Selecting Response Form for Nonverbal Persons: Facilitated … · 2019. 5. 22. · verbal children, especially for those who have very little vocal behavior under echoic control

The Analysis of Verbal Behavior 1993, 11, 99-116

Selecting a Response Form for Nonverbal Persons:Facilitated Communication, Pointing Systems, or Sign Language?

Mark L. SundbergBehavior Analysts, Inc., Danville, CA

The three major types of augmentative communication for nonverbal persons consist of writing(or typing), pointing, and signing. These alternative response forms are examined in terms oftheir advantages and disadvantages for establishing effective verbal behavior. In addition, thesesystems are examined using the concepts from Skinner's (1957) analysis of verbal behavior (i.e.,mand, tact, intraverbal, and autocitic). The results of this analysis show that sign language hasthe most advantages and the fewest disadvantages, and more closely parallels speech in termsof the verbal operants. Although, the current trend is to favor facilitated communication (typing)and pointing systems, both of these response forms have several disadvantages that impede thedevelopment of the verbal operants. It is suggested that for many nonverbal individuals signlanguage is a better alternative response form, and has a better chance of improving speech.

The recent interest in facilitated commu-nication (FC), especially by the media, hasdrawn substantial attention to the lan-guage needs of nonverbal persons. How-ever, many of the issues concerning how tobest meet these needs remain unresolved.It is clear that many developmentallydisabled (DD) individuals with severe lan-guage disorders can benefit from sometype of augmentative communication (fora review, see Zangari, Lloyd, & Vicker, inpress). But questions as to which augmen-tative system might be the most effectivefor an individual, and how to best teachverbal behavior to nonverbal persons, havebecome more complicated by the entranceof FC into this already controversial arena.A close look at the different augmentativesystems using both structural and func-tional analyses may clarify some of theseissues.A critical element in establishing a lan-

guage intervention program for a nonver-bal person is selecting a response form.That is, what type of response topographyshould be used for developing the verbal

The author wishes to acknowledge Genae A. Hall,James W. Partington, Mary Ann Powers, and CindyA. Sundberg for their helpful comments on earlierversions of this paper. Reprints may be obtained fromthe author, Behavior Analysts, Inc., 888 Podva Road,Danville, CA 94526.

repertoires? There are four general options:(1) speech, (2) independent writing or typ-ing, or facilitated communication, (3)pointing and exchange systems (includingcomputer generated speech), and (4) signlanguage. There is an extensive body ofresearch on each of these alternatives; how-ever, there is relatively little empirical orconceptual research comparing them (for areview, see Shafer, 1993). Often decisionsto use one system or another are based onthe personal preference of the trainers,rather than on the student's individualabilities and needs, or on any empiricalevidence supporting a specific system(Reichle, Sigafoos, & Remington, 1991).The purpose of the current paper is to usethe concepts from Skinner's (1957) analysisof verbal behavior to examine the differentresponse form options for a nonverbal per-son.

SPEECH AS THE PRIMARY GOAL

Speech is the most desired responseform for verbal behavior for a number ofreasons. Perhaps the most significantadvantage of speech is the availability of alarge verbal community that can easilyprompt and reinforce vocal verbal behav-iors without special training. In addition,

99

100 MARK L. SUNDBERG

this sizable number of speakers producesendless vocal models which are oftenpaired with strong forms of reinforcement,resulting in specific speech sounds andpatterns becoming conditioned reinforcers.In addition, these pairings may establishvocalizations as automatically reinforcing,thereby naturally strengthening vocalbehavior (Bijou & Baer, 1965; Skinner, 1957;Vaughan & Michael, 1982). As a result,vocal behavior can be the easiest form ofverbal behavior for humans to acquire, asevidenced by the rapid acquisition ofspeech by most children even though par-ents have no special training. Vocal behav-ior is also portable, and responses can beemitted quickly and efficiently. However,if speech fails to develop in the typicalmanner, which is often the case for manydevelopmentally disabled (DD) children,speech may become a difficult responseform to establish.There are many causes for delayed or

defective language development, bothorganic and nonorganic (Drash & Tudor,1993). However, it is clear that the longer achild goes without speaking, the more dif-ficult it becomes to establish vocal behav-ior (Lovaas, 1977; Van Riper, 1978). Directspeech therapy, while often very effective,may not establish speech for some individ-uals, especially for those who have longhistories of failing to communicate. Theresponse forms required for speech maysimply be too hard to emit for some non-verbal children, especially for those whohave very little vocal behavior underechoic control. For these individuals, analtemative response form may be the mosteffective way to generate successful verbalbehavior (Zangari et al., in press). How-ever, the ultimate objective when teachingverbal behavior is still to establish vocalresponse forms. The three most commontypes of augmentative communication,writing, pointing and signing, may all beeffective, but only for specific students andunder certain circumstances. These altema-tives to speech and the issues involved intheir use will now be examined.

SPELLING, WRITING, TYPING, ANDFACILITATED COMMUNICATION

There are a number of nonvocal personswho can effectively communicate byspelling out words. Many disabled indi-viduals have demonstrated the indepen-dent use of this form of responding (e.g.,LaVigna, 1977). Perhaps most obvious is itsvalue to deaf individuals and persons withcerebral palsy (CP). Also, many individu-als with traumatic brain injury (TBI) retaintheir ability to read and write, even thoughthey are unable to speak. Finally, there area number of individuals with autism whoare nonvocal but have demonstrated theability to read and write. Rimland (1964)identified these autistic individuals as"criptic savants." While less desirable thanspeech, spelling words as a type of verbalbehavior can be effective. For example, aperson can use written words to ask forreinforcers (mand), identify nonverbalstimuli (tact), and engage in conversations(intraverbal). These repertoires require anumber of prerequisite behaviors, such asthose involved in literacy. However, thisform of verbal responding can be effectivefor establishing all types of verbal behaviorfor a nonvocal person. If a person can inde-pendently write (or type), but can notspeak, then this repertoire should certainlybe used. If a person does not indepen-dently speak or write, then other alterna-tive communication methods should beconsidered. The current paper will focusprimarily on these other alternatives.The recent technique of facilitated com-

munication makes use of transcription, butdiffers in many important ways from theindependent verbal behavior discussedabove. Facilitated communication involvesa response form consisting of spelled-outwords, produced by typing or pointing toletters, with the assistance of a facilitatorwho guides the learner's arm, hand, or fin-ger (Biklen, 1990; Haskew & Donnellan,1992). The basic premise is that the facilita-tion allows a nonverbal person to emit ver-bal behavior that he already possessed, buthad been unable to emit due to globalapraxia (Biklen, 1990).

SELECTING A RESPONSE FORM 101

FC as a form of verbal behavior has pro-duced a substantial amount of attentionfrom the media, due to demonstrations ofsophisticated levels of responding beingemitted by previously nonverbal students.Proponents of FC have claimed that manypreviously nonverbal developmentally dis-abled individuals can demonstrate nearnormal levels of verbal functioningthrough facilitation (Biklen, 1990; Haskew& Donnellan, 1992). However, severalrecent studies have demonstrated that inalmost all cases examined, the facilitatorwas the individual emitting the verbalbehavior (e.g., Green & Shane, 1993;Wheeler, Jacobson, Paglieri, & Schwartz,1993). In a few cases, it appears that someindividuals do have the necessary prereq-uisite repertoires, and could possibly bene-fit from FC. Therefore, a closer look at FCis warranted.

Advantages ofFacilitated Communication

Facilitation may be effective in generat-ing verbal behavior in previously nonvocalindividuals for a number of reasons (Table1). First, facilitation allows a trainer to eas-ily get an individual to emit a response.That is, many students are willing to havetheir hands touched and guided, and withespecially cooperative individuals, it mayseem much easier than teaching words orsigns. It may be possible that some individ-uals can already emit textual and transcrip-tive behaviors, but refuse to do so, or willnot do so without prompts. The facilitationmay also allow for the easier emission ofpreviously strengthened responses includ-ing mands, tacts, intraverbals, and autocli-tics. This may be most probable with indi-viduals suffering from certain types oftraumatic brain injury (TBI), or individualswith Cerebral Palsy (CP), and perhapseven some individuals with autism. Inaddition, there are a number of DD per-sons who have become elective mutes andhave later been found to have strong ver-bal behavior including the textual andtranscriptive repertoires.

Certain students may learn to communi-cate with FC because facilitation providesaccess to strong reinforcers (physical con-

Table 1The advantages and disadvantages

of facilitated communication.

Advantages of Facilitated Communication (FC)The response form is easy to establish with facilitation

FC may capitalize on existing verbal operants (espe-cially for elective mutes, those with CP, TBI, and someautistic individuals)

FC training may provide a powerful schedule of rein-forcement change, sometimes from near extinction toCRF

FC may capitalize on existing functional equivalencerelations (e.g., written-words with spoken-words orobjects)

The response form is written English, so the literatelistener does not need any special training to react tothe response

Transcriptive and textual behaviors are topography-based verbal behavior

The response form of FC may not be associated withany negative emotional history related to speech(aversive pairings)

Disadvantages of Facilitated CommunicationThe response form is dependent on two types of envi-ronmental support: access to a keyboard and physicalprompts from a trained facilitator; therefore, the stu-dent may not be able to emit verbal behavior in thenatural environment when EOs and SDs are strong

With FC there are frequently problems with general-ization to other facilitators and parents, this is espe-cially problematic because of the financial burden of atrained facilitator

The response form (typed words) involves a verycomplex type of verbal behavior; the student must beable to scan and discriminate among a wide array oftextual stimuli, be able to spell and type, and be ableto emit a complex chain of behaviors involving condi-tional and multiple discriminations

There is an extremely slow response time comparedto speech and sign, and the controlling variables andlistener's attention can be easily lost

tact, one-to-one attention). Subjects withstrong establishing operations (EOs) forthis type of reinforcement may be very suc-cessful with FC. Also, many of the individ-uals who enter an FC program may becoming from a near extinction reinforce-ment schedule. They now have easy accessto strong and consistent reinforcement,and it might not be surprising to observe alearning rate never demonstrated before bythe individual. FC may also capitalize onexisting functional equivalence relationssuch as those formed between writtenwords, spoken words, and objects (Hall &


Chase, 1991). A final advantage of FC pre-sented here is that the listener does notneed any special training to react to theresponse. Most adults working with thispopulation enter the profession as effectivereaders. However, as long as a facilitator isinvolved in the response form it will be dif-ficult to determine exactly who is thespeaker (Hall, 1993). It is likely though,that some literate individuals may be emit-ting their own mands, tacts, and intraver-bal responses through the help of facilita-tion.

Disadvantages ofFacilitated Communication

There are several complications with FCas a type of verbal behavior which make ita relatively undesirable response form(Table 1). Perhaps the most problematic, isthe fact that successful verbal interaction isdependent on two types of environmentalsupport. First, in order to verbally interactthe speaker must have access to the key-board. Second, a trained facilitator must bepresent to physically prompt the response.These requirements seriously limit theopportunities to engage in verbal interac-tions. If the keyboard is not available (e.g.,in another room or building, broken, or noaccess to a power source), or if the facilita-tor is absent, verbal responses cannotoccur. In addition, the need for a facilitatorplaces a financial burden on the studentwhich may be impossible to maintain overtime. This problem, along with the com-mon problems of generalization to otherfacilitators and parents, makes it quitelikely that a person will spend a substantialamount of time without a facilitator. As aresult of these limitations, FC is a highlyrestrictive form of verbal behavior thatcould not make much use of thecontingencies in the natural environment,which play a critical role in shaping verbalbehavior.

If the student is not literate, then FCwould be an extremely complex responseform to use to shape verbal behavior.There are several behavioral repertoireswhich are necessary for the successfulemission of verbal behavior in this manner.First, the speaker must acquire letter dis-

crimination skills and be able to scan anddiscriminate among a wide array of stimuliin order to find and type (or point to) theletters when needed. In addition, thespeaker must be able to spell the words toemit the correct sequence of letters neededto make functional words. Finally, thespeaker must be able to emit this complexchain of behaviors involving conditionaland multiple discriminations in a relativelyrapid manner in order to affect a listener ina timely way (if the response is too slowcontrolling variables or the listener's atten-tion can be easily lost). Facilitation mayseem to avoid the need for the student tohave these repertoires, but if the student isnot literate he cannot be the primaryspeaker.

An Analysis ofFC Using the Concepts FromSkinner's Book Verbal Behavior

The mand repertoires. Mands, which arecontrolled by EOs and specific conse-quences (Michael, 1988; Skinner, 1957), arean important type of verbal behavior.When an EO is in effect, manding may leadto specific reinforcement. In other words,when someone wants something, a mandmay get it for them. However, EOs occurunder a wide variety of conditions in thenatural environment. In order to effectivelymand, a speaker must have a responseform readily available. A typical mandrepertoire will be difficult to develop usingFC because the response form may not bequickly available to the student when EOsare naturally strong (e.g., an EO for a spe-cific food). The facilitator may be unavail-able, or the typing device may not beimmediately accessible, especially if theperson is highly mobile. This lack of anavailable response form in the naturalenvironment may lead to aggression, SIB,or other types of negative behavior whichhave been reinforced as mands in the past.In addition, the mand training that wouldbe possible, would be restricted to mostlycontrived EOs occurring under a limitedset of conditions involving sitting at thekeyboard with the facilitator present. Thistype of training may also involve multiplecontrol, in that other verbal and nonverbal


stimuli are frequently present during train-ing, making generalization and puremands less likely.

The tact repertoires. Tacts, which are ver-bal responses controlled by nonverbalstimuli, would be limited to nonverbalstimuli presented while the facilitator waspresent. The number of tacts emitted oracquired would also be related to the facili-tator's ability to present nonverbal stimulito the student. Because the facilitator'shands are busy facilitating, this task maybe particularly difficult. Also, attending tothe letters on the keyboard may competewith attending to visual, nonverbal stimuli.Furthermore, tacting like all the other ver-bal operants would be impossible to emitin the natural environment without thefacilitator or the FC device. Therefore, itwould be difficult to develop a typical tactrepertoire using FC. Even if the tacts werealready strong in the student's repertoirethe opportunities to emit them arerestricted, which may evoke negativebehaviors.

The intraverbal repertoires. Intraverbalbehavior, which consists of verbalresponses controlled by nonmatching ver-bal stimuli, would also be limited to verbalstimuli presented while the facilitator waspresent. However, verbal stimuli are easyto present while using FC, which is proba-bly why much of the typical FC traininginvolves intraverbal behavior consisting ofquestions and answers. An interestingproblem arises with the use of transcriptivebehavior as a response form in intraverbalinteractions. An intraverbal that would becompletely consistent with the responseform of transcription would involve verbalstimuli presented in the textual form whichwould control typed responses. However,the typical intraverbal interactions with FCconsist of spoken words and typedresponses. This interaction involves themixing of two different response forms. Itis unclear what the effects, if any, thiswould have on verbal development. Thissame problem arises in the analysis ofreceptive behavior, and in the develop-ment of intraverbal and receptive behaviorwith pointing systems.

The codic repertoires. Codic behavior(Michael, 1982) consists of textual (reading)and transcriptive (writing and spelling)behaviors, and are the prerequisite reper-toires for FC. If a student fails to empiricaldemonstrate these skills it would beimpossible to say that he was the personemitting mands, tacts, intraverbals, orautoclitics.

The autoclitic repertoires. Skinner (1957)distinguishes between primary verbalbehavior (e.g., mands, tacts, intraverbals)and secondary verbal behavior (autocli-tics). Secondary verbal behavior neverstands on its own, rather it is alwayscaused by some aspect of a primary verbaloperant. For example, a speaker can mod-ify a primary tact such as It's a moose withthe qualifying autoclitic I think, resulting inthe total response of I think it's a moose. Ithink is an autoclitic tact that enjoins the lis-tener to react to the primary tact involvinga nonverbal stimulus and the word moosewith caution. If the variables controllingthe primary tact were stronger the speakermay have emitted the response I'm sure it'sa moose, resulting in a different effect on thelistener. Autoclitic behavior can be consid-ered "verbal behavior about verbal behav-ior" (Peterson, 1978), and is reinforced bythe special effects it has on listener behav-ior.

Autoclitic responses can be classified asautoclitic tacts or autoclitic mands(Peterson, 1978). Autoclitic tacts involvetacts of some nonverbal aspect of the pri-mary controlling relation, while autocliticmands involve responses controlled byEOs which are emitted because the valueof a particular reaction to the primaryresponse is strong. For example, saying I'msure it's a moose may involve an autocliticmand to get the camera, rather than a tactof the strength of the controlling relationbetween the nonverbal stimulus and theprimary response.With facilitated communication, auto-

clitic behavior would be impossible if thestudent was not the genuine author of theprimary verbal behavior. However, if thestudent was the author, autoclitic mandsand tacts could occur in much the same


way as other topography-based systems.They would be limited by the same vari-ables that limit the development of the pri-mary responses with FC, such as a slowand complex response form, and the lim-ited opportunity to emit the verbal behav-ior in the natural environment.The receptive repertoires. Receptive dis-

criminations involve a verbal stimulus(usually a mand by a speaker) which evokesa nonverbal response by the listener.Touching the door when asked to is anexample of receptive behavior. Receptivelanguage training with FC users is typi-cally not conducted using the sameresponse form that the student uses foremitting verbal behavior. Rather, receptivediscriminations are accomplished in thesame manner as with speaking persons.That is, vocal stimuli are used to evokenonverbal responses. This complicationwould seem to have some effect on verbaldevelopment and the formation of func-tional equivalence relations. However, theblending of spoken and written Englishmay be effective in advancing a student'sreceptive repertoires with FC, but thisquestion is an issue for empirical research.

If one were to use the codic repertoiresof FC in receptive discriminations, which isunlikely, a textual stimulus must evoke atextual response prior to the nonverbalresponse. For example, in order to cor-rectly respond to the instruction to touchthe door the written word door must evokea textual response which in turn mustevoke the nonverbal behavior. Withspeech, the vocal stimulus alone evokes thenonverbal response. This multipleresponse requirement with FC makesreceptive discriminations somewhat morecumbersome. Therefore, training on recep-tive discriminations with FC are not typi-cally conducted in the student's primaryresponse form.

Why the Controversy Over FacilitatedCommunication?

There are several possible reasons for thecurrent controversy over FC. First, thereare undoubtedly some DD persons whoare literate and/ or fast learners, and may

benefit from this method of communica-tion. Some of these literate individuals maybe nonvocal from birth or elective mutes,but may show some immediate and rela-tively high level verbal skills given facilita-tion. It is also possible that the significantchange in the reinforecment scheduleavailable for these students while engagedin FC, along with some careful 1-to-1 shap-ing, may be sufficient to evoke (or rapidlydevelop) mands, tacts, intraverbals, andautoclitics. It is not uncommon to find pre-viously nonverbal students who, whengiven instruction, have demonstrated asurprisingly rapid acquisition of textualand transcriptive behaviors, but continue toshow little or no improvement in speech.Second, many of the facilitators truly

believe that they are not helping the stu-dents respond, but are only providingemotional support. Hall (1993) has pointedout, how in many cases, facilitators may beengaging in automatic verbal behavior (i.e,verbal behavior in which the self-editingprocess has been suspended), and fail tonotice that they are the producing thetyped message. It certainly is possible thatthe facilitators are unable to tact their owninvolvement in the facilitation process. Forexample, the facilitators at the 0. D. HeckCenter on the Frontline Program on PBS,were surprised at the results of theresearch that demonstrated the criticalrole they played in providing correctresponses. Wheeler et al. (1993) haveincluded an appendix in their paper towarn researchers about the devastatingemotional effects on facilitators upon find-ing that it is the facilitators who are emit-ting the verbal behavior. These examplesclearly demonstrate that many facilitatorsare probably unaware of the role that theyplay in the communication process.A third reason for the controversy is that

many of the supporters of FC assume thatcognitive abilities exist, and that FC simplyallows access to these cognitions (Biklen,1990). This cognitive theory proposes thatverbal behavior is already strong in thecognitive processing system, but due toglobal apraxia cannot be emitted. Theobservation of high level verbal behavior


during FC seems to support this theory,while drawing attention away from thefacilitator's behavior.A fourth issue that has caused many to

try FC is the general hope that it will work.Many parents and professionals are look-ing for quick cure for DD individuals. Likemany "miracle cures" before it, FC seemsto provide that hope to many. This hope,like the belief in cognitions, often pre-cludes observers from looking for the truesource of the verbal behavior.

Finally, there is a general lack of under-standing of the complexity of the issuesrelated to training verbal behavior to non-verbal persons (Sundberg, 1990). Knowingwhere or how to start language interven-tion programs for nonverbal individuals isoften difficult. Especially for nonverbaladults who may demonstrate high levels ofreceptive and other nonverbal behaviors. Itis often tempting to start expressive lan-guage training at relatively high verbal lev-els, rather than at basic mand, tact, andintraverbal levels. When the training failsto generate expressive verbal behavior,other factors are blamed (which is oftenwhy many individuals are still nonverbalas adults). FC, and its premise that theseverbal behaviors exist at the cognitivelevel, seems to conveniently sidestep all ofthe work and problems involved in teach-ing an early verbal repertoire. However,given the evidence which points to thefacilitator as the speaker (e.g., Green &Shane, 1993), and the conceptual barriers ofFC which highly restrict verbal development,FC is probably not a response form that willbenefit many nonverbal individuals.

Alternatives to Speech, Transcription, and FC

Pointing systems and sign language arethe two major alternatives to speech, inde-pendent transcription, and facilitated com-munication. There is an extensive body ofresearch on each of these two methods ofaugmentative communication (for reviews,see Fristoe & Lloyd, 1977; Shafer, 1993;Zangari et al., in press). However, there isvery little research which directly com-pares pointing systems and sign language(Shafer, 1993). The research that does exist

suggests that there are several practicaland conceptual differences between thesetwo types of verbal behavior. These differ-ences may have significant implications forthe procedures used to establish verbalbehavior for nonverbal students (e.g.,Hodges & Schwethelm, 1984; Michael,1985; Sundberg & Sundberg, 1990;Wraikat, Sundberg, & Michael, 1991).Further identification and clarification ofthese differences may help parents, teach-ers, and other professionals decide whichsystem is most appropriate for an individ-ual, and what special intervention tech-niques may be required to establish,develop, and maintain effective verbalbehaviors.

Perhaps one of the most important dis-tinctions between these two types ofresponse forms is the one Michael (1985)makes between stimulus selection-basedverbal behavior and topography-basedverbal behavior. In selection-based verbalbehavior, the response form consists ofselecting a stimulus from an array of stim-uli by scanning, then pointing to thedesired stimulus. The response form ofscanning and pointing consists of a multi-ple response that has essentially the sametopography for all verbal relations; what isdifferent is the stimulus selected.

Selection-based verbal behavior can becontrasted with topography-based verbalbehavior where the topography of theresponse, rather than a selected stimulus, isthe distinguishing part of the verbal rela-tion. Speech, writing, and sign language allconstitute topography-based verbal behav-ior because there is a different responsetopography for each controlling variable.For example, there is a specific sign orword for shoe which involves differentmuscle movements than the sign or wordfor hat, while pointing to a symbol for shoeinvolves basically the same muscle move-ments as pointing to a symbol for hat. Thedifferences between these two types of ver-bal behavior will be further identified inthe form of the advantages and disadvan-tages of each system, and by an analysis ofeach system using the concepts fromSkinner's (1957) analysis of verbal behavior.


POINTING SYSTEMS (INCLUDINGCOMPUTER SYSTEMS)

Pointing to (or touching, looking at, etc.)pictures, symbols, or the keyboards ofcomputer-operated devices that electroni-cally produce verbal stimuli, has provento be an effective verbal response formfor previously nonverbal students(e.g., Hurlbut, Iwata, & Green, 1982;McNaughton, 1976; Mirenda, 1985; Reichle,York, & Sigafoos, 1991; Romski & Sevcik,1988; Shafer, 1993; Vanderheiden & Lloyd,1986; Zangari et al., in press). There areseveral advantages of a pointing systemwhich make it an attractive form of aug-mentative communication. In fact, it is cur-rently the most preferred system to use bylanguage intervention specialists (Shafer,1993). However, there are several disad-vantages which must be considered aswell. Both the advantages and disadvan-tages of pointing systems will now be pre-sented (Table 2).

Advantages ofPointing Systems

Perhaps the primary advantage of point-ing systems is that the listener does notneed any special training. Most symbolshave the English word written on them,which means the response can be emittedin the presence of any attentive and literatelistener. In addition, many of the first sym-bols or pictures are easy to acquire becausethey consist of simple matching-to-sample.For example, the symbol for ball or cupmay look very much like a real ball or cup,and many students who are nonverbal caneasily match similar stimuli (Keogh &Reichle, 1985). This matching-to-samplerepertoire may facilitate the early acquisi-tion of pointing, much in the same waythat a strong echoic repertoire can facilitatespeaking, and a strong imitative repertoirecan facilitate signing (Sundberg, 1990).Another advantage of pointing systems

is that the response topography (pointing)is the same for each verbal operant, sotraining differential responding is not nec-essary. This feature makes this form ofresponse especially effective for individu-als with muscle control problems such asthose with CP and TBI. Also, there has


of pointing systems.

Advantages of Pointing SystemsThe listener does not need special training, theEnglish word typically accompanies the symbol orpicture

The first responses involve simple matching-to-sam-ple which may be easy to acquire

There is no special shaping of individual responseforms, scanning and pointing are about the same foreach verbal relation, and it is much easier than shap-ing the individual responses required for speech orsign

Pointing at stimuli may avoid any negative emotionalhistory associated with speech

The response of pointing may already be strong in theperson's repertoire

Disadvantages of Pointing SystemsPointing at stimuli requires environmental support,the speaker must have the stimuli available to emitverbal behavior, therefore, the student cannot alwaysemit verbal behavior when EOs and SDs are strong inthe natural environment

The symbols and pictures become increasinglyabstract as the complexity as the words increase, andthere are space limitations on the boards

The pointer needs to have the listener in close proxim-ity (speech and signs can be successful across a room)

The response is slow compared to speech and sign,and the controlling variables and listener's attentioncan be easily lost

Pointing systems constitute stimulus selection-basedverbal behavior and may be more difficult to acquire;the response form involves a complex type of verbalbehavior consisting of conditional discriminations(two or more controlling variables), and a multiple-component response form (scan and point)

There is no point-to-point correspondence betweenthe response and the response product, this corre-spondence facilitates the acquisition of verbal behav-ior with self-feedback

There is not an existing or natural verbal communitywhich uses pointing systems as a form of verbalbehavior; trainers tend not to use pointing systems tointeract with the students, rather English is used

Typically there is no improvement in speech

With pointing systems it is harder to establish func-tional equivalence

been substantial progress in the technolog-ical development of communicationdevices, computers, and software which isgreatly improving the efficiency of thisform of verbal behavior (e.g., Zangari et al.,in press). These developments may pro-


vide some solutions to the disadvantagesof pointing systems identified below.

Disadvantages ofPointing Systems

Unfortunately, there are several practicaland conceptual limitations with pointingsystems that may impede verbal develop-ment. These limitations should be consid-ered before designing a language interven-tion program for a nonverbal student.Perhaps one of the most significant practi-cal barriers with pointing systems is that,like with writing, typing, and FC, respond-ing is dependent on environmental sup-port. This presents practical limitationsbecause, unlike speech and sign language,the response form is dependent on otherstimuli and cannot be emitted when thosestimuli are absent. For example, to success-fully mand with a pointing system theboard must be present. If the board is notavailable, the response cannot occur. Thistype of constraint is significant becauseverbal behavior may be impossible to emitwhen EOs and SDs are strong in the naturalenvironment. These conditions may alsoevoke negative behavior such as tantrum-ming or SIB, because these behaviors maybe the only response form available to theindividual.Many other practical limitations of

pointing systems have been identified inthe literature (e.g., Shafer, 1993; Sisson &Barrett, 1983; Sundberg & Sundberg, 1990;Wraikat et al., 1990). Some of these limita-tions include the difficulties of portrayingcomplex words in symbol form, the spacelimitation of most boards, and the need toalways have the listener in close proximity(speaking and signing can be successfuleven when a listener is across the room).Also, the response of pointing to stimuli isnaturally slower than speaking (or signing)because of the required time to scan thearray of stimuli to find the appropriatestimulus. However, these problems couldbe overcome with substantial training andutilization of the many new technologicaldevelopments in presenting and arrangingstimuli. Lana the chimpanzee (Rumbaugh,1977), for example, acquired a very rapidrate of verbal responding, using an easily

accessible computer-operated responseboard, strong reinforcers, and careful shap-ing.The conceptual limitations of pointing

systems are more problematic. A majordisadvantage of a selection-based form ofverbal behavior is that the response formof scanning and pointing is a multiple-component response, rather than a single-component response like that of speakingand signing. Also, unlike speech and signlanguage, the controlling variables consti-tute conditional discriminations in that thepresentation of stimulus or establishingoperation (e.g., a cup), alters the evocativeeffect of a second stimulus (the appropriatesymbol for cup or drink). Therefore, effec-tive responding with a selection-based sys-tem always involves two or more control-ling variables and a two-componentresponse. These unique features of selec-tion-based systems make teaching this typeof verbal behavior more complicated(Michael, 1985; Shafer, 1993; Sundberg &Sundberg, 1990; Wraikat et al., 1991).Another limitation with pointing sys-

tems is the absence of a functioning verbalcommunity that uses these systems tocommunicate. Students who learn to usepointing systems typically do so withoutthe advantages of observing competentspeakers engage in verbal interactions bypointing. Also, students do not haveopportunities to observe competent speak-ers using more advanced symbols from thecommunication system. Rarely is it thecase that adults (e.g., teachers, parents,staff) use symbol boards to communicatewith each other, nor do they regularly usethe boards when verbally interacting withthe students; they typically use spokenEnglish. This lack of contact with modelsin the natural environment would seem tomake language acquisition more difficult,especially given the fact that exposure to afluent verbal community is essential for thedevelopment of the verbal repertoires(Skinner, 1957).A final issue about symbol systems con-

cerns the effects of learning to point on thedevelopment of vocal behavior. In theirreview of the literature on pointing sys-


tems, Sisson and Barrett (1983) note thatnone of the studies mention vocal behavioror vocal improvement. This lack of vocalimprovement may be due to the fact thatwith pointing there is not a differentialresponse form which can get paired with aspecific word. That is, the same pointingresponse is associated with all words.While it seems plausible that vocalizationscould improve due to the effects of suc-cessful verbal interaction, the degree ofimprovement may be small.

The Picture-Exchange Communication System

The Picture-Exchange CommunicationSystem (Bondy & Frost, 1993) is a uniquetype of verbal behavior where a studentscans, selects and hands a picture to a lis-tener as a form of verbal behavior. Thistype of selection-based verbal behavior dif-fers somewhat from the other types dis-cussed above. It does involve the scanningand selecting of specific stimuli and a mul-tiple-component response, but theexchange aspect involves some importantfeatures which may facilitate the acquisi-tion of verbal behavior. Specifically, theexchange requires that the adult interactwith the response form (i.e., they mustreceive the picture). This makes it easier forthe adult to function as a speaker and a lis-tener in the language system. The picturesmay become conditioned reinforcers moreeasily, and functional equivalence relationsmay form quicker. This mitigates the prob-lem of the adult not using the symbolboard. In addition, it may be that this sys-tem avoids the problem of weak echoic orimitation skills, and avoids any negativehistories often associated with attempts tospeak or sign. Finally, it appears that thereis a reported effect on vocal behavior(Bondy, personal communication). Futureresearch on this system should prove inter-esting.

An Analysis ofPointing Systems Using theConceptsfrom Skinner's Book Verbal Behavior

There are several aspects of a pointingsystem, and stimulus selection-based sys-tems in general, which impede the naturalshaping of the verbal operants. Perhaps the

main barriers are the dependence on envi-ronmental support, the complexity of thestimuli and responses involved, the lack ofa natural verbal community, the slow rateof responding, and the difficulty of por-traying complex words in symbol or pic-ture form. Below is an analysis of howthese barriers will affect the developmentof each of the elementary verbal operants,as well as the development of the autocliticand receptive repertoires.

The mand repertoires. Pointing systemssuffer from the same problem identifiedfor FC in that the emission of a responsedepends on environmental support. WhenEOs occur in the natural environment thestudent often may not have immediateaccess to the board or computer. This situa-tion is common for many users of pointingsystems who are very mobile (e.g., autisticchildren), or find carrying a communica-tion device (or board) cumbersome oraversive. The occurrence of EOs in the nat-ural environment cannot produce mandingby pointing to a stimulus, so negativebehaviors (e.g., screaming) may occur asmands. In addition, the loss of the oppor-tunity to emit the trained verbal responsesunder the control of EOs in the naturalenvironment severely limits teachingopportunities (e.g., Hart & Risley, 1975)and verbal development (Michael, 1988,1993). It may be difficult, for example, toteach complex mands such as asking ques-tions. These problems may not be as seri-ous for students who are wheelchairbound. However, even these students arefrequently in situations where they will beunable to mand when EOs are strong (e.g.,in bed, on the toilet, on the floor mat).These problems can be overcome some-

what by ensuring that the board is alwaysavailable to the student, or devising a sys-tem that is highly portable. Picture walletsand picture necklaces have been used suc-cessfully for some clients, but the numberof the responses available are limited, andthe rate of emitting a verbal response con-taining several components will be slow.This is a problem because if the response istoo slow, the temporal contiguity betweenspecific antecedents, behavior, and conse-


quences will be broken (i.e., searching forthe right stimulus takes time and the origi-nal controlling variables or the listener'sattention can be lost, also the consequencesmay become too delayed).

The tact repertoires. Tacting would proba-bly be the least affected by the uniquenature of selection-based verbal behavior.Nonverbal stimuli can easily be presentedand responding can be prompted, shaped,and differentially reinforced. However, asthe nonverbal stimuli become more com-plex it becomes increasingly difficult topresent these stimuli pictorially or symbol-ically. This complexity may require moretraining trials than a similar verbal operantusing a topography-based system (Sund-berg & Sundberg, 1990; Wraikat et al.,1991). In addition, the number of responsesavailable may be limited due to spacerequirements, and the length of theresponse may be short due to time andresponse effort requirements.

The intraverbal repertoires. There are sev-eral general difficulties in establishing anintraverbal repertoire with pointing sys-tems. First, the lack of a natural verbalcommunity means that the symbols are notsystematically paired with adults andstrong reinforcers in the manner thatspeech is paired with such reinforcers.Much of a speaking child's intraverbalbehavior is affected by regular exposure toa highly reinforcing natural verbal commu-nity. Intraverbal connections (e.g., stringsof words, songs, phrases) are often pairedwith powerful reinforcers such as food,attention, and physical contact. As a result,words become automatically reinforcing toemit because the response products (sen-sory feedback or what you hear, see or feelwhen verbally responding) have reinforc-ing properties and as a result have func-tional control as a consequence (Vaughan& Michael, 1982; Skinner, 1957; Sundberg,Sundberg, & Partington, 1993). There is aself-shaping process where the responsebecomes stronger each time it is emittedbecause the response is automatically rein-forced. This effect may not occur if rein-forcing adults do not use the symbols toverbally interact with the student.

A second complication in developing theintraverbal repertoire with pointing sys-tems involves the presentation of verbalstimuli. In order to be consistent with theestablished response form of pointing, pic-tures or symbols would need to be pre-sented as verbal stimuli, and trainingwould involve establishing intraverbalconnections between one picture andanother. This repertoire would require spe-cial training procedures and symbols toestablish. For example, a trainer couldpoint to a general symbol for food andestablish pointing to a number of the sym-bols for specific foods. Or, when askingabout a field trip, a trainer could point outthe symbols What did you see on your trip?However, this type of intraverbal presenta-tion is not typical. Instead trainers usuallyask the question in spoken English whichrequires an additional intraverbal relationbe established (i.e., English and pointing).It is unclear what effect this might have onintraverbal development, but it is notuncommon to see pointing system userswho are unable to emit a substantialamount of intraverbal behavior. (However,the same point could be made for all of theforms of augmentative communication.The real problem is that intraverbal behav-ior is typically not systematically targetedfor intervention.)

The codic repertoires. It is certainly possi-ble that the written word which is pairedwith the symbol may eventually be able todemonstrate control over the response.However, if English words did come toevoke pointing to specific stimuli the rela-tion would actually be intraverbal becauseof the lack of point-to-point correspon-dence between the verbal stimulus and theverbal response (Skinner, 1957). Thus, itwould be impossible to have true codicbehavior with pointing, unless it involvedpointing to letters to spell out words.

The autoclitic repertoires. The lack of afunctioning verbal community in the natu-ral environment, that models and rein-forces autoclitic behavior by pointing tostimuli, would seem to have an effect on astudent's autoclitic development. How-ever, most of the autoclitic functions could


be accomplished with a pointing system,but would require special training proce-dures. The responses would probably beslower due to the general complexities ofemitting primary responses, and it may bedifficult to differentially establish symbolsfor some of the rather complex autocliticresponses.

The receptive repertoires. Receptive lan-guage with pointing systems is often hardto distinguish from intraverbal behavior. Atruly receptive response with pointing sys-tems would be, for example, touching aspecific object given the presentation of theassociated symbol. However, receptivetraining is typically not systematically pro-vided in this way. Rather, spoken wordsare used as SDS to touch specific symbols(e.g., "Show me car"), probably becausemany students already have a number ofbehaviors under receptive stimulus controlusing spoken words (often receptive trialsoccur in the absence of the board).Therefore, it is probably the case that mostof what is assumed to be receptive discrim-ination trials between words and pointingto symbols or pictures is probably intraver-bal rather than receptive, because the stim-ulus and the response are both verbal andlack point-to-point correspondence. On theother hand, what is thought to be intraver-bal may actually be receptive, it dependson the verbal or nonverbal properties ofthe stimuli involved. In any case, it is cum-bersome to conduct receptive trials consis-tent with the response form of pointing, soEnglish is typically used instead.

In conclusion, it appears that there aremany disadvantages of a pointing system,including several restrictions on the devel-opment of the verbal operants. For someindividuals there may be no alternatives,but for others sign language may be worthconsidering. An analysis of the advantagesand disadvantages of sign language, fol-lowed by an analysis of sign language andthe acquisition of the verbal operants, willnow be presented.

SIGN LANGUAGE

The use of sign language with nonverbalpersons has also proven to be an effective

way to generate verbal behavior. There arenow several studies in the literature whichdemonstrate that a wide variety of nonver-bal people can acquire sign language (e.g.,Bonvillian, Nelson, & Rhyne, 1981; Carr,1979; Clarke, Remington, & Light, 1986;Fristoe & Lloyd, 1977; Partington et al., inpress; Reichle, York, & Sigafoos, 1991).Sign language has many advantages as aresponse form and only a few disadvan-tages (Table 3).

The Advantages of Sign Language

Sign language is a very efficient form ofverbal behavior and has many of the sameadvantages as speech. Sign language con-stitutes a topography-based language, so itlike speech, benefits from a differentialresponse form that is independent from thephysical environment. Signs are com-pletely portable so verbal behavior can beemitted under all potential types of con-trolling variables. Also, the deaf popula-tion provides a natural verbal community


of sign language.

Advantages of Sign LanguageSign language constitutes topography-based verbalbehavior that is free from environmental support,making it portable and conceptually similar to speechThe deaf community constitutes a natural verbal com-munity that uses sign language, so materials andtrainers are available

The response form (imitation) may already be strongin the person's repertoire

Trainers can use and fade physical prompts to obtainthe response form

The stimulus and the response often resemble eachother (an iconic relation) providing a built-in prompt

Sign language can improve speechSign language involves a single response and a simplediscrimination

The response and response product have point-to-point correspondence

Sign language may avoid the negative emotional his-tory involved with speech

Disadvantages of Sign LanguageThe listener must have special training, and there is aneed to establish a signing verbal community

The trainer must shape individual responses


where sign language is the primaryresponse form. This makes sign language aliving language in the sense that it is stillgrowing, changing, and affecting andbeing affected by, a community of listenersand speakers. This also ensures a variety ofmaterials, teachers and trainers, videos,research, and other auxiliary support itemsnecessary for teaching verbal behavior tononverbal persons may be available.

Sign language has several specific advan-tages over speech, and the other types ofaugmentative communication which maymitigate the disadvantages identifiedbelow. Sign language training may be eas-ier than speech training because many DDpersons who cannot echo sounds, can emitmotor responses under imitative control.These responses can quickly be broughtunder EO, nonverbal, and verbal controlwhile fading out the imitative control, thusthe establishment of a multiply controlledmand (Sundberg, 1990). For example,under many circumstances it may be easierto teach a student the sign for drink, than itwould be to teach the vocal word drink, orto teach the student to reliability point tothe symbol for drink.

If a student does not have a strong imita-tive or echoic repertoire it will probably beeasier to teach him to imitate a motormovement than to echo a word, because ofthe advantage of physical prompting andfading procedures. For imitative training,the student's hands can be physicallyguided by the teacher to the appropriateposition and then the physical prompts canbe faded out. This physical prompting pro-cedure is impossible with the vocal muscu-lature since one cannot directly manipulatethe parts of the vocal system to producespecific sounds. The use of physicalprompting may make the shaping easier,while also providing clear and unambigu-ous models of the appropriate responseform.

Sign language also benefits from the factthat many of the signed response formsclosely resemble the controlling variablesin the environment. The sign ball for exam-ple, is made by placing the curved fingertips of both hands together out in front of

the body. The sign looks like a ball madewith the hands. The iconic relationbetween controlling variables and appro-priate responses may make sign acquisi-tion easier than vocal response acquisition.Spoken English has only a few of theseiconic, or onomatopoetic relations (e.g.,"hum," "bow wow," "buzz"), and they areof little help in early language acquisition.

Finally, there are several ways in whichsign language may facilitate the establish-ment of vocal response forms and improvelanguage acquisition in general. First, signlanguage may solve the immediate prob-lem of not being able to verbally interactwith others. Signs allow the listeners (e.g.,parents, teachers, peers) to immediatelyunderstand what the person might be try-ing to say, thereby permitting the deliveryof the specific reinforcement necessary toreduce the evocative effect of the EO. Inaddition, this successful verbal interactionprovides an excellent opportunity to shapearticulation. These points could be madefor all types of augmentative communica-tion, but sign language, as a topography-based system free from environmentalsupport, has some unique features whichmay result in substantial vocal improve-ment.

First, if trainers speak as they sign, andrequire and reinforce approximations tospoken words, specific words can becomedifferentially associated with specific signsand highly reinforcing verbal interactions(e.g., successful manding). Therefore, notonly might specific speech sounds becomepaired with specific signs, but specificsounds may also become conditioned rein-forcers, and even automatic reinforcers aswell. These new forms of reinforcementcan strengthen vocal behavior in manyways (e.g., Skinner, 1957; Sundberg et al.,1993). For example, in mand training if thespoken word eat is consistently paired withthe sign eat and the delivery of food, thespoken word eat may acquire new evoca-tive and reinforcing effects (Michael, 1983).As a result, the spoken word eat may alsoenter into a number of different functionalequivalence relations such as those wherevocal stimuli control signing and receptive


responding (Hall & Chase, 1991; Sundberg& Sundberg, 1991). Also, the establishmentof echoic stimulus control, and other typesof verbal behavior involving vocal behav-ior, may now be easier, especially if vocalbehavior already occurs to some degree(Clarke, Remington, & Light, 1988). It isalso possible that the successful establish-ment of a generalized imitative repertoire(which is typical for many accomplishedsigners), may facilitate echoic behavior.

Sign language may improve articulationfor a number of other reasons as well. Ifsigns begin to evoke specific vocalizations,then signs can be used as a new type ofprompt to evoke these vocalizations. Thistype of prompting may be more effectivethan typical echoic prompts which providethe response form, making it harder totransfer control to the other types of verbalbehavior. Also, students learn to sequencemotor movements which are often easier tosequence than vocal movements. Once themotor movements are learned, specificvocalizations can be matched with thesigns. This sign-vocalization prompt canhelp in other ways as well. A student canuse signs to prompt his own vocalizations.That is, if a nonverbal stimulus can evoke asign, and a student is able to emit a vocal-ization under the control of a sign, then hecan self-prompt his own vocalizations. Forexample, when a signing student wants acup, but there is not one present, the signcup can prompt the articulation of theword cup. This self-prompting may resultin more successful vocal-verbal interac-tions and less punishment for attempts tospeak.

The Disadvantages of Sign Language

The primary disadvantage of sign lan-guage is that the listeners must learn anduse sign language to function as an effec-tive verbal community. Learning to signmay involve a substantial commitment toattend classes and practice sessions (likeother second languages). Also, a signingenvironment must be established wheresigns are consistently reinforced, and mod-els of signing are frequently provided andpaired with other forms of reinforcement.

The establishment of this verbal commu-nity is essential for verbal development(Skinner, 1957; Sundberg, Milani, &Partington, 1977), but these barriers maybe difficult to overcome because of the lackof available training and resources, andfrequent staff turnover. However, it shouldbe pointed out that even a small signingrepertoire on the part of trainers may besufficient to begin sign training.

Sign language has other disadvantagesas well. First, since sign language is atopography-based system each responseform (sign) must be individually shaped(unless the individual has acquired a gen-eralized imitative repertoire). This requiresthat staff have special training in shaping,prompting and fading, and in differentialreinforcement procedures. Pointing tostimuli and FC responding eliminate thisproblem (but at a later cost) by using thesame behavior for all verbal operants andphysically prompting the response form,respectively. The use of sign language alsorequires different training procedures thanthose used with speech and pointing sys-tems. For example, physical and imitativeprompts can be used along with echoic andintraverbal prompts. These training meth-ods may require extra effort to leam.

An Analysis of Sign Language Using theConcepts From Skinner's Book Verbal Behavior

Sign language is functionally closer tospeech than the other two types of aug-mentative communication (Table 4). This isprimarily because sign language consti-tutes a topography-based verbal systemthat is free from environmental support.As a result, the acquisition of the verbaloperants using sign language as a responseform is similar to the acquisition of the ver-bal operants using speech as a responseform. Sign language development seems toparallel speech development in every way,providing there exists a consistent reinforc-ing verbal community. Data show thatwhen a deaf child is raised by deaf parentswho sign, the child acquires verbal behav-ior similarly to a hearing child raised byspeaking parents (Moores, 1978; Vernon &Koh, 1970; Zwiebal, 1987). Therefore, if the


Table 4Comparing the response forms by the potential strength of the verbal operants.

Speech Facilitated Pointing SignCommunication Systems Language

Mand Strong Weak Weak-Medium Strong

Tact Strong Weak-Medium Medium Strong

Intraverbal Strong Medium Weak Strong

Codic Strong Strong Weak Weak-Medium

Autoclitic Strong Weak-Medium Weak-Medium Strong

Receptive Strong None (Speech) None (Speech) Strong

critical people in a signing person's envi-ronment do not acquire sign language,then the probability is low that the verbalrepertoires will develop properly.

The mand repertoires. Signs can be equallyeffective for manding as speech, providedthe people in the student's environmentuse and reinforce sign language.

The tact repertoires. Tacting in sign lan-guage differs in no significant way fromtacting with speech. The key issue is again,is the presence of a reinforcing signing ver-bal community.

The intraverbal repertoires. The establish-ment of sophisticated intraverbal behaviorprimarily depends on contact with a sign-ing verbal community. Limited exposureto signing models and audiences will makeit difficult to establish complex intraverbalbehavior. However, with a signing verbalcommunity, the development of signedintraverbal behavior can parallel that ofspoken intraverbal behavior.The codic repertoires. Since there is no

point-to-point correspondence betweensigns and letters, it is unlikely that codicbehaviors would occur via signing (thiswould require speaking, or fingerspelling).However, intraverbal relations betweenwritten words and signs can occur (likewith pointing systems) providing studentswith a form of literacy.

The autoclitic repertoires. Autoclitic devel-opment with sign language also differs inno significant way from speech. Secondarymands and tacts can be acquired, given areinforcing verbal community.

The receptive repertoires. Responding to a

signed verbal stimulus does not differconceptually from responding to a spokenverbal stimulus. Therefore, all forms ofreceptive behavior are possible with signlanguage, given a reinforcing verbal com-munity.

In conclusion, sign language has severaladvantages that make it an appealing alter-native response form. Especially attractiveis the fact that it is a topography-based sys-tem that is free from environmental sup-port. In addition, verbal development withsign language appears to closely parallelverbal development with speech, whilethis is not the case with the other responseforms (Table 4). However, the major disad-vantage of sign language is the require-ment that the trainers, and other persons inthe student's environment, learn to sign.Deciding which augmentative systemmight be the most appropriate for a spe-cific individual will now be discussed.

SELECTING A RESPONSE FORM

Speech is the response form of choice.Effort should always be given to develop-ing vocal verbal behavior prior to consider-ing an augmentative system. Particularprocedures like mand training and auto-matic reinforcement pairings may substan-tially improve vocal behavior, as might anumber of other proven speech therapytechniques. In considering speech as aviable response form, one should assessthe strength of the person's echoic reper-toire. If echoic behavior is moderate orstrong, then a vocal response form should


be pursued. Even a very small amount ofechoic behavior may be enough to immedi-ately get started with vocal mand training.If repeated attempts to establish vocal ver-bal behavior as mands or tacts fail, then anaugmentative system should be considered.Given the three alternative response

forms available for nonverbal individuals,how does one decide which system is mostappropriate for a specific individual? First,one must consider the individual and hisdisabling condition. If the nonverbal per-son is severely physically involved such asthose with CP and TBI, the differentialmuscle control necessary for speaking,signing, or writing may be impossible. Forthese individuals a pointing system or FCmay be the only options. However, inorder to use FC as the genuine speaker, theperson would have to be able to demon-strate literacy. Typing and spelling couldconceivably be taught through facilitation,but it would seem to take a large numberof training trials. In addition, the possibil-ity of automatic verbal behavior on thepart of the facilitator would have to beruled out, which would be difficultbecause it occurs so readily (Hall, 1993).Even if the person is literate, the facilitatorcould still engage in automatic verbalbehavior and control some of the typedmessages.

If the nonverbal person is literate, thenindependent writing or typing may be areasonable response form to use, especiallyif echoic and imitative behaviors are weak.Typically, the existence of these verbaloperants is accompanied by related mands,tacts, and intraverbals, however, for someindividuals the other operants might bequite weak. It is not unreasonable to usethe transcriptive mode to establish or fur-ther develop these other verbal operants.This may even be possible with facilitation,through an arduous shaping process. Butin some cases (e.g., CP and TBI), it is cer-tainly possible that independent verbaloperants could be acquired easier, or befurther developed with facilitation.However, if the person is not literate, ordoes not have pre-existing verbal operants,

then FC as a response form will probablynot be effective.When considering a system for nonver-

bal students who are not literate, norseverely physically involved (e.g., mostDD individuals, autistic children), the pri-mary choice is between pointing systemsand sign language. Given the many advan-tages of sign language as a topography-based system which parallels and possiblyimproves speech, and the many disadvan-tages of pointing systems as a selection-based system requiring environmentalsupport, sign language should be theresponse form of choice. However, it maybe the case that individual differences orhistories favor one system over another. Inaddition, it is possible that a blend of dif-ferent systems may be beneficial for someindividuals (Shafer, 1993), especially asthey become older and need a method tocommunicate with untrained audiences.However, the current trend is to favor

pointing systems over sign language(Shafer, 1993), primarily because it is easierfor the listener to understand what is beingpointed to by simply looking at the picture,symbol or written word touched. Signlanguage requires that the listener receivespecial training to effectively react as a lis-tener. This practice may be supported bycognitive theory which focus on whatwords mean to listeners, thereby placingthe emphasis on listener behavior ratherthan on speaker behavior (Skinner, 1957).This emphasis on the listener would favorpointing systems over sign languagebecause of the readily available supply oflisteners. Therefore, from this point of viewit would not be unreasonable to assumethat pointing would be easier to acquireand more functional than sign language.However, behavioral conceptual analyses(Michael, 1985; Shafer, 1993) and empiricalanalyses show that not only are signscloser to speech, but when all things areequal, sign language is easier for a studentto acquire than pointing at stimuli(Sundberg & Sundberg, 1990; Wraikat etal., 1991).

In conclusion, systems of augmentativecommunication have played a major role


in the development of successful verbalskills for nonverbal individuals. The deci-sion as to which type of augmentativesystem to use is often complicated by anindividual's specific needs and the generalcomplexities of the options available. Signlanguage is conceptually closest to speechand has the most advantages and fewestdisadvantages of the various options pre-sented, with pointing systems being thenext most preferred. Finally, it is clear thatonly a small number of nonverbal personscould benefit from facilitated communica-tion.

REFERENCESBiklen, D. (1990). Communication unbound: Autismand praxis. Harvard Educational Review, 60, 291-314.

Bijou, S. W., & Baer, D. M. (1965). Child Development II:Universal stage of infancy. Englewood Cliffs, NJ:Prentice-Hall.

Bondy, A. S., & Frost, L. A. (1993). Mands across thewater: A report on the application of the picture-exchange communication system in Peru. TheBehavior Analyst, 16, 123-128.

Bonvillian, J. D., & Nelson, K. E. (1978). Sign languageacquisition in a mute autistic boy. Journal of Speechand Hearing Disorders, 41, 339-347.

Bonvillian, J. D., Nelson, K .E., & Rhyne, J. (1981). Signlanguage and autism. Journal of Autism andDevelopmental Disorders, 1, 125-137.

Carr, E. (1979). Teaching autistic children to use signlanguage: Some research issues. Journal of Autismand Developmental Disorders, 9, 345-359.

Clarke, S., Remington, B., & Light, P. (1988). The roleof referential speech in sign language learning bymentally retarded children: A comparison of totalcommunication and sign-alone training. Journal ofApplied Behavior Analysis, 21, 419-426.

Drash, P. W., & Tudor, R. M. (1993). A functionalanalysis of verbal delay in preschool children:Implications for prevention and total recovery. TheAnalysis of Verbal Behavior, 11, 19-29.

Fristoe, M., & Lloyd, L. L. (1977). Manual communica-tion for the retarded and others with severe com-munication impairments: A resource list. MentalRetardation, 15, 18-21.

Green, G., & Shane, H. C. (1993). FacilitatedCommunication: The claims vs. the evidence.Harvard Mental Health Letter, 10, 4-5.

Hall, G. A. (1993). Facilitator control as automatic ver-bal behavior: A verbal behavior analysis. TheAnalysis of Verbal Behavior, 11, 89-98.

Hall, G. A., & Chase, P. N. (1991). The relationshipbetween stimulus equivalence and verbal behavior.The Analysis of Verbal Behavior, 9, 107-119.

Hall, G. A., & Sundberg, M. L. (1987). Teachingmands by manipulating conditioned establishingoperations. The Analysis of Verbal Behavior, 5, 41-53.

Hart B., & Risley T. R. (1975). Incidental teaching oflanguage in the preschool. Journal of AppliedBehavior Analysis, 8, 411-420.

Haskew, P., & Donnellan, A. M. (1992). Emotionalmaturity and well-being: Psychological lessons offacili-tated communication. Danbury CT: DRI Press.

Hodges, P., & Schwethelm, B. (1984). A comparison ofthe effectiveness of graphic symbol and manualsign training with profoundly retarded children.Applied Psycholinguistics, 5, 223-253.

Hurlbut, B. I., Iwata, B. A., & Green, J. D. (1982).Nonvocal language acquisition in adolescents withsevere physical disabilities: Blissymbol versusiconic symbol formats. Journal of Applied BehaviorAnalysis, 15, 241-258.

Keogh, W., & Reichle, J. (1985). Communication inter-vention for the "difficult to teach" severely handi-capped. In S. Warren & A. Rodgers-Warren (Eds.),Teachingfunctional language (pp. 157-196). Baltimore:University Park Press.

LaVigna, G. (1977). Communication training in mute,autistic adolescents using the written word. JournalofAutism and Childhood Schizophrenia, 7, 135-149.

Lovaas, 0. I. (1977). The autistic child: Language develop-ment through behavior modification, New York:Irvington Publishers.

McNaughton, S. (1976). Bliss symbol - an alternativesymbol system for the non-vocal pre-reading childin G. Vanderheiden & K. Gillery (Eds.), Non-vocalcommunication techniques and aids for the severelyphysically handicapped (pp. 85-104). Baltimore:University Park Press.

Michael, J. (1982). Skinner's elementary verbal rela-tions: Some new categories. The Analysis of VerbalBehavior, 1, 1-3.

Michael, J. (1983). Evocative and repertoire-alteringeffects of an environmental event. The Analysis ofVerbal Behavior, 2, 19-21.

Michael, J. (1985). Two kinds of verbal behavior plus apossible third. The Analysis of Verbal Behavior, 3, 1-4.

Michael, J. (1988). The establishing operation and themand. The Analysis of Verbal Behavior, 6, 3-9.

Michael, J. (1993). Establishing operations. TheBehavior Analyst, 16, 191-206.

Mirenda, P. (1985). Designing pictorial communica-tion systems for physically able-bodied studentswith severe handicaps. Augmentative and AlternativeCommunication, 1, 58-64.

Moores, D. F. (1978) Educating the deaf. Boston:Houghton Mifflin.

Partington, J. W., Sundberg, M. L., Newhouse, L., &Spengler, S. M. (in press). Overcoming an autisticchild's failure to acquire a tact repertoire. Journal ofApplied Behavior Analysis.

Peterson, N. (1978). An introduction to verbal behavior.Grand Rapids, MI: Behavior Associates.

Reichle, J., Sigafoos, J., & Remington. B. (1991).Beginning an augmentative communication systemwith individuals who have severe disabilities. In B.Remington, (Ed.), The challenge of severe mental hand-icap (pp. 189-213). Chichester: Wiley & Sons.

Reichle, J., York, J., & Sigafoos, J. (1991). Communi-cation programming for persons with severe handicaps:Vocal and augmentative strategies. Baltimore: Paul H.Brookes.

Rimland, B. (1964). Infantile Autism. New York:Appleton-Century-Crofts.

Romski, M. A., & Sevcik, R. A. (1988). Augmentativeand alternative communication systems: Consider-ations for individuals with severe intellectual dis-abilities. Augmentative and Alternative Communi-cation, 4, 83-98.


Rumbaugh, D. M. (1977). Language learning by a chim-panzee. New York: Academic.

Shafer, E. (1993). Teaching topography-based andstimulus selection-based verbal behavior to devel-opmentally disabled individuals: Some considera-tions. The Analysis of Verbal Behavior, 11, 117-134.

Sisson, L. A., & Barrett, R. P. (1983). A review of non-speech communication systems with autistic andmentally retarded individuals. In S. E. Breuning, J.L. Matson, & R. P. Barrett (Eds.), Advances in mentalretardation and developmental disabilities, Vol. 1 (pp.97-123). Greenwich, CT: JAI Press.

Skinner, B. F. (1957). Verbal behavior, EnglewoodCliffs, NJ: Prentice Hall.

Sundberg, C. T., & Sundberg, M. L. (1990). Comparingtopography-based verbal behavior with stimulusselection-based verbal behavior. The Analysis ofVerbal Behavior, 8, 31-41.

Sundberg, M. L. (1990). Teaching verbal behavior to thedevelopmentally disabled. Danville, CA: BehaviorAnalysts, Inc.

Sundberg, M. L., Milani, I., & Partington, J. W. (1977,August). The use of sign language with hearing men-tally impaired persons. Paper presented at the 85thAnnual meeting of the American PsychologicalAssociation, San Francisco, CA.

Sundberg, M. L., & Sundberg, C. A., & Partington, J.W. (1993, May). The use of automatic reinforcementprocedures to increase the verbal behavior of languagedelayed children. Paper presented at the 19th Annualmeeting of the Association for Behavior Analysis,Chicago, IL.

Van Riper, C. (1978). Speech correction principles andmethods (6th ed.). New York: Prentice-Hall.

Vanderheiden, G. C., & Lloyd. L. L. (1986).Communication systems and their components. InS. W. Blackstone (Ed.), Augmentative communication:Introduction (pp. 49-161). Rockville, MD: AmericanSpeech-Language-Hearing Association.

Vaughan, M. E., & Michael, J. (1982). Automatic rein-forcement: An important but ignored concept.Behaviorism, 10, 217-227.

Vernon, M., & Koh, S. D. (1970). Early manual com-munication and deaf children's achievement.American Annals of the Deaf, 115, 527-536.

Wheeler, D. L., Jacobson, J. W., Paglieri, R. A., &Schwartz, A. A. (1993). An experimental assessmentof facilitated communication. Mental Retardation, 31,49-60.

Wraikat, R., Sundberg, C. T., & Michael, J. (1991).Topography-based and selection-based verbalbehavior: A further comparison. The Analysis ofVerbal Behavior, 9, 1-17

Zangari, C., Lloyd, L. L., & Vicker, B. (in press).Augmentative and alternative communication: Anhistoric perspective. Augmentative and AlternativeCommunication.

Zwiebel, A. (1987). More on the effects of early man-ual communication on the cognitive developmentof deaf children. American Annals of the Deaf, 132,16-20.

Documents

Selecting Response Form for Nonverbal Persons: Facilitated … · 2019. 5. 22. · verbal children, especially for those who have very little vocal behavior under echoic control