Verbal Behavior and Autism Intervention Mark L Sundberg

Verbal Behavior and Autism Intervention Mark L. Sundberg, Ph. D. , BCBA marksundberg@astound. net

Introduction • • It was suggested by Dr. Gina Green (2005) at last year’s ABA convention that language training procedures for children with autism that are based on Skinner’s (1957) analysis of verbal behavior should not be disseminated until data supporting those procedures are obtained The current presentation is a response to Dr. Green’s concerns about the use of Skinner’s analysis of verbal behavior for language assessment and intervention for children with autism

Introduction • • • Behavioral interventions in general have been quite successful for children with autism With this success comes good news and bad news The good news. . . Increase in the demand for behavioral services Increase in the recognition of behavior analysis Increase in the demand for BCABAs & BCBAs Increase in the demand for training and university courses Increase in the demand for teachers of behavior analysis Increase in research opportunities Increase in ABA: International membership

Membership in ABA: International

Introduction • • However, with this success there is some bad news Everybody is now an expert in behavior analysis Widespread dissemination of behavioral techniques and services, often by unqualified people Simplifying the concepts and procedures beyond recognition Common to see procedures promoted without the analysis Parallels to B-mod and education in the 60 s? “What happened to the promise of behavior modification? ” Similar concerns for the current popularity of the use of behavior analysis for the treatment of children with autism

Introduction • • • • What constitutes a “behavioral approach” to treatment of children with autism? Consumers must be confused. There are many models out there, often quite different from each other, but all claiming to be a “behavioral approach” DTT Lovaas model CARD model ABA Pivotal response training VB approach CABAS Competent learner model Natural language paradigm Milieu language training Incidental teaching

Introduction • • In addition, there are many other approaches and treatments such as. . . Floortime RDI Son-Rise Holding therapy TEACCH Secretion Auditory training • • • Sensory integration Weighted jackets Deep pressure Special diets Vitamins Medications Swimming with dolphins Decompression chambers Chelation Facilitated communication

Introduction • • Thanks to the efforts of Dr. Gina Green and others, many of these “pseudoscientific” approaches have been identified and consumers have been warned about their ineffectiveness and even potential danger to children However, at least year’s ABA convention Dr. Green added the “Verbal Behavior Approach” to her list of pseudoscientific approaches and suggested that it had similarities to facilitated communication In her recent presentation at the NY-ABA titled “Verbal Behavior; ” An evidence-based technology for autism intervention? ” Dr. Green (2005) concluded “the ‘VB’ approach to autism intervention does not appear to meet accepted criteria for evidence-based practice or transferable behavioral technology” Others have expresses concern about the dissemination of verbal behavior procedures. Carr and Firth (2005) stated “little research exists to support such widespread dissemination (of the VB approach)”

Introduction • • The purpose of the current presentation is to address the concerns raised by Dr. Green, Carr and Firth, and others about the “unwarranted dissemination of the verbal behavior approach” to language training for children with autism The goal is to clarify what constitutes a “verbal behavior approach, ” while demonstrating its empirical foundation, on-going research agenda, its value to children with autism, and hopefully, to get Dr. Green to remove verbal behavior from her list of pseudosciences

What Constitutes a Verbal Behavior Approach to Autism Treatment? • • First, I share Dr. Green’s concern for the need for additional verbal behavior research. I believe this point is uncontroversial. (Sundberg, 1991: “ 301 Research topics from Skinner’s book Verbal Behavior”) Verbal behavior research was the primary purpose for starting the journal The Analysis of Verbal Behavior, now in its 22 nd Volume, and published by ABA: International Second, I strongly share Dr. Green’s concern about the improper dissemination of behavioral concepts and procedures And, her concerns about the ubiquitous dissemination of pseudoscientific treatments for children with autism

What Constitutes a Verbal Behavior Approach to Autism Treatment? • • • The basic teaching procedures consist of the standard methodology found in applied behavior analysis (e. g. , Cooper, Heron, & Heward, 1987) Prompting Fading Pairing Modeling Shaping Chaining Differential reinforcement procedures (e. g. , DRO, DRI, DRL) Intermittent reinforcement procedures (e. g. , FR, VR, FI, VI)

What Constitutes a Verbal Behavior Approach to Autism Treatment? • • • Extinction procedures (e. g. , planned ignoring) Punishment procedures (e. g. , reprimands) Generalization Discrimination training Errorless learning Transfer of stimulus control Task analysis Fluency procedures Contingency contracting Token economies

What Constitutes a Verbal Behavior Approach to Autism Treatment? • • • Additional procedural elements include, for example. . Individualized assessment and intervention program Frequent opportunities to respond Use of discrete trial teaching procedures Incidental & natural environment teaching procedures Data collection Interspersel techniques Behavioral momentum techniques Peer and social interaction Functional analyses On-going analyses of performance by formally trained behavior analysts

What Constitutes a Verbal Behavior Approach to Autism Treatment? • • These procedures are (to varying degrees) common to most behavioral intervention programs for children with autism (e. g. , Greer & Keohane, 2006: Howard, Sparkman, Cohen, Green, & Stanislaw, 2005; Koegel & Koegel, 1996; Leaf & Mc. Eachin, 1999; Lovaas, 2003; Maurice, Green, & Luce, 1996: Sundberg & Partington, 1998), and thus all benefit from the same empirical foundation found in applied behavior analysis However, these programs vary substantially in terms their treatment of language

What Constitutes a Verbal Behavior Approach to Autism Treatment? • • The major difference between the verbal behavior programs and the majority of discrete trial (DTT) and ABA programs available in the literature is the conceptual analysis of language that underlies the assessment and curriculum used in each program How is language measured, classified, and assessed? What is the unit of analysis? What causes the emission of words and sentences? How is language acquired? What causes language errors and deficits? Most DTT/ABA programs are based on the traditional linguistic classification system of expressive and receptive language, and the associated vernacular, concepts, and theoretical constructs related to language, which has its roots in cognitive psychology The verbal behavior approach employs Skinner’s (1957) functional analysis of language, which has its roots in radical behaviorism

Skinner’s Analysis of Verbal Behavior • • • Language is learned behavior under the functional control of environmental contingencies “What happens when a man speaks or responds to speech is clearly a question about human behavior and hence a question to be answered with the concepts and techniques of psychology as an experimental science of behavior” (Skinner, 1957, p. 5) The analysis of verbal behavior involves the same behavioral principles and concepts that make up the analysis of nonverbal behavior. No new principles of behavior are required Chapter 1 of Verbal Behavior is titled “A Functional Analysis of Verbal Behavior” In Chapter 2 Skinner identifies the dependent and independent variables for a functional analysis of verbal behavior

A Functional Analysis of Verbal Behavior: The Basic Principles of Operant Behavior Stimulus Control (SD) Motivating Operation (MO/EO) Response Reinforcement Punishment Extinction Conditioned reinforcement Conditioned punishment Intermittent reinforcement

Skinner’s Analysis of Verbal Behavior • • • The traditional linguistic classification of words, sentences, and phrases as expressive and receptive language blends important functional distinctions between types of operant behavior, and appeals to cognitive explanations for the causes of language behavior (Skinner, 1957, Chapter 1) Thus, in Chapter 1 of VB Skinner recommends against what has become the linguistic foundation of most DTT and ABA programs While there are many conceptual and practical distinctions between a cognitive and behavior analysis of language, this presentation will focus on: Research on the distinction between the mand, tact, and intraverbal A functional analysis of verbal assessment and intervention

Skinner’s Analysis of Verbal Behavior • • At the core of Skinner’s analysis of language is the distinction between the mand, tact, and intraverbal (traditionally all classified as “expressive language”) Is there conceptual and empirical support for this distinction? Skinner identified three separate sources of antecedent control for these verbal operants EO/MO control------->Mand Nonverbal SD----->Tact Verbal SD------->Intraverbal In cognitive analyses of language these three sources are commonly grouped together under the rubric of “referent or meaning”

Skinner’s Analysis of Verbal Behavior • • The empirical question is: Are these three antecedent variables functionally separate, or is there no value in making this distinction? From a clinical standpoint, the two most common language problems demonstrated by children with autism that I have encountered over the past 32 years is a defective mand repertoire and/or a defective intraverbal repertoire, despite often having strong tact and listener discrimination repertoires

Charlie: Quick Assessment

Nathan: Quick Assessment

Matt: Quick Assessment

The Distinction Between the Mand the Tact • • Based on the distinction between the establishing operation (EO/MO) and stimulus control (SD) as separate sources of control Skinnerian psychology (“radical behaviorism, ” see Skinner, 1974) has always maintained that motivational control is different from stimulus control In Behavior of Organisms (Skinner, 1938) Skinner devoted two chapters to the treatment of motivation; Chapter 9 titled “Drive” and Chapter 10 titled “Drive and Conditioning: The Interaction of Two Variables” Skinner also made it clear in the section titled “Drive (is) Not a Stimulus” (pp. 374 -376) that motivation is not the same as discriminative, unconditioned, or conditioned stimuli

The Distinction Between the Mand the Tact • • • Keller and Schoenfeld (1950) titled Chapter 9 “Motivation” and further developed Skinner’s point, “A drive is not a stimulus” (p. 276), and suggested “a new descriptive term. . . ‘establishing operation’” (p. 271) In Science and Human Behavior (1953) Skinner devoted three chapters to motivation: Chapter 9: “Deprivation and Satiation, ” Chapter 10: “Emotion, ” and Chapter 11: “Aversion, Avoidance, Anxiety” In Verbal Behavior (1957) Skinner had a full chapter on motivation and language (The Mand), and throughout the book provided many elaborations on motivational control -- as an antecedent variable

The Distinction Between the Mand the Tact • • • Holland Skinner’s (1961) book contained four chapters on motivation; Chapters 7: “Deprivation, ” 8: “Emotion I, ” 9: “Avoidance and Escape Behavior, ” and 10: “Emotion II” Millenson (1968) contained four chapters on motivation and presented an excellent summary of the relevant empirical research (p. 364 -384); Chapters 15: “Motivation I, ” 16: “Motivation II, ” 17: “Aversive Contingencies, ” and 18: “Emotional Behavior” However, the topic of motivation was for the most part dropped from the first generation of Applied Behavior Analysis/Behavior Modification textbooks that followed Millenson’s book (e. g. , Fantino & Logan, 1979; Kazden, 1975; Martin & Pear, 1978; Powers & Osborne, 1976; Whaley & Malott, 1971)

The Distinction Between the Mand the Tact • • In explaining what happen to the analysis of motivation in behavior analysis, Michael (1993) pointed out, “In applied behavior analysis or behavior modification, the concept of reinforcement seems to have taken over much of the subject matter that was once considered a part of the topic of motivation” (p. 191) There was a shift from the analysis of motivation as an antecedent variable to motivation as a consequence In addition, motivation as a topic of research was absent from the behavioral journals. For example, The Journal of Applied Behavior Analysis contained no entries for “establishing operations” or “motivation” in the cumulative index (1978) covering the first 10 years of publication During the next 10 years (1979 -1988) there were still no entries for “establishing operation. ” However, there were 5 entries for “motivation, ” but they all involved the use of motivation as a consequence, rather than as an antecedent variable

The Distinction Between the Mand the Tact • • Motivation as an antecedent variable has returned to behavior analysis textbooks and now is a common topic in JABA thanks to Jack Michael, Brian Iwata, Wayne Fisher, and others The JABA index for the years 1999 -2005 contains 29 entries for the EO, and 2 for the MO (motivating operations) Malott, Whaley, & Malott (1997) contains a full chapter on the EO. Catania (1998), Martin & Paer (2002), and Pierce & Epling (1995) all contain analyses of motivation throughout their books The 2 nd Edition of the Cooper, Heron, & Heward book Applied Behavior Analysis (In press) contains a full chapter on motivation as well as a full chapter on Skinner’s analysis of verbal behavior

Clinical Value of EO and the Mand to Children with autism • • Many children with autism have absent or defective mand repertoires Manding is a critical part of a typical child’s language development Tact training does not typically produce manding in early learners A functional analysis of the child’s verbal behavior often reveals that the response called a mand or a request is not under EO control, but rather SD control, thus not, by definition a mand

Defective Mand - Ally EO Does not evoke a mand ___________________________ EO Does not evoke a mand Object ___________________________ Intraverbal prompt (e. g. “Sign cookie”) Evokes a response Imitative prompt (not a mand) (ASL sign) __________________________________

A Sample of the Published Empirical Research on the EO, and the Mand Tact • • • Hung (1980) Simic & Bucher (1980) Lamarre & Holland (1985) Pierce, Epling, & Boer (1986) Hall & Sundberg (1987) Carroll & Hesse, (1987) Stafford, Sundberg, & Braam (1988) Yamamoto & Mochizuki (1988) Mc. Pherson & Osborne (1988) De Freitas & Ribeiro (1989) Sigafoos, Doss, & Reichle (1989) Sundberg, San Juan, Dawdy, & Arguelles (1990)

A Sample of the Published Empirical Research on the EO, and the Mand Tact • • • Sigafoos, Reichle, Doss, Hall, & Pettitt (1990) Baer & Detrich (1990) Braam & Sundberg (1991) Sprague & Horner (1992) Williams & Greer (1993) Twyman (1996) Drasgow, Halle, & Ostrosky (1998) Drash, High, & Tudor (1999) Brown, Wacker, Derby, Peck, Richman, & Sasso (2000) Knutson & Harding (2000) Barnes-Holmes & Barnes-Holmes (2000) Goh, Iwata, & De. Leon (2000)

A Sample of the Published Empirical Research on the EO, and the Mand Tact • • • Sundberg, Loeb, Hale, & Eigenheer (2002) Arntzen & Almas (2002) Ewing, Magee, & Ellis (2002) Winborn, Wacker, Richman, Asmus, & Geier (2002) Chambers & Rehfeldt (2003) Ross & Greer (2003) Nuzzolo-Gomez & Greer (2004) Taylor, Hoch, Potter, Rodriguez, Spinnato, & Kalaigian (2005) Petursdottir, Carr, & Michael (2005) Hartman & Klatt (2005) Wallace, Iwata, & Hanley (2006) Sweeney, Carbone, O’Brien, Zecchin, & Janecky (2006)

A Sample of the Published Empirical Research on the EO, and the Mand Tact • • • Taylor, Hoch, Potter, Rodriguez, Spinnato, & Kalaigian (2005) EO must be present to evoke mands (initiations to peers) Petursdottir, Carr, & Michael (2005) Mand training resulted in tacts, but tact training did not result in mands There doesn’t appear to be a body of research that contradicts the separation of the mand tact at the time of initial acquisition, or manding without EOs Empirical research reviews: Greer, R. D. , & Keohane D. (2005). The evolution of verbal behavior in children. Behavioral Development, 1, 31 -48. Oah, S. , & Dickinson, A. M. (1989). A review of empirical studies on verbal behavior. The Analysis of Verbal Behavior, 7, 53 -68. Sautter, R. , & Le. Blanc, L. (2006). The empirical applications of Skinner’s analysis of verbal behavior with humans. The Analysis of Verbal Behavior.

A Sample of the Published Empirical Research on the EO, and the Mand Tact • • • EO and Mand research across a variety of populations with similar effects Children with autism (e. g. , Ross & Greer, 2003) Language delayed children (e. g. , Twyman, 1996) Typical children (e. g. , Petursdottir, Carr, & Michael, 2005) Deaf/autistic teenagers e. g. , (Hall & Sundberg, 1987) Children with developmental disabilities (e. g. , Sigafoos, Doss, & Reichle, 1989) Adults with developmental disabilities (e. g. , Chambers & Rehfeldt, 2003) Adults with traumatic brain injury (e. g. , Sundberg, San Juan, Dawdy, & Arguelles, 1990) Pigeons (e. g. , Sundberg, 1985) Rats (e. g. , Pierce, Epling, & Boer, 1986) Chimpanzees (e. g. , Savage-Rumbaugh, 1984) Parrots (e. g. , Pepperberg, 1988)

A Sample of the Published Empirical Research on the EO, and the Mand Tact • • • EO and Mand research and reviews across a variety of behavioral journals JEAB (e. g. , Lamarre & Holland, 1985) JABA (e. g. , Wallace, Iwata, & Hanley, 2006) TAVB (e. g. , Petursdottir, Carr, & Michael, 2005) TBA (e. g. , Michael, 1993) Behavioral Development (e. g. , Greer & Koehane, 2005) Research in Developmental Disabilities (e. g. , Taylor, Hoch, et al. , 2005) Journal of Autism and Developmental Disabilities (e. g. , Hung, 1980) Behavior Modification (e. g. , Rogers-Warren & Warren, 1980) Journal of Early and Intensive Behavior Intervention (e. g. , Pistoljevic & Greer, 2006)

EO Antecedent Control and Mand Assessment and Intervention in DTT and ABA Programs • • Lovaas, 1977, 1981, 2003 (clearly the most influential, outcome data) Expressive-receptive framework for language Terminology and analysis derived from traditional linguistics No mention of EO/motivation antecedent control All language skills are presented as SD control Closest mand training activity is the “I Want___” program found late in the program (between the adjective and preposition chapters). “Requesting” does not appear in the 2003 index No focus on the fact that a single word, phrase, or sentence can be strong in one repertoire and not another No functional analysis of words as behavior

How ABA is perceived in the press: Time Magazine, May 15, 2006 • • Biased presentation of behavior analysis and an excellent behavioral program that is clearly guided by behavioral principles, including the analysis of VB, EOs, etc. Floortime presented with a very favorable bias, despite the acknowledged lack of data. “Greenspan…is responding with a series of studies just getting under way” What does Floortime have that the reporter seemed so impressed with? “David, 6, goes down a slide again and again. A teacher playfully blocks his way, She wants him to. . . say “Move. ” She's got an agenda; he doesn't know it. He keeps going back for more because it's fun. . Knowing to ask. . . is part of learning to regulate oneself. ” Manding!!! We have the analysis and data. Floortime gets the credit How about natural environment teaching, making learning fun, and pairing? Again, we have the analysis and data, Floortime gets the credit DTT/ABA cannot afford to wait any longer, or it will go the route of education in the 60 s

The Distinction Between the Tact and the Intraverbal • • • A substantial number of children with autism have extensive tact and RD repertoires, but a weak, absent, or defective intraverbal repertoire The existing body of research supports the conceptual analysis that a response acquired under nonverbal stimulus control may not automatically transfer to verbal stimulus control For example, a child may be able to say Pool when he sees a swimming pool (Nonverbal SD), but not say Pool when asked Where do you go swimming? (Verbal SD)

Empirical Research on the Distinction Between the Tact and Intraverbal • • • Braam & Poling (1983) Chase, Johnson, & Sulzer-Azaroff (1985) Luciano (1986) Daly (1987) Lodhi & Greer (1989) Tenenbaum & Wolking (1989) Watkins, Pack-Teixeira, & Howard (1989) Sundberg, San Juan, Dawdy, & Arguelles (1990) Partington & Bailey (1993) Sundberg, Endicott, & Eigenheer (2000) Finkel & Williams (2001) Miguel, Petursdottir, & Carr (2005)

Empirical Research on the Distinction Between the Tact and Intraverbal • • Two examples of research Braam & Poling (1983) found that children with autism who could emit specific responses under tact control could not emit the same response forms under intraverbal control. Transfer of stimulus control between nonverbal SDs and verbal stimuli were successful Miguel, Petursdottir, & Carr (2005) replicated the basic procedures from Braam and Poling (1983) and concluded “ while participants were able to tact…and point to the pictures…they were not necessarily able to reliably produce thematically related intraverbal responses…(until) intraverbal training was used” No body of research has emerged to show the tact and intraverbal are the same

Clinical Value of Verbal Stimulus Control and the Intraverbal to Children with Autism • • • Many children with autism have absent or defective intraverbal repertoires Verbal behavior evoked by verbal discriminative stimuli constitute a significant element of human verbal interaction A functional analysis of the child with autism’s verbal behavior often reveals that the response called intraverbal or conversational is not under verbal stimulus control, but rather under nonverbal stimulus control, or EO control, thus by definition, not an intraverbal Many common verbal errors by children with autism are related to defective verbal stimulus control Verbal stimulus control is extremely complicated, usually involving verbal conditional discriminations where one verbal stimulus alters the evocative effect of another verbal stimulus (e. g. , What do you wear to the beach? vs. What do you take to the beach? )

Verbal Stimulus Control and Verbal Conditional Discriminations in ABA • • Research on verbal stimulus control, conditional discriminations, and especially verbal conditional discriminations is absent from the applied journals A review of the JABA indexes from 1968 to 2005 shows only 5 entries for conditional discriminations, and none for verbal conditional discriminations Thus, like the EO and the mand, the antecedent variables that evoke intraverbal behavior have not been a focus of behavioral research However, there is no argument that much of our verbal behavior is controlled by verbal SDs (e. g. , answering questions)

Verbal Stimulus Control and Intraverbal Assessment and Intervention in DTT and ABA Programs • • • Lovaas (1977, 1981, 2003) No sections on answering questions, fill-in’s, verbal categories, etc. or what could be identified as intraverbal training No mention of verbal antecedent control of verbal behavior (all under receptive language) No mention of verbal conditional discriminations Cognitive theory of words, referents, and meanings drives the language curriculum

Summary of the Empirical Research Supporting Verbal Behavior • • • Sautter & Le. Blanc (2006) “The volume of empirical support for Skinner’s (1957) analysis of verbal behavior has increased three fold over the past 15 years. ” “To date, empirical investigations have provided initial support for. . . Skinner’s notion of functional independence, the importance of the mand as the preliminary focus of language training, the utility of transfer of stimulus control procedures the benefits of multiply controlled language in the acquisition and development of more complex verbal behavior the notion of the establishing operation as a critical controlling variable for the mand”

Summary of the Empirical Research Supporting Verbal Behavior • • Greer & Koehane (2005) “While a large portion of the literature on verbal behavior has been theoretical, we have identified over 88 experiments devoted to testing theory. In our program of research we have completed approximately 44 experiments and a number of replications” (Greer & Koehane, 2005, p. 32).

Summary of the Empirical Research Supporting Verbal Behavior • • Horne & Lowe (1996) “Naming” Horne, Hughes, & Lowe (2006) conclude that a “route for future research may be to concentrate directly on the extraordinary behavioral repertoires that we term language or verbal behavior and the issue of how it transforms human learning. Skinner (1957) had clearly recognized the importance of this behavior and had begun to establish a theoretical framework for language research. . The naming account. . . in conjunction with the present series of experimental tests, is an attempt to advance this theory and to establish a coherent empirical base for its further development” (p. 271)

Conclusion • • • There is a rapidly growing body of empirical research supporting various aspects of Skinner’s analysis of verbal behavior This conference has contained many excellent examples of verbal behavior research on a wide range of topics (including pick-up lines) There are no contradictory lines of research on the distinction between the mand, tact, and intraverbal The VB approach is just behavior analysis The “VB approach” shares the same procedures and methods as the other behavioral approaches, but is based on a functional rather than structural analysis of language

Conclusion • • Should we continue to disseminate verbal behavior procedures? Sundberg & Michael (2001) suggested five major contributions that Skinner’s analysis of verbal behavior could make to the existing DTT/ABA programs for children with autism: Mand training Motivation (EO) as an independent variable in language training Intraverbal training Automatic reinforcement A functional analysis of verbal responses, verbal errors, language assessment, and curriculum development

Conclusion • • • Is there enough empirical support for the dissemination of these suggestions? Mand training Yes Motivation (EO) as an independent variable Yes Intraverbal training Yes Automatic reinforcement No (but, see Tim Vollmer’s presentation at 11 am this morning) A functional analysis of verbal responses, verbal errors, language assessment, and curriculum development Yes

Conclusion • “The ‘VB approach’ is simply normative applied behavior analysis with a few refinements. That is, it incorporates all of the standard methodology of applied behavior analysis, but it explicitly adopts Skinner's interpretive framework for analyzing verbal contingencies. In other words, it is a small variation on a methodology that has an enormous empirical foundation. The worst-case scenario is that the added framework doesn't help. But even in that case the child is still getting a full-fledged program of applied behavior analysis procedures. It is simply hard to believe that a set of procedures guided only by a distinction between receptive and expressive language can be as sharp as one that respects all of the various types of contingencies analyzed by Skinner” (Palmer, 2005)

Conclusions • • Recommendations Continue to conduct experimental and applied research on verbal behavior (Carr & Firth, 2005; Sautter & Le. Blanc, 2006; Sundberg, 1991) Conduct case histories Obtain more VB outcome data

Conclusion • • • In 1978 B. F. Skinner wrote … “Verbal Behavior…will, I believe, prove to be my most important work” (p. 122) Let’s get on with the proving!

Thank You! For an electronic version of this presentation email: marksundberg@astound. net