steele, Joseph, the common practice of focusing on partic- and Tager-Flusberg (2003) observed signifi- ular aspects of ToM for describing knowl- cant improvements in ToM scores of fifty- edge or performance such as false beliefs, seven 4- to 14-year children and adolescents researchers largely agree that ToM is most with ASD over a 1-year span. Moreover, ad- appropriately conceptualized as a broad and vances in ToM knowledge were represented 342 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008 relatively equally across the age range sur- false beliefs has obscured the importance veyed, challenging the view that individuals of other ToM achievements (Carpendale & with ASD show no improvements in ToM over Lewis, 2006). This is evidenced by a tendency time. Because individual variation characteris- to describe earlier achievements as simple tic of the autism spectrum far outstrips varia- precursors or prerequisites foundational to tion related to age, the development of ToM in the more sophisticated ToM characterized by ASD is tied not to chronological timetables but false-belief understanding (Tomasello, 1995). rather distinguished along dimensions of rel- In a related vein, investigations of the develop- atively early emerging, simple, or basic skills ment of ToM in typically developing children to relatively late emerging, complex, or ad- beyond the age of false-belief acquisition are vanced skills. This is evidenced most clearly scarce (Barr, 2006). Thus, comparatively little in attempts to develop a range of ToM tasks is known about the development of mental- (Happe, 1994; Hutchins, Prelock, & Chace, in izing in middle and late childhood and adult- press; Muris et al., 1999; Steele et al., 2003) as hood (Happe & Winner, 1998). This gives the discussed further below. erroneous impression that few or unimpor- For our purposes, it is important to describe tant ToM developments occur after the mas- current thinking about the general order in tery of false-belief tasks and leads to descrip- which certain aspects of ToM or ToM-related tions of developmental timetables (Barr, 2006; abilities emerge. Although this article deals Miller, 2006) that are artificially truncated. with how to support the development of ToM of children with ASD, understanding typical ToM and language relations ToM development can facilitate educational Much empirical evidence exists that ToM planning by determining what may be chil- and language competencies are strongly re- dren’s ToM strengths and challenges and iden- lated in typically developing children, but tifying where children’s functioning lies in re- even more so for children with ASD (Happe, lation to typical ToM achievements. This is 1995; Prior et al., 1990; Yirmiya, Solomonica- not to say that the pathways and developmen- Levi, Shulman, & Pilowsky, 1996). It makes tal processes operating in the development sense that language and ToM would be closely of ToM of typically developing children and associated. Language provides the means by children with ASD are equivalent. There is, in which children become aware of unobserv- fact, accumulating evidence to the contrary able mental states, and successful communi- (Frith, Happe, & Siddons, 1994; Happe, 1995; cation requires an understanding of others’ Tager-Flusberg, 2000). Nonetheless, the ToM minds. The links between language and ToM challenges faced by individuals with ASD are complex and are not, for the most part, may be best understood, explained, and re- theoretically disentangled in this discussion. sponded to through awareness of the form of It is appropriate to acknowledge, however, typical ToM development. that developments in language are expected to parallel those in ToM because this has im- Why false beliefs are not the whole story plications for ToM assessment and treatment Typically developing children are unable to planning. attribute a false belief to another person un- Although the nature of ToM develop- til approximately 4 years of age, and this re- ment has been debated (for a review, see sult persists after the effects of a number of Carpendale & Lewis, 2006), social construc- task variables have been statistically removed tivists emphasize the importance of guided (Wellman, Cross, & Watson, 2001). Although participation and joint attention and argue false-belief understanding represents an im- that children acquire a ToM by internaliz- portant component in the development of ing the meanings that are co-constructed ToM and has generated much excitement and during language-mediated social interaction theorizing in the field, preoccupation with (Hutchins, Bond, Silliman, & Bryant, in press). Supporting Theory of Mind Development 343

In accordance with this view, we conceptu- words, to the use of appropriate language alize ToM development as a continuous and in a range of communicative contexts. For coherent process of social–cognitive devel- example, the relatively advanced understand- opment (Chandler & Hala, 1994) and reject ing of one’s own utterances in the social the view that false-belief understanding is the and situational context is framed by a speak- most interesting or important ToM developing partner’s current knowledge and attention mental milestone. Given the unwieldy num- (Tomasello, 2003). Thus, an impaired ability ber of studies in this area and the goals of to understand others as intentional agents and this article, we describe some general trends to participate in episodes of joint attention is in typical ToM development by selectively re- believed to be causally linked to significant de- viewing the literature most commonly cited lays in language, which is a defining feature to describe ToM development in young child- of ASD (Baron-Cohen, Baldwin, & Crowson, hood while extending our description be- 1997; Tager-Flusberg, 2000). yond 4 and 5 years of age. Developments in The understanding that others can have language are known to accompany develop- desires different from one’s own is typically ments in ToM, and these links and their im- demonstrated in toddlerhood. Repacholi and plications for understanding ToM in ASD are Gopnik (1997) showed that 18-month-olds are elaborated upon when relevant. able to engage in behaviors to satisfy an adult’s desire even when the desire conflicts with ToM development in typically their own. Consistent with this observation, developing children children at 2 years of age are able to communicate using desire-state terms (e.g., want) Early ToM development spontaneously and appropriately during con- Perhaps, the earliest manifestation of ToM versation (Bartsch & Wellman, 1995). develops around the 9th month of life and By 3 years of age, children begin to make involves understanding intentionality or the reference not only to desires but also to be- ability to construe others as intentional agents liefs (e.g., think, know) in explaining ac- who have concrete goals or motives, which tion (Bartsch & Wellman, 1989; Wellman & drive their behavior (Tomasello, 1995). Al- Bartsch, 1988). This age is also characterized though not without controversy, the under- by an emerging ability to distinguish mental standing of intentions may be foundational from physical entities. Thus, a child knows, to another skill that develops at around the for example, that an actual puppy and not same time—that is, the ability to participate a thought of a puppy can be fed, and, con- in episodes of joint attention (Tager-Flusberg, versely, that a thought of a puppy and not 2000; Tomasello, 1995). Although gaze fol- an actual puppy can be transformed in size lowing, pointing, and the use of communica- (Wellman & Estes, 1986). Around this same tive gestures accompany and facilitate joint time, children also come to understand that attention, it is not established when two per- seeing leads to knowing. For instance, most sons merely coordinate their eye gaze. Rather, children can now understand that if one per- joint attention is established when each in- son lifts a box and another person looks inside dividual is aware of and monitors simultane- a box, only the person who looked inside will ously the other’s attention to an outside en- know the contents of the box (Pratt & Bryant, tity (Tomasello, 1995). This often includes 1990). shared enjoyment or shared affect (Prizant, The traditional hallmark achievement char- Wetherby, Rubin, Laurent, & Rydell, 2004). acteristic of 4 years of age is the abil- The ability to establish and respond to bids ity to consistently pass the standard false- for joint attention is involved in every aspect belief and Smarties tasks (Wellman et al., of ToM, from basic symbol learning, as evi- 2001), which are described more fully in denced in the acquisition of children’s first the subsequent text. A number of other ToM 344 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008 accomplishments are usually demonstrated at dren generally understand that prior expecta- this time, which undoubtedly contributes to tions and biases can influence interpretation the notion that this age represents a major of an ambiguous event, but it is not until 7 ToM developmental milestone. For example, or 8 years of age that children can adequately by asking children to talk about objects that justify their responses. Similar results explor- have misleading identities (e.g., a candle fash- ing children’s ability to attribute a variety of ioned to look like an apple), Flavell, Green, interpretations to others have been observed and Flavell (1986) demonstrated that this age using ambiguous visual stimuli (Carpendale & is characterized by the ability to distinguish Chandler, 1996; Lalonde & Chandler, 2002). appearance from reality (e.g., by reporting In a related vein, 7–8 years of age is associ- that the apple-shaped candle looks like an ated with the ability to understand humor on apple but is really a candle). Age 4 also is the basis of lexical ambiguity (McGhee, 1979). cited as the age at which children can ex- Consider the following joke: “Time flies like plain emotions on the basis of beliefs (Harris, an arrow. Fruit flies like a banana.” Although Johnson, Hutton, Andrews, & Cooke, 1989), children 5 and 6 years of age usually demon- distinguish mental state (e.g., think, dream, strate knowledge of the two meanings of the pretend) from nonmental state (e.g., run, word like in isolation, they are unable to ap- jump) terms (Baron-Cohen et al., 1994), and preciate the humor in a pun such as this be- monitor and report on their own intended cause, unlike 7- and 8-year-olds, they cannot versus unintended outcomes (Phillips, Baron- negotiate the dual-word meanings simultane- Cohen, & Rutter, 1998). By age 4, children ously (Carpenter & Lewis, 2006). also begin to take an interest in and engage Other aspects of social understanding in- in deception (Sodian, Taylor, Harris, & Perner, volve the intimate connection between ToM 1992), although they do so rather clumsily and and communicative competence. “Almost ev- unsuccessfully. ery aspect of pragmatics involves sensitivity to speaker and listener mental states, Later ToM development and hence mindreading”(Baron-Cohen, 2000, A more mature understanding of mind p. 13). Given the range of speech acts and the requires thinking about what other people contexts (including physical, social, and situa- are thinking (Carpendale & Lewis, 2006). tional contexts) in which they are embedded, This is known as second-order false-belief un- tying communicative competence to specific derstanding, which is estimated to emerge developmental timetables for the present pur- 1 between 5 /2 years (Sullivan, Zaitchik, & poses is not helpful. The point here is that Tager-Flusberg, 1994) and 6–7 years (Perner abilities to use language appropriately (e.g., & Wimmer, 1985) of age depending on the to take turns or maintain a topic of conver- complexity of the task and the language used. sation) and to understand language appropri- Of course, mastery of even second-order false- ately (e.g., to interpret metaphor and sarcasm) belief understanding is not adequate. Sophis- are viewed as relatively advanced skills that ticated social understandings are more com- are contingent upon the understanding of oth- plex and involve, for instance, the knowledge ers’ minds within the larger context (Happe, that two people whose experiences are iden- 1995; Tager-Flusberg, 2000). tical may nevertheless arrive at different interpretations. For example, two people who watch the same movie may have dissimilar in- ASSESSMENT OF ToM terpretations of the plot or impressions of the characters. Traditional ToM assessment procedures In examining the ability to construe others Assessment of ToM has been dominated as interpretive agents, Pillow (1991) demon- by what has come to be known as the strated that between 5 and 8 years of age, chil- standard false-belief (Wimmer & Perner, 1983) Supporting Theory of Mind Development 345 or Sally-Anne (Baron-Cohen et al., 1985) dures exist for evaluating a range of ToM com- task and the Smarties task (Perner, Leekam, petencies (for a review of additional selected & Wimmer, 1987), which have come to methodologies, see Baron-Cohen, 2000). serve as general markers of ToM competence One limitation relevant to this discussion in- (Astington, 2001; Wellman et al., 2001). The herent in each of these procedures may be traditional false-belief task involves telling a best illustrated through examination of the story in which an object is moved from an old false-belief task, which has been the most location to a new location without the knowl- widely used, but also most scrutinized, mea- edge of the main protagonist. The observing sure. False-belief task performance is tradi- child is then typically asked either a thinking tionally assessed on a dichotomous pass–fail question (“Where does Sally think the choco- basis, leading to the impression that ToM is late is?”) or a looking question (“Where will something that one does or does not have. Sally look for the chocolate?”) and a mem- That is, the ToM literature “often equates ory control question (“Where is the choco- performance on a false-belief task to the late really?”). Children who answer the test presence or absence of a theory of mind, question(s) with the new and incorrect loca- reducing what should be a rich, complex, un- tion fail the task, whereas children who an- folding mentalistic conception of people to swer with the old and correct location pass a categorical capacity” (Tager-Flusberg, 2001, the task, presumably by demonstrating their p. 178). In fact, much previous research appreciation that others’ thoughts or actions has shown that although more able individ- are guided by a false belief. uals with ASD and Asperger’s disorder often Concern that the Sally-Anne task may be pass first- and second-order false-belief tasks, unnecessarily difficult because it does not di- they nonetheless demonstrate ToM deficits rectly involve the young child as a partici- in more naturalistic situations (Happe, 1995; pant prompted the development of the Smar- Klin, 2000). Theory of mind is most appropri- ties task (Perner et al., 1987). The Smarties ately conceptualized as lying on a continuum task involves presentation of a container that of competence, and the inability of false-belief is revealed to hold surprising contents (tra- tasks to quantify developing competency of- ditionally a Smarties candy box that contains ten masks ToM deficits (Klin, 2000) or abilities pencils). The child is typically asked a test (Tager-Flusberg, 1999). question that requires predicting what a naive subject (i.e., someone who has not looked in- Alternative ToM assessment procedures side the box) will say or think is in the box To address concerns associated with tra- along with a memory control question (“What ditional ToM assessment procedures, re- is really in the box?”). Thus, the Smarties task searchers have developed procedures de- is similar to the false-belief task in that it re- signed to tap more advanced ToM skills. For quires inference of another’s false belief, but example, the “Strange Stories” task (Happe, it differs by involving the child more centrally 1994) was designed to assess the understand- as the protagonist who becomes aware of the ing of story characters’ thoughts and feelings unexpected contents. through identification of several pragmatic Although false-belief and Smarties tasks functions (e.g., lying, joking, pretending). (and their variants) are the most commonly More recently, Baron-Cohen, Wheelwright, employed indices of ToM, tasks have been de- Hill, Raste, and Plumb (2001) have refined an veloped to tap other aspects of mental state advanced ToM measure intended to assess the understanding such as those described above ability to read thoughts and feelings as evident (e.g., the mental–physical distinction, the no- in the appearance of the region of the face tion that seeing leads to knowing, and the surrounding the eyes. This test has been de- appearance–reality distinction to name just scribed as a measure of social sensitivity that a few). Thus, a variety of assessment proce- is one index of how well one can “tune in” 346 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008

(Baron-Cohen et al., 2001, p. 241) to the men- ies may be influenced by attention, mem- tal states of others. ory, linguistic, motivational, and situational Researchers also have recognized the need factors (Tager-Flusberg, 1999). The impact of for task batteries that assess different com- these shortcomings varies with the individ- ponents of ToM across levels of complexity ual and the assessment procedures employed. on the basis of the rationale that a broader To avoid these shortcomings, Hutchins et al. range of tasks allows for more sensitive mea- (2008) developed a psychometrically sound surement of ToM ability (Hutchins, Prelock, informant measure (see brief description of et al., in press; Muris et al., 1999; Steele the Perceptions of Children’s Theory of Mind et al., 2003). For example, Steel et al. (2003) Measure [PCToMM] in the case vignette de- developed a ToM task battery for assessing the scribed in the subsequent text), which is ap- development of ToM of verbal children and propriate for evaluating a range of ToM com- adolescents with ASD. The battery progresses petencies among ASD individuals who are from “early” to “basic” to “advanced,” with functioning at verbal and preverbal levels. tasks designed to capture ToM at each level. This tool is particularly attractive because it Mastery of the early level is judged on the abil- takes advantage of the persons in the child’s ity to engage in pretend play and predict ac- life who know the child best (Hutchins et al., tion based on a character’s explicitly stated de- 2008). Although there are notable exceptions sire. Basic-level competence is credited when (Hutchins & Prelock, 2006), combining quan- a child demonstrates understanding that see- titative and qualitative approaches to ToM as- ing leads to knowing, as well as by the abil- sessment is rare in the ToM literature, which ity to pass false-belief and Smarties (what they has been highly empirical and theoretical and call an “unexpected contents”) tasks and en- concerned with group performance and is- gage in episodes of deception. Finally, skills sues of experimental control. However, when reflecting advanced levels of ToM understand- considering evaluation and intervention for ing include second-order false-belief under- individuals with ASD, there is a need to com- standing, understanding the distinction be- plement a selection of more traditional tasks, tween lies and jokes, and the ability to pre- such as those described above, with qualita- dict behavior on the basis of personality traits tive and observational data. These data are im- and engage in moral reasoning. In a similar ef- portant for developing a fuller understanding fort to tap aspects of ToM representing a range of the ToM competencies of the individual in of content and complexity, Hutchins, Prelock, question. Not only do they acknowledge the et al. (in press) developed a task battery that value of collaborative efforts among families can be used appropriately and reliably with and professionals (Crais, 1993; Diehl, 2003) verbal or nonverbal individuals with ASD. The but also, they are natural choices for parents skills tapped range from easy tasks, such as and professionals who have opportunities to recognition of emotion, to difficult and com- gather and reflect on the information such plex tasks that require second-order false- data provide. belief understanding. In our own work to support the develop- Theory of mind task batteries are impor- ment of ToM of individuals with ASD, we have tant because they reflect growing recogni- found a variety of qualitative assessment meth- tion that there is more to ToM than false- ods to be useful. These include, but are by belief understanding and because they have no means limited to, two related approaches: the potential to highlight the specific ToM (a) observation during naturalistic activities strengths and challenges that an individual and routines and in more formal and struc- brings to the social problem-solving situa- tured testing situations and (b) triangulation tion. Of course, task batteries have their lim- to seek the impressions of parents, educa- itations as well. As on the false-belief task, tors, and other professionals who know the performance on more comprehensive batter- individual best. Because the presence of an Supporting Theory of Mind Development 347 observer may influence behavior, contexts taneous behaviors. This suggested that partic- (e.g., the home, playground, classroom, or ipants were generating and applying nonmen- therapy room) in which the observer (e.g., talistic rules to pass tasks, as opposed to de- the parent, the educator, or the therapist) veloping a genuine understanding of others’ is typically a functioning member of the ac- mental states. tivity provides an excellent perspective from Taking another tack, others have evaluated which to observe behavior in situations that the use of a “picture in the head” strategy matter to children (Brinton & Fujiki, 2003, to train ToM to children with ASD (McGre- p. 167). Parents and professionals have nu- gor, Whiten, & Blackburn, 1998; Swettenham, merous opportunities to observe children’s Baron-Cohen, Gomez, & Walsh, 1996). For social functioning and form impressions this method, photographs representing the about the nature of children’s underlying content of beliefs (including false beliefs) are ToM capacities. In line with this perspec- slotted into a mannequin’s or a doll’s head tive, parents have been found to be re- to make explicit the contents of another’s markably accurate in predicting their chil- mind during false-belief tasks. Using a differ- dren’s performance on a series of ToM ent pictoral analogy, Wellman et al. (2002) ex- tasks that require different forms and de- plored the usefulness of “thought bubbles,” grees of understanding of others’ men- such as those that appear in comics and car- tal states (Hutchins et al., 2008). Because toons, to train ToM. Using these approaches, parents and professionals typically observe it was found that children with ASD could un- children in different physical settings and derstand and make use of the strategy; how- communicative contexts, it is preferable to re- ever, training was usually associated with only cruit multiple observers (representing multi- modest generalization to transfer ToM tasks. ple disciplines) to contribute to the ToM as- More importantly, transfer tasks were remark- sessment process and treatment planning. Tri- ably similar to the training tasks in administra- angulation of this kind “not only demands tion and content and none assessed children’s seeking different voices, it also requires valu- functioning in naturalistic settings in which ing each of those voices” (Brinton & Fujiki, the problem surrounding the target mental 2003, p. 167). Such recognition is particularly state was not explicitly created and defined. welcomed from an interdisciplinary perspec- In summary, previous efforts to teach ToM tive in which professionals pool their exper- to children with ASD have failed to demon- tise to better suit the needs of children. strate generalization, particularly in novel social situations that would constitute most real- TEACHING ToM world applications of ToM knowledge. Several ToM training studies with typically developing Initial efforts to teach ToM were primar- children have also focused on a similar narrow ily concerned with whether children could set of outcomes operationalized by false-belief be taught to pass traditional false-belief mea- task performance. However, research in this sures. For example, early studies in this area area has been useful for identifying factors in essentially administered multiple trials and in- typical development that are associated with corporated explicit feedback and teaching of ToM advances. These factors include more so- principles (e.g., that seeing leads to know- cial and pragmatic variables and suggest that ing), which were believed to be foundational rich discourse incorporating explanations of for an understanding of false beliefs (Hadwin, others’ mental states facilitates ToM under- Baron-Cohen, Howlin, & Hill, 1996; Ozonoff standing (Appleton & Reddy, 1996; Clements, & Miller, 1995). Although children with ASD Rustin, & McCallum, 2000; Guajardo & could be taught to pass false-belief tasks, ef- Watson, 2002). It is important to remember, fects of teaching did not generalize either to however, that what works with typically de- transfer tasks or to more meaningful and spon- veloping children may not work for children 348 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008 with ASD. The hallmark impairments in as parents’ access to and participation in sup- language and social cognition among chil- port services. To more fully comprehend the dren with ASD have traditionally made this child’s functioning in home and school con- kind of explanatory approach exceedingly texts, the child’s parent(s) and primary ed- unfeasible. ucator(s) participated in in-depth semistruc- A few possible reasons exist why results tured interviews. Both were asked to reflect from previous efforts to teach ToM to children on social situations in which the child demon- with ASD have been so disappointing. Chief strated inappropriate behaviors, experienced among them is that they have not focused on communicative challenges, and appeared to what is relevant, motivating, and meaningful fail to understand others’ thoughts, feelings, in children’s lives. Similarly, they have neither and perspectives. As an example, when gath- been tailored to suit the strengths and chal- ering information about their children’s com- lenges of the child in question nor been con- munication challenges, it was explained to cerned with clinical insights that have the po- parents that children with ASD often expe- tential to inform assessment and intervention rience difficulty asking questions, greeting, efforts. initiating/maintaining topics of conversation, In what follows, we describe the case of a etc. Parents were asked to identify a commu- 5-year-old boy named Zach (pseudonym), nication skill they would like to see increase who is a participant in an ongoing, large- in their children. Once they identified a de- scale, and experimentally controlled study to sired communication skill, they were asked examine the efficacy of a specific interven- to identify specific situations in the home tion for remediating the core deficits of ASD when challenging communication events oc- (Hutchins & Prelock, 2008). Zach’s inclusion curred, who was usually involved, what was as a research participant in this study required said, how their children responded, and what the use of fixed data collection procedures they would like to see happen. This informa- (described in the subsequent text). Given his tion provided the context for writing a so- strengths and challenges and the priorities of cial story for children. Similar information was his family and the professionals who worked gathered about each child’s problem behav- with him, Zach’s case was selected for de- iors and impairments in perspective-taking scription here because the assessments and abilities. Interviewees also were asked to intervention ultimately implemented would identify their children’s strengths and strate- be fitting choices in a clinical setting that al- gies or accommodations that may help chil- lowed for great flexibility. Zach also repre- dren to be more successful in social situa- sented a classic case of autism and presented tions, along with the parents’ priorities for special challenges to our efforts to support intervention. his ToM development, making his case most As stated previously, professionals working illustrative. with children in other settings also may develop valuable insights and impressions of the children’s functioning. A review of rel- ToM INTERVENTION evant and available records (e.g., IEPs, diag- nostic evaluations) was therefore conducted Learning about the child and the family to learn about the impressions and clinical Sound clinical judgment requires enough recommendations of other professionals in- knowledge about the child’s functioning to volved (e.g., pediatricians, social workers, tailor intervention goals and methods to the speech–language therapists, physical and oc- child’s specific needs. To learn about the child cupational therapists, and psychologists). Fi- and the family, a detailed personal history was nally, standard tests were administered to as- first conducted to understand the nature of sess children’s social, communicative, and the child’s current educational setting, as well ToM functioning. All of these components are Supporting Theory of Mind Development 349 described in greater detail in the following behaviors were noted when either slight sections. or considerable changes in routine were introduced in the school setting; for example, Zach when exiting through one door rather than When we met Zach, he was a verbal another or when the fire alarm at the school 5-year-old who had been diagnosed with ASD was sounded. a few months earlier by a developmental pe- Strategies to support Zach’s ability to tran- diatrician. Zach was an only child who lived sition from one activity to the next more suc- with his birth parents. He attended an Essen- cessfully at home and at school included regu- tial Early Education program (services for chil- lar reference to picture schedules of the day’s dren 3–5 years of age) and received services activities and frequent breaks to allow Zach to from a special educator three times per week, calm down, count to 10, or take a few deep adaptive physical education once per week, breaths. He was also encouraged to “use his and speech–language therapy three times per words”to talk about his thoughts and feelings week, which focused primarily on communi- and told firmly that his aggressive behaviors cation skills and speech production. He also were “not okay.” In the school, social stories received one-on-one support for social inter- (about how to play nicely and ask for toys) actions three times per week. were developed and read to Zach, but his mother and teacher reported that they were Interview not used enough. Unfortunately, all of these When asked to reflect on problem be- strategies were characterized as generally in- haviors, communicative challenges, and effective, with fluctuations in effectiveness at- perspective-taking deficits, one theme dom- tributed mainly to Zach’s state of mind and inated interview comments made by Zach’s whether he was having a “good day or a bad mother and his special educator. Both re- day.” ported that Zach would often become defiant Zach was described as a sensitive and af- and aggressive at home and at school and fectionate child who strived to please his agreed that this behavior represented a clear parents. One of Zach’s greatest strengths priority for intervention. They further ex- was his frequent remorsefulness following his plained that aggressive behaviors were most tantrums. He would become tearful and sor- likely to occur in situations in which “things rowful following outbursts at home. He could don’t go the way he wants” and associated also reliably forecast his mother’s approval on with rigidity and difficulties surrounding one of his “good days”at school when he was transitions between activities or a change in eager to share with her how he “didn’t throw plans. Zach’s mother and his special educator a fit”for which he received great praise, affec- offered some examples they had recently tion, and high-fives. observed. For instance, his mother reported that when the family went out to dinner a few Record review days earlier, Zach learned that the restaurant According to a record review, results from was out of apple juice, which was contrary to a battery of standard tests conducted at an his desire and expectation. Zach attacked the initial evaluation by a social worker, a psy- waitress and tried to hit her saying “You’re chologist, and a developmental pediatrician bad.”At school, Zach often enjoyed riding on revealed scores on measures of verbal and a small yellow scooter at recess. When the nonverbal intelligence that fell almost 1 SD scooter was unavailable, he would lunge at below the mean. Of greater concern, how- his teachers or other children, grab things ever, were suspected significant impairments and throw them to the ground, or hit, scratch, in social/emotional and behavioral functions. or kick others while yelling things such as On the basis of informal observations and “you are mean” or “you go away.” Similar consistent with the testimony of his mother 350 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008 and special educator, Zach’s aggressive behav- will be happy if they get what they want) iors were characterized as “willful,” “opposi- to difficult and complex (first- and second- tional,” and “explosive,” particularly in novel order false-belief understanding), with several situations or those requiring adjustment and degrees in between. Zach was able to main- transitioning. He also engaged in avoidant be- tain attention during this activity with mini- haviors such as crawling under the desk, at- mal difficulty. Zach passed almost all of the tempting to leave the room, and using fleet- control questions, thus demonstrating suffi- ing eye contact. Notably, Zach demonstrated cient executive and linguistic skills to perform some recognition that his difficult behaviors the tasks. That is, he could understand the lan- affected his parents, although he seemed un- guage used and recall relevant details on even able to communicate accurately these men- the longest and most complex tasks in this tal states. For instance, following a tantrum, battery. Therefore, it was revealing that Zach he vacillated in his comments about his par- failed all but the earliest ToM test questions, ents’ probable emotional reactions, stating which represented the easiest items in the bat- that “Dad gonna be angry” and conversely tery. In line with his mother’s evaluation of his “Dad gonna be happy.” ToM abilities, he was able to identify emotions on the basis of their visual characteristics (al- ToM assessment beit inconsistently) and understand situation- Measures administered to gain insights into based emotion (i.e., that a boy would be angry Zach’s ToM functioning included a parent- when someone did something that he did not informant ToM measure (i.e., the PCToMM; like) and desire-based emotion (i.e., that a girl Hutchins et al., 2008) and a ToM task bat- would be happy if she got what she wanted). tery (Hutchins, Prelock, et al., in press), both At the same time, he was unable to pass the of which were designed to assess ToM skills more advanced seeing-leads-to-knowing, false- across a range of content and complexity. belief, and second-order reasoning tasks, also The PCToMM was administered first. For the consistent with his mother’s evaluation of his PCToMM, an informant is asked to read sev- ToM deficits. eral statements and indicate the level of his In conclusion, the findings from these two or her confidence that the target child pos- ToM measures converged to present a portrait sesses a particular understanding (e.g., “My of Zach’s ToM development that was consis- child understands that when people get what tent with the earliest abilities that emerge in they want, they will be happy”). Items that typical development such as the ability to en- scored lowest (i.e., ranked by Zach’s mother gage in joint attention and understand how as “definitely not present”)reflected the areas desires are (or are not) satisfied. Skills devel- of ToM functioning related to understanding oping around and later than age 3 were not that seeing leads to knowing, intended versus demonstrated. This is not to say that Zach’s unintended outcomes, false beliefs, second- understanding of others’ minds was function- order thinking, and speech acts (e.g., white ally equivalent to that of a 3-year-old, but the lies, deception). Items that scored highest patterns of performance were interpretable (i.e., ranked by Zach’s mother as “definitely in light of typical developmental timetables. present”)reflected the ability to appropriately They also revealed areas of relative strength use desire terms (e.g., want) and identify and weakness to be considered when design- emotional expressions (e.g., happiness). ing intervention. The ToM task battery of Hutchins, Prelock, et al. (in press) consisted of a series of test Social stories and comic strip and control questions across several tasks that conversations progressed generally from easy or basic (e.g., As noted previously, Zach was a participant emotion recognition and desire-based emo- in a larger study that was designed to assess tion, which is the understanding that people the efficacy of an intervention for remediating Supporting Theory of Mind Development 351 the core deficits of ASD. Of particular interest ries and CSCs can be powerful for advanc- to the current discussion were our attempts to ing ToM development (Hutchins & Prelock, support his ToM development using social sto- 2006). This has important implications be- ries and comic strip conversations (CSCs; de- cause increments in children’s ToM have been veloped by Gray, 1994a, 1994b, respectively). associated with more appropriate social and Social stories are short stories developed to communicative behaviors that can generalize help the child reflect on social situations and to socially valid outcomes in a range of situa- inform and advise the child on how to be tional contexts (Hutchins & Prelock, 2006). more successful. Social stories are presented in a storybook format and incorporate reading Individualizing the intervention as a major part of the activity (for detailed in- Interviews with Zach’s mother and special formation on writing effective social stories, educator, reviews of evaluations conducted see Gray 1995, 1998; Gray & Garand, 1993; by a social worker, a psychologist, and a devel- Hutchins & Prelock, 2006). Comic strip con- opmental pediatrician, and our own observa- versations are similar to social stories because tions of Zach during low- and high-structured they are both visual systems designed to sup- testing situations revealed remarkable conver- port the understanding of social situations, gence in assessment. In particular, frequent but CSCs differ from social stories by relying concerns were raised regarding Zach’s aggres- more heavily on the child who co-constructs sive behaviors, and all parties identified this (most often by drawing and writing) the as a priority for intervention. Thus, remedi- conversations. ation of these behaviors was seen as criti- Social stories and CSCs are designed to min- cal to building a successful educational pro- imize factors during social interaction that gram. Given the frequency and explosiveness may be confusing to children with ASD be- of Zach’s tantrums, as well as the diverse situ- cause of impairments in ToM and the more ations in which they occurred, we sought to general tendency to engage in detail-focused promote Zach’s understanding of the causes processing at the expense of global and con- and consequences of his own and others’ textually meaningful processing (Gray, 1998). mental states in hopes of remediating this Social stories and CSCs highlight missing so- behavior. cial information, which “establish social understanding as an integral and prerequisite Zach’s social story component to teaching social skills” (italics Gray (1995) and Gray and Garand (1993) added, Gray, 1998, p. 169). Furthermore, they suggested that sentences in a social story are in line with what is considered best prac- be descriptive, perspective, or directive, al- tice for working with children with ASD. For though sentences in social stories are some- instance, they make use of the visual mode, times categorized using slightly different cod- which is typically an area of strength of chil- ing schemes. Zach’s social story comprised dren with ASD, and they are written from chil- several descriptive sentences to provide in- dren’s perspective so that they draw on ex- formation about the relevant context and de- periences that are personally relevant and po- fine the setting, people, or activities. Multi- tentially motivating to the child (Gray, 1995, ple perspective sentences were incorporated 1998). to explain the causes and consequences of Although previous research has docu- his tantrums in terms of his own and oth- mented the efficacy of social stories for ers’ mental states (e.g., “It makes my mom decreasing specific inappropriate behaviors and dad sad when I throw a fit”). To elab- (Norris & Dattilo, 1999; Kuttler, Myles, & orate on these perspectives, we borrowed Carlson, 1998; Swaggart, Gagnon, Bock, & from Gray’s description of “thinking stories” Earles, 1995), evidence is just beginning to (Gray, 1994b) and included “thinking” sen- emerge suggesting that the use of social sto- tences, which make explicit the thoughts and 352 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008 feelings of others (e.g., “They might think ‘I office, home) and situational contexts (e.g., wish we could make Zach happy’”). Direc- no apple juice available at a restaurant, visiting tive sentences were also included to advise the doctor, losing while playing a board game, Zach how to avoid tantrums by offering him or missing a favorite TV show). Each topic more socially acceptable behavioral alterna- was revisited several times over the course of tives (e.g., “If I get scared or mad or disap- the intervention. An example of one of these pointed, I can use my words to talk to my CSCs is presented in Figure 1. Because evi- mom and dad”). In striving to ensure that dence supports the practice of repeated pre- the language used was meaningful to Zach sentations of social stories and CSCs for es- and responsive to his language level, Zach’s tablishing new and more appropriate social mother reviewed and edited the social story. interactions (Hutchins & Prelock, 2006), The social story also was designed to cap- these were presented to Zach a total of 15 italize on Zach’s ToM competencies, which times over the course of 5 weeks (approxi- included the ability to engage in joint atten- mately three times per week) by the first au- tion, recognize some facial expressions, and thor in the family’s home. understand situation- and desire-based emotions. Each sentence in the social story was Evaluating the effects of intervention presented on a separate page, with accompa- A simple ABA design with a 6-week follow- nying color illustrations. Zach’s social story is up was used to assess the effects of interven- presented in the Appendix. tion. Brief individualized diaries were developed to obtain the mother’s daily and gen- Zach’s comic strip conversations eral impressions of the specific behaviors and Comic strip conversations incorporate social understandings targeted in the social drawing, writing, and conversation as major story and CSCs. Specifically, Zach’s mother components of the activity, and they use a rated her general and subjective impression basic set of symbols (e.g., talking and thinking regarding change by indicating the degree bubbles) to make explicit what people say to which she agreed with several statements and think. The use of CSCs is a dynamic on a 10-point Likert-type scale anchored by intervention process that allows a child to “strongly disagree” to “strongly agree.” The revisit and elaborate on a recurring problem, effects of this particular intervention could highlighting the talking and thinking around be reflected in at least three different out- different events that lead to behavioral re- comes. First, effectiveness could be evident in sponses and perspective-taking challenges. a reduction of aggressive and violent behav- As Gray explained: iors. We characterized this as the ability “to Good opportunities for [comic strip conversations] stay calm” when things did not go the way include situations that are causing difficulty, ex- Zach wanted, when there were unforeseen plaining the responses of others, or preparing for changes, or when he was required to transi- a new situation or unfamiliar event. The content of tion to a new activity. Thus, Zach’s mother a conversation is simultaneously illustrated, guided was asked to respond to the statement “Based by carefully selected questions that assist a student on my judgments today, Zach is able to stay in sharing information. Each [comic strip conversa- calm in these situations.” Second, effective- tion] systematically identifies what people do, say, ness could be evident in an enhanced abil- and think. (1998, p. 183). ity or tendency to talk about the distressing For Zach, a previous CSC was revisited and event, which was one of the strategies em- elaborated on some days, whereas the need phasized in the social story and CSCs. Zach’s for a new situation was identified by Zach or mother was asked to respond to the state- Zach’s mother on other days. The specific top- ment “Based on my judgments today, Zach is ics of the comic strips surrounded a limited able to use his words to talk about these sit- number of settings (e.g., restaurant, doctor’s uations more appropriately.” Finally, we were Supporting Theory of Mind Development 353

Figure 1. Zach’s comic strip conversation. 354 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008 interested in impressions of Zach’s mentaliz- not seen until the final weeks of the with- ing ability, which are believed to underlie his drawal period during which a modest but con- social and communicative behaviors. Thus, sistent rise in slope is evident. Zach’s mother was asked to respond to the We also asked Zach’s mother to elaborate general perspective-taking question. In recog- on her quantitative ratings by recounting any- nition that such a question could be inter- thing that was different or interesting about preted with wide variation, the question was her son’s behavior, communications, and so- phrased as follows: cial understandings, and she provided these on a highly regular basis. These qualitative re- Based on my judgments today, Zach is able to ports (see Table 1 for selected examples) are perspective-take. That is, based on my judgments today, Zach is able to understand that other peo- important because they reveal more about the ple have thoughts and feelings that are different nature of the effects associated with interven- from his own. This can be a difficult question to tion. Notable was Zach’s nascent ability to use answer. Remember, we are asking for your overall language to negotiate difficult situations. Ac- or most general feeling in relation to what you have cording to maternal reports, Zach was able to observed today. engage with others more frequently in significant discussions about his own and others’ If no opportunities arose to form impres- thoughts and feelings, as well as the causes sions about the content of the statements and consequences of these. Moreover, he did on a particular day, Zach’s mother was in- so in multiple settings with a variety of in- structed to select a “don’t know”response op- terlocutors (i.e., peers, teachers, parents). Al- tion, which was treated as missing data. She though Zach’s emotional reactions remained also was encouraged to report any informa- intense, they were not accompanied by vio- tion that would be helpful for understanding lent and potentially injurious outbursts as they the nature, context, and frequency of the be- had been. To his mother’s surprise, he would haviors that were relevant to the targets of now cry and seek dialogue. Not only was this intervention. “more manageable” for his mother, but it was also a means for coping that provided oppor- Findings tunities for the language-mediated social in- The ratings of Zach’s mother on the out- teraction so important to the development of comes across ABA phases of the study are ToM. presented in Figure 2. Clear baseline stabil- ity is evident for all three outcome measures. DISCUSSION On the 11th day of the intervention phase (at which time the social story and CSCs had Although we have emphasized the impor- been presented five times), change in a thera- tance of supporting the child’s ToM, it is dif- peutic direction is evident in Zach’s mother’s ficult to disentangle the direction of effects subjective ratings of Zach’s ability to stay calm between increased use of dialogue and ToM and use words to talk about difficult situa- development and to remove the potential bias tions. Although there are fluctuations in daily introduced by Zach’s mother’s knowing that performance (as would be expected), marked he was receiving the experimental treatment. changes in level and slope are clear between Examination of trends in the quantitative ma- baseline and intervention phases of the study. ternal ratings reveals that increases in the abil- High ratings during the 6-week withdrawal ity to talk about difficult situations precede phase are maintained and show even more im- later occurring (and modest) ratings of en- provement with less variability and more con- hanced generalized perspective-taking abili- sistent ratings in the higher range. With regard ties. These findings are consistent with in- to ratings of Zach’s generalized perspective- terpretations offered by previous researchers taking abilities, reliable increases in ratings are (Dunn, Bretherton, & Munn, 1987; Dunn, Supporting Theory of Mind Development 355

Figure 2. Maternal subjective ratings across A (baseline), B (intervention), and A (withdrawal) phases of study for three outcomes (i.e., being able to stay calm in this situation, being able to use words to talk about the situation, and being able to take others’ perspectives). aDays on which the intervention was delivered are noted on the top figure only.

Brown, & Beardsall, 1991; Lewis, Freeman, led to shifts in ToM understanding, which, Kyriakidou, Maridaki-Kassotaki, & Berridge, in turn, facilitated communicative changes. In 1996) that the ability to engage in discussions this case, the lag in perspective-taking ratings about the causes of mental states across di- might simply reflect the fact that it is through verse physical, social, and situational contexts these kinds of communicative exchanges (and may be of special importance in the develop- over time) that a parent informant becomes ment of ToM. This discussion works to focus aware of the mentalizing ability of the child. attention on mental processes and thus foster As noted in this article’s introduction, it may a greater sensitivity to the existence and na- be most appropriate to conclude that these ture of mental states. On the other hand, the relationships are not strictly unidirectional content of the social story and CSCs may have but rather are dynamic and transactional, 356 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008

Table 1. Examples of maternal comments regarding child’s behaviors by day (see Figure 2)

Day Comments

29 Usually Zach has a major fit (screams, bangs on the PC with hands) when playing games on the PC. Today Zach came to me and told me he was upset and “angry.” He did this three times! Yeah! 31 Played “chutes and ladders” and “Candy Land” with Zach. He was very excited about playing. He said that he was going to “follow the rules” and “not have a fit.” I actually had a nice time playing a game with him. We had no problems. 35 After a bad morning, we went out for a hike. We were not able to get our usual spot on the trail. We had to drive to a new trail. Zach was very cooperative and flexible, more so than usual in a situation where the routine changes. 37 Zach asked for help again with his computer instead of having a tantrum. He was having trouble with his train program and came and got me for help. 43 Played games with mom and dad today. A little different with the three of us. Zach lost one game and cried. He asked to play again. I was surprised he cried but very proud of him for not lashing out at me or dad and not throwing a fit. 50 Asked for help to fix a train track set up that wasn’t coming together for him easily. He didn’t get upset about the pieces not fitting together. He helped me change it so that it would work. 52 Played outside. Zach usually plays in the backyard and deck. Unable to do that today (beehive). He followed directions and stayed in the front. No problems. I notice over the last few days that Zach still gets upset in situations; however, he seems to cry more instead of lashing out and trying to hit or break things. 54 Zach got a haircut today. It is usually a very upsetting experience for him; however, he didn’t lash out or hurt anyone or anything. He cried and was upset, but his behavior was much more manageable. 55 Played with classmate today. Followed her lead with play choices. Did have a tantrum over his schedule at school, but was able to talk about it and move on. 58 Zach walked on the nature trail with class. He was promised he would cross two bridges, but they were flooded. He wasn’t able to cross them. He screamed but then talked about why he was upset and talked about ways to fix the bridges. 63 Half day at school. Even though Zach knew what was going to happen, he still became very emotional and upset. He was not aggressive but very upset. 66 Able to calm down at school with help. Spilt milk on himself and calmed down and listened to his friends who said it was OK. 71 Altercation at school with two other kids. They weren’t doing what Zach wanted to do. Zach imposed a self time-out, calmed down on his own, and returned to play with the same kids. 73 Tough day with Zach. Easily upset with everything today (games, toys, even meals). Not physically aggressive though. 77 Was tested at school today (had to demonstrate some fine motor skills). Had some frustrations. Talked about it. Finished the test. 78 Played outside with classmates. Followed their lead. Played chase. Did what they wanted to do. Didn’t control play or get upset. He comes to me when he is upset with games/play saying “I’m angry” and asks for help. 81 Very upset about a situation where we were not able to follow Zach’s plan. He was very willing to talk about it afterwards and make a new plan. (continues) Supporting Theory of Mind Development 357

Table 1. Examples of maternal comments regarding child’s behaviors by day (see Figure 2) (Continued)

Day Comments

85 Better at remaining calm in most situations in general. Dropped lunch tray at school—food everywhere—remained calm. Helped pick up food and asked for another tray. No problems. 89 Surprise visit from a friend today. Zach did great. Some problems with toys, but he came and asked for help. 93 Zach was upset with some kids in class choosing sand table to play instead of doing what Zach wanted. Zach talked about it with aid in hall and calmed down and rejoined the class for the rest of the afternoon. Also played musical chairs with class when the game got too much for him, he asked to leave and remained calm.

underscoring the importance that “the two Unlike previous attempts to teach ToM not be treated as competing hypotheses but that describe alternative, artificial, or compen- rather complementary accounts” (Astington, satory pathways to ToM that serve as “pros- 2001, p. 686) thetic devices” (Wellman et al., 2003), social Our intervention with Zach produced pre- stories and CSCs arguably facilitate access to liminary evidence that intervention using so- the processes believed to operate in the typi- cial stories and comic book conversations can cal development of social understanding. The achieve clinically meaningful changes across primary difference may involve the fact that a range of real-life contexts with a child with whereas typically developing children iden- ASD. Although the observation of a single tify and understand the relevant metacogni- case and our use of ratings from an informant tive aspects of a social situation easily and im- who was not blind to her child’s treatment plicitly (Peskin & Astington, 2004), children status (including the time at which the in- with ASD do not. From a social-constructivist tervention phase of study began) limited our perspective, social stories and CSCs may help ability to rule out threats to validity (e.g., children with ASD engage in shared meaning maturation, history, response demands, ob- making and enable them to reason through so- server bias, placebo effects), the results are cial phenomena during episodes of language- consistent with the notion that social stories mediated joint attention. In the context of and CSCs may have potential for advancing guided participation with a mentor, the use of behavioral and cognitive development. The repetition, visual stimuli, structured activities, fact that the mother’s ratings in the daily di- and topics that are applicable, pertinent, and aries were supported by her anecdotal elabo- motivating may clarify the relevant aspects of rations and justifications of these ratings (see the social situation. Table 1) lends further confidence to the interpretation that observable changes in Zach’s IMPLICATIONS AND FUTURE behaviors were associated with the interven- DIRECTIONS tion. Although research is needed to clarify whether the intervention was responsible for Research on the effectiveness of using so- these changes, we suspect that social stories cial stories and CSCs to support development and CSCs may be particularly effective be- of ToM and more appropriate social and because they are specifically designed to give in- havioral routines is in its infancy. The limita- dividuals with ASD explicit social information tions posed by an uncontrolled single-subject (Gray, 1994b). ABA design does not allow for inferences of 358 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008 causality. Therefore, our interpretations are agnosed with ASD, was nonverbal and ex- speculative. A full-scale, controlled, longitudi- hibited a seizure disorder and visual impair- nal study is needed to meet standards of in- ment in one eye secondary to complications ternal validity and more fully examine causal as a premature infant. He experienced con- linkages. It should be noted, however, that al- siderable difficulty establishing and respond- though Zach’s mother was aware of when the ing to bids for joint attention. Brady’s so- intervention was initiated, change in maternal cial stories and CSCs emphasized only the ratings was evident only after several interven- earliest and most basic ToM skills related to tion sessions and that incremental improve- emotion recognition (i.e., being able to iden- ment in ratings persisted into the withdrawal tify happy, sad, and mad faces). Although phase of study. This suggests that repeated ex- social stories and CSCs were reasonably ef- posure and experience with this kind of inter- fective in helping Brady recognize emotions, vention is needed to establish more appropri- other strategies that were not a part of our ate and enduring social interactions, and this experimental protocol represent appropriate is in line with previous literature in this area choices to further enhance ToM and com- (Hutchins & Prelock, 2006; Norris & Dattilo, municative functioning. More specifically, a 1999). Furthermore, the timing in the pattern focus on reinforcing joint attention would observed is inconsistent with the interpreta- be desirable for a child who lacks this basic tion that changes were due to placebo effects skill. Recall that joint attention is an impor- and would be highly coincidental (although tant mechanism underlying ToM and language not impossible) if due to maturation or other development. Strategies to facilitate joint at- co-occurring treatments. No other support tention might make use of relationship-based services or educational programming were interventions (Greenspan & Wieder, 1998; implemented during the course of this inter- Gutstein & Sheely, 2002a, 2002b; Ingersoll, vention because it was administered over the Dvortcsak, Whalen, & Sikora, 2005; Prizant summer months when the school was out of et al., 2004) facilitating the child’s ability to session. establish two-way communication, share en- Further work in this area is also needed to joyment, and engage in symbol use through identify factors that are related to efficacy. We motivating activities and frameworks for play. do not wish to give the false impression that Through experiences with children in our this kind of intervention will be uniformly suc- study who represented a range of function- cessful or easy to implement. To the contrary, ing across the autism spectrum, we learned some of the children that we worked with that progress often required creative tailor- in the larger study demonstrated very few ing to the child’s skills and, in particular, the changes (Hutchins & Prelock, 2006, 2008). child’s ability and willingness to engage so- This is to be expected considering that no one cially with a clinician. Zach’s intervention was intervention has been found to be effective for particularly challenging given his explosive every individual with ASD (Diehl, 2003). To and avoidant behaviors, which continued in some degree, the lack of therapeutic change the context of treatment until the final week in the larger study was probably due to the use of the intervention. Thus, a combination of of our fixed experimental procedures. One visual schedules, frequent breaks, token re- advantage of working in a clinical setting is the wards, the use of timers, and gentle physical freedom to modify assessment or intervention control were incorporated during interven- procedures to support movement toward be- tion. We imagine that professionals who ex- havioral or performance goals. perience similar challenges may benefit from To illustrate how the intervention might be employing a variety of strategies to ensure modified in a clinical setting, we offer the child engagement. Although Zach strongly ob- case of another boy who participated in our jected to the intervention and confided in his research. Brady (pseudonym), a 4-year-old di- mother that the stories were “bad”and “scary,” Supporting Theory of Mind Development 359 his discomfort might have revealed insights professionals who seek to conduct interven- into his own inability to express emotions and tions to promote ToM in a way that can meet understand situations that he knew were dif- standards for a higher level of evidence. Our ficult for him. This is consistent with previous data collection procedures would have ben- observations (Hutchins & Prelock, 2006) and efited from triangulation through the use of provides a rationale for persevering with a dif- multiple raters, some of whom could be kept ficult intervention but in a way that contin- blind to the point in the intervention process. ually attempts to understand the child’s per- Convergent findings across raters would of- spectives and experiences that may not be fer more convincing evidence for an interven- fully understood by parents or professionals tion’s efficacy than could be provided using who work with the child. the current methods and would argue against Our findings agree with others who have the interpretation that responses were the re- used subjective evaluation of informants close sult of altered response demands or interac- to the child to determine whether those who tant bias (Prelock & Hutchins, 2008). Unfortu- interact with the child regularly see changes nately, other usual daily observers could not that are perceived as important and mean- be engaged in this research because Zach’s ingful (Kazdin, 1977). The use of such mea- intervention took place over the summer sures is also consistent with a family-centered months when he did not have regular con- approach (McCauley, 2001). Because the es- tact with educators or other professionals tablishment of clinical significance requires involved in his education planning and health- a frame of reference or perspective (Kazdin, care. Of course, another alternative for pro- 1999), we recommend regular evaluations fessionals would involve the reliable cod- over baseline and intervention phases, using ing of observations in relevant contexts carefully designed rating scales that tap areas in which target behaviors are expected to of functioning identified as priorities for reme- occur. diation. We also recommend that parents be Subjective ratings also benefit from trian- centrally involved in the identification of spe- gulation through the use of multiple mea- cific targets of intervention and the develop- sures. We encouraged informal diary descrip- ment of the intervention materials such as so- tions to add qualitative data regarding the cial stories (Prelock & Hutchins, 2008). This is nature of change being reported to the quan- particularly important to ensure that the con- titative ratings. Postintervention administra- tent of the social stories (including the lexical tion of the ToM informant measure (Hutchins items chosen, the level of grammatical com- et al., 2008) might also be appropriate to ex- plexity, and the objectivity and accuracy of amine whether intervention was associated the descriptions of social context) is appro- with a change in score. Standard ToM task bat- priate to the child’s language level. Whether teries or individual ToM task probes represent interventions take place at the home, the another possibility for measuring change, al- school, or the clinic, parents should be given though improvement over time may be diffi- the opportunity to confirm more broadly cult to interpret given the repeated presenta- that collaborators have understood and cap- tion of tasks and the potential for test practice tured the context and nature of the targets of effects (Hutchins et al.). Formal tests may be intervention. used most appropriately either in the context We chose to involve parents as informants of assessment or in controlled experiments. to gauge the outcome of the intervention. In any case, performance on formal tasks to The informant described in this article was indicate change (or lack of change) must be not blind to the treatment status of her child, interpreted with caution. which introduced the threat of biased re- We encourage professionals who provide sponding. The limitations posed by our data services to individuals with ASD to consider collection procedures could be instructive to incorporating assessment and intervention 360 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008 procedures that acknowledge the importance benefit from further controlled evaluations of of ToM, as well as the connections between the intervention described here, with involve- ToM and communicative and behavioral func- ment of other professionals working with chil- tioning. This is meant as a call for the match- dren with ASD who can be valuable partners ing of intervention to the family’s vision and in this endeavor. Furthermore, the use of so- the situational contexts important to the fam- cial stories and CSCs should not be utilized in ily and the professionals who work with lieu of other interventions but rather incorpo- children. Such matching can be achieved rated into a comprehensive program that ad- through a combination of thoughtful inter- dresses the core deficits of communication, views, progress monitoring, and standard as- social interaction, behavior, and ToM for chil- sessments. Evidence-based practice could also dren with ASD.

REFERENCES

Appleton, M., & Reddy, V. (1996). Teaching three year- attribution of action to beliefs and desires. Child De- olds to pass false belief tests: A conversational ap- velopment, 60, 946–964. proach. Social Development, 56, 275–291. Bartsch, K., & Wellman, H. M. (1995). Children talk Astington, J. W. (2001). The future of theory-of-mind re- about the mind. New York: Oxford University Press. search: Understanding motivational states, the role of Brinton, B., & Fujiki, M. (2003). Blending quantitative and language, and real-world consequences. Child Devel- qualitative methods in language research and interven- opment, 72, 685–687. tion. American Journal of Speech-Language Pathol- Astington, J. W. (2005). Introduction: Why language mat- ogy, 12, 165–171. ters. In J. W. Astington & J. A. Baird (Eds.), Why lan- Carpendale, J., & Chandler, M. J. (1996). On the distinc- guage matters for theory of mind (pp. 3–25). New tion between false belief understand and subscribing York: Oxford University Press. to an interpretive theory of mind. Child Development, Baron-Cohen, S. (1995). Mindblindness: An essay on 67, 1686–1706. autism and theory of mind. Cambridge, MA: MIT Carpendale, J., & Lewis, C. (2006). How children develop Press. social understanding. Malden, MA: Blackwell. Baron-Cohen, S. (2000). Theory of mind in autism: A Carpenter, J., & Lewis, C. (2006). How children develop fifteen year review. In S. Baron-Cohen, H. Tager- social understanding. Malden, MA: Blackwell. Flusberg, & D. J. Cohen (Eds.), Understanding other Chandler, M. J., & Hala, S. (1994). The role of personal in- minds: Perspectives from developmental cognitive volvement in the assessment of early false belief skills. neuroscience (pp. 3–20). New York: Oxford Univer- In C. Lewis & P. Mitchell (Eds.), Children’s early un- sity Press. derstanding of mind: Origins and development (pp. Baron-Cohen, S., Baldwin, D., & Crowson, M. (1997). Do 403–425). Mahwah, NJ: Erlbaum. children with autism use the speaker’s direction of Clements, W.A., Rustin, C. L., & McCallum, S. (2000). Pro- gaze strategy to crack the code of language? Child De- moting the transition from implicit to explicit under- velopment, 68, 48–57. standing: A training study of false belief. Developmen- Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does tal Science, 3(1), 81–92. the autistic child have a theory of mind? Cognition, Crais, E. R. (1993). Families and professionals as collabo- 21, 37–46. rators in assessment. Topics in Language Disorders, Baron-Cohen, S., Ring, H., Moriarty, J., Shmitz, P., Costa, 14(1), 29–40. D., & Ell, P.(1994). Recognition of mental state terms. Diehl, S. (2003). The SLP’s role in collaborative assess- British Journal of Psychiatry, 195, 640–649. ment and intervention for children with ASD. Topics Baron-Cohen, S., Wheelwright, S., Hill, J., Raste, Y., & in Language Disorders, 23(2), 95–115. Plumb, I. (2001). The “reading of the mind in the eyes” Dunn, Bretherton, & Munn. (1987). test revised version: A study with normal adults and Dunn, J., Brown, J., & Beardsall, L. (1991). Family talk adults with Asperger’s syndrome or high-functioning about feeling states and children’s later understanding autism. Journal of Child Psychology and Psychiatry, of others’ emotions. Developmental Psychology, 27, 42(2), 241–251. 448–455. Barr, R. (2006). Developing social understanding in a so- Flavell, J. H., Green, F. L., & Flavell, E. R. (1986). cial context. In K. McCartney & D. Phillips (Eds.), Development of knowledge about the appearance- Blackwell handbook of early childhood development reality distinction. Monographs of the Society for Re- (pp. 188–207). Malden, MA: Blackwell. search in Child Development, 51 (1, Serial No. 212). Bartsch, K., & Wellman, H. M. (1989). Young children’s Frith, U., Happe, F., & Siddons, F. (1994). Autism and Supporting Theory of Mind Development 361

theory of mind in everyday life. Social Development, psychometric evaluation of the Perceptions of Chil- 3, 108–124. dren’s Theory of Mind Measure—Experimental Ver- Gray, C. (1994a). The new social storybook. Arlington, sion (PCToMM-E). Journal of Autism and Develop- VA: Future Horizons. mental Disabilities, 38(1), 143–155. Gray, C. (1994b). Comic strip conversations. Jenison, MI: Hutchins, T. L., Bond, L. A., Silliman, E. R., & Bryant, J. The Morning News. (in press). Maternal epistemological perspectives and Gray, C. (1995). Teaching children with autism to read so- variation in mental state talk. Journal of Speech, Lan- cial situations. In K. A. Quill (Ed.), Teaching children guage, and Hearing Research. with autism: Strategies to enhance communication Hutchins, T. L., & Prelock, P. A. (2006). Using social sto- and socialization (pp. 219–242). New York: Delmar. ries and comic strip conversations to promote socially Gray, C. (1998). Social stories and comic strip conversa- valid outcomes for children with ASD. Seminars in tions with students with Asperger syndrome and high- Speech and Language,27(1), 47–59. functioning autism. In E. Schopler (Ed.), Asperger syn- Hutchins, T. L., & Prelock, P. A. (2008). The effects of drome or high-functioning autism? (pp. 167–194). social stories for remediating the behavioral, com- New York: Plenum Press. municative, and theory of mind deficits charac- Gray, C., & Garand, J. D. (1993). Social stories: Improving teristic of autism spectrum disorder. Manuscript in responses of students with autism with accurate social preparation. information. Focus on Autistic Behavior, 8, 1–10. Hutchins, T. L., Prelock, P.A., & Chace, W.(in press). Test- Greenspan, S. I., & Wieder, S. (1998). The child with spe- retest reliability of theory of mind tasks representing cial needs: Encouraging intellectual and emotional a range of content and complexity and adapted to fa- growth. Reading, MA: Addison-Wesley. cilitate the performance of children with autism spec- Guajardo, N. R., & Watson, A. C. (2002). Narrative dis- trum disorder. Focus on Autism and other Develop- course and theory of mind development. The Journal mental Disabilities. of Genetic Psychology, 163(3), 305–325. Ingersoll, B., Dvortcsak, A., Whalen, C., & Sikora, Gutstein, S. E., & Sheely, R. K. (2002a). Relationship de- D. (2005). The effects of a developmental, social- velopment intervention with children, adolescents pragmatic language intervention on rate of expressive & adults: Social and emotional development activ- language production in young children with ASD. Fo- ities for Asperger syndrome, autism, PDD & NLD. cus on Autism and Other Developmental Disabili- Philadelphia: Jessica Kingsley. ties, 20(4), 213–222. Gutstein, S. E., & Sheely, R. K. (2002b). Relationship de- Kazdin, A. E. (1977). Assessing the clinical or applied sig- velopment intervention with young children: Social nificance of behavioral change through validation. Be- and emotional development activities for Asperger havior Modification, 1, 427–452. syndrome, autism, PDD & NLD. Philadelphia: Jessica Kazdin, A. E. (1999). The meanings and measurement of Kingsley. clinical significance. Journal of Consulting and Clin- Hadwin, J., Baron-Cohen, S., Howlin, P., & Hill. K. (1996). ical Psychology, 67, 332–339. Can we teach children with autism to understand Klin, A. (2000). Attributing social meaning to ambigu- emotions, belief, or pretence? Development & Psy- ous visual stimuli in higher-functioning autism and chopathology, 8, 345–365. Asperger syndrome: The social attribution task. Jour- Happe, F.(1994). An advanced test of theory of mind: Un- nal of Child Psychology and Psychiatry, 7, 831– derstanding the story characters thoughts and feelings 846. by able autistic mentally handicapped and normal chil- Koegel, L. K., & Keogel, R. L. (2006). Pivotal response dren and adults. Journal of Autism & Developmental treatments: Communication, social and academic Disorders, 24, 129–154. development. Baltimore: Paul H. Brookes. Happe, F. (1995). The role of age and verbal ability in Koegel, L. K., Koegel, R. L., & Carter, C. M. (1998). Pivotal the theory of mind task performance of subjects with responses and the natural language paradigm. Semi- autism. Child Development, 66, 843–855. nars in Speech & Language, 19, 355–372. Happe, F., & Winner, E. (1998). The getting of wisdom: Kuttler, S., Myles, B., & Carlson, J. K. (1998). The use of so- Theory of mind in old age. Developmental Psychol- cial stories to reduce precursors to tantrum behavior ogy, 43(2), 358–362. in a student with autism. Focus on Autism and Other Harris, P., Johnson, C. N., Hutton, D., Andrews, G., & Developmental Disabilities, 13, 176–182. Cooke, T. (1989). Young children’s theory of mind and Lalonde, C. E., & Chandler, M. J. (2002). Children’s under- emotion. Cognition and Emotion, 3, 370–400. standing of interpretation. New Ideas in Psychology, Holroyd, S., & Baron-Cohen, S. (1993). How far can peo- 20, 163–198. ple with autism go in developing a theory of mind? Lewis, C., Freeman, N. H., Kyriakidou, C., Maridaki- Journal of Autism and Developmental Disorders, 23, Kassotaki, K., & Berridge, D. M. (1996). Social 379–385. influences on false belief access: Specific sibling influ- Hutchins, T. L., Bonazinga, L. A., Prelock, P. A., & Taylor, ences or general apprenticeship? Child Development, R. (2008). Beyond false beliefs: The development and 67, 2930–2947. 362 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008

Lovaas, O. I. (1987). Behavioral treatment and normal ed- Pillow, B. H. (1991). Children’s understanding of bi- ucational and intellectual functioning in young autistic ased social cognition. Developmental Psychology, 17, children. Journal of Consulting and Clinical Psychol- 263–276. ogy, 55, 3–9. Pratt, C., & Bryant, P. (1990). Young children understand McCauley, R. (2001). Assessment of language disorders that looking leads to knowing (so long as they are look- in children. Mahwah, NJ: Erlbaum. ing into a dingle barrel). Child Development, 61(4), McGhee, P. E. (1979). Humor: Its origin and develop- 973–982. ment. San Francisco: W. H. Freeman. Prelock, P. A. (2006). Autism spectrum disorders: Issues McGregor, E., Whiten, A., & Blackburn, P. (1998). Teach- in assessment & intervention. Austin, TX: Pro-Ed. ing theory of mind by highlighting intention and illus- Prelock, P. A., & Hutchins, T. L. (2008). The role of trating thoughts: A compassion of their effectiveness family-centered care in research: Supporting the so- with 3 year olds and autistic individuals. British Jour- cial communication of children with autism spectrum nal of Developmental Psychology, 16, 281–300. disorder. Topics in Language Disorders, 28(4), 323– Miller, C. A. (2006). Developmental relationships be- 339. tween language and theory of mind. American Jour- Prior, M., Dahlstrom, B., & Squires, T. L. (1990). Autistic nal of Speech-Language Pathology, 15, 142–154. children’s knowledge of thinking and feeling in other Muris, P., Steerneman, P., Meesters, C., Merckelbach, H., people. Journal of Child Psychology & Psychiatry, Horselenberg, R., van den Hogen, T., et al. (1999). The 31, 587–601. TOM test: A new instrument for assessing theory of Prizant, B. M., Wetherby, A. M., Rubin, E., Laurent, A. C., mind in normal children and children with pervasive & Rydell, P. (2004). The SCERTS model: Enhancing developmental disorders. Journal of Autism and De- communication and socioemotional abilities of chil- velopmental Disorders, 29, 67–80. dren with autism spectrum disorders. Port Chester, National Research Council. Committee on Educational In- NY: National Professional Resources Inc. terventions for Children with Autism. Division of Be- Repacholi, B. M., & Gopnik, A. (1997). Early reasoning havioral and Social Sciences and Education. (2001). about desires: Evidence from 14- and 18-month-olds. Educating Children With Autism. Washington, DC: Developmental Psychology, 33, 12–21. National Academy Press. Sodian, B., Taylor, C., Harris, P., & Perner, J. (1992). Early Norris, C., & Dattilo, J. (1999). Evaluating effects of a so- deception and the child’s theory of mind: False trails cial story intervention on a young girl with autism. Fo- and genuine markers. Child Development, 62, 468– cus on Autism and Other Developmental Disabili- 483. ties, 14(3), 180–186. Steele, S., Jospeh, R. M., & Tager-Flusberg, H. (2003). Brief Ozonoff, S., & McEvoy, R. E. (1994). A longitudinal study report: Developmental change in theory of mind abil- of executive function and theory of mind develop- ities in children with autism. Journal of Autism and ment in autism. Development and Psychopathology, Developmental Disorders, 33(4), 461–467. 6, 415–431. Sullivan, K., Zaitchik, D., & Tager-Flusberg, H. (1994). Ozonoff, S., & Miller, J. N. (1995). Teaching theory of Preschoolers can attribute second-order beliefs. De- mind: A new approach to social skills training for indi- velopmental Psychology, 30, 395–402. viduals with autism. Journal of Autism and Develop- Swaggart. B., Gagnon, E., Bock, S., & Earles, T. (1995). mental Disorders, 25(4), 415–433. Using social stories to teach social and behavioral skills Perner, J., Frith, U., Leslie, A. M., & Leekam, S. R. (1989). to children with autism. Focus on Autistic Behavior, Exploration of the autistic child’s theory of mind: 10, 1–16. Knowledge, belief, and communication. Child Devel- Swettenham, J. G., Baron-Cohen, S., Gomez, J. C., & opment, 60, 689–700. Walsh, S. (1996). What’s inside someone’s head? Con- Perner, J., Leekham, S. R., & Wimmer, H. (1987). 3-year- ceiving of the mind as a camera helps children with old’s difficulty with false belief: The case for a con- autism acquire an alternative to a theory of mind. Cog- ceptual deficit. British Journal of Developmental Psy- nitive Neuropsychiatry, 1 (1), 73–88. chology, 5, 125–137. Tager-Flusberg, H. (1999). The challenge of studying Perner, J., & Wimmer, H. (1985). “John thinks that Mary language development in children with autism. In thinks that...”: Attribution of second-order beliefs by L. Menn, & N. Bernstein Ratner (Eds.), Methods 5- and 10-year-old children. Journal of Experimental for studying language production (pp. 313–331). Child Psychology, 39, 427–471. Mahwah, NJ: Erlbaum. Peskin, J., & Astington, J. W. (2004). The effects of adding Tager-Flusberg, H. (2000). Language and understanding: metacognitive language to story contexts. Cognitive Connections in autism. In S. Baron-Cohen, H. Tager- Development, 19, 253–273. Flusberg, & D. J. Cohen (Eds.), Understanding other Phillips, W., Baron-Cohen, S., & Rutter, M. (1998). Un- minds: Perspectives from developmental cognitive derstanding intention in normal development and in neuroscience (pp. 124–149). New York: Oxford Uni- autism. British Journal of Developmental Psychol- versity Press. ogy, 16, 337–348. Tager-Flusberg, H. (2001). A reexamination of the Supporting Theory of Mind Development 363

theory of mind hypothesis of Autism. In J. Burack, T. bubbles help children with autism acquire an alterna- Charman, N. Yirmiya, & P. R. Zelazo (Eds.), The de- tive to theory of mind. Autism, 6(4), 343–363. velopment of autism: Perspectives from theory and Wellman, H. M., & Bartsch, K. (1988). Young children’s research (pp. 173–193). Mahwah, NJ: Erlbaum. reasoning about beliefs. Cognition, 30, 239–277. Tomasello, M. (1995). Joint attention as social cognition. Wellman, H. M., Cross, D., & Watson, J. (2001). Meta- In C. Moore & P. J. Dunham (Eds.), Joint attention: analysis of theory-of-mind development: The truth It’s origins and role in development (pp. 103–130). about false beliefs. Child Development, 72, 655–684. Hillsdale, NJ: Erlbaum. Wellman, H. M., & Estes, D. (1986). Early understanding of Tomasello, M. (2003). Constructing a language. Cam- mental entities: A reexamination of childhood realism. bridge, MA: Harvard University Press. Child Development, 57, 910–923. Watson, L. R., Lord, C., Schaffer, B., & Schopler, E. Wimmer, H., & Perner, J. (1983). Beliefs about be- (1989). Teaching Spontaneous Communication to liefs: Representation and the constraining function of Autistic and Developmentally Handicapped Chil- wrong beliefs in young children’s understanding of de- dren (TEACCH). NY: Irvington. ception. Cognition, 13, 103–128. Wellman, H. M., Baron-Cohen, S., Caswell, R., Gomez, J. Yirmiya, N., Solmonica-Levi, D., Shulman, C., & Pilowsky, C., Swettenham, J., Toye, E., et al. (2003). Thought- T. (1996). Theory of mind abilities in individuals with bubbles help children with autism acquire an alterna- autism, down syndrome, and mental retardation of un- tive to a theory of mind. Autism, 6(4), 343–363. known etiology: The role of age and intelligence. Jour- Wellman, H. M., Baron-Cohen, S., Caswell, R., Gomez, J. nal of Child Psychology and Psychiatry and Allied C., Swettenham, J., Toye, E., et al. (2002). Thought- Disciplines, 37(8), 1003–1014. 364 TOPICS IN LANGUAGE DISORDERS/OCTOBER–DECEMBER 2008

Appendix

Zach’s social story “When things don’t go how I think they will”

It makes me feel good when I know what is going to happen. [perspective] But things don’t always go the way I think they will. [perspective] Sometimes things surprise me like when the fire alarm at school goes off. [perspective] Sometimes I get disappointed because someone told me one thing, and we do something else. [perspective] I can get upset when there is a change of plans and my routine gets all messed up. [perspective] This can make me scared because I don’t know what is going to happen. [perspective] It can also make me upset. [perspective] I might throw a fit and hit, kick, bite, scream, or break things. [descriptive] It makes my mom and dad sad when I throw a fit. [perspective] They might think “I wish we could make Zach happy”or “I wish he would talk to us about what he was thinking and feeling. ”[thinking] They might also think “Zach is not acting very nice right now” or “I wish he wouldn’t throw a fit.”[thinking] When things don’t go the way I think they will—that’s okay and I’m okay. [perspective] If I get scared or mad or disappointed, I can use my words to talk to my mom and dad. [directive] I can calmly say “I’m not happy”“I don’t like this”or “I need a break.”[directive] This makes my mom and dad happy. [perspective] They will think “I’m so proud of Zach”and “I’m so happy that he is talking to us about how he thinks and feels.”[thinking] They might even give me a “high five.”[descriptive] It’s good to talk to my mom and dad about the things that make me sad, mad, or scared. [perspective] When I talk to them, they can help me think about things. [descriptive] It’s okay when things don’t go exactly how I think they will. [perspective] And when they don’t, I will try to remember to talk calmly about how I feel. [directive]