words as behavior - university of maryland, baltimore countycatania/vb/acc avb words and...

The Analysis of Verbal Behavior

87

The article that follows, “Antecedents andConsequences of Word,” is reprinted with per-mission from the European Journal of Behav-ior Analysis. It is one more piece of an on-go-ing longer work. The article especially empha-sizes the role of multiple causation in verbalbehavior in the converging effects of verbalgovernance, verbal shaping, attention to ver-bal stimuli, and the replication of verbal be-havior. It considers a variety of contingencies,and by way of introduction it may be appro-priate to review them here in terms of anteced-ents, behavior and consequences. Table 1 clas-sifies three-term contingencies by sorting themon the basis of the nonverbal or verbal statusof antecedents and responses and the nonso-cial or social status of consequences. The tophalf of the table (I) shows cases we usuallyspeak of as instances of contingency-governedbehavior. When both antecedents and responsesare nonverbal, the behavior can be maintainedby either nonsocial consequences (A) or so-cial consequences (B). When the behavior isverbal (C and D), it takes social contingenciesto create classes of behavior that are occasionedby nonverbal antecedents (D), but once theyhave been established (as a higher-orderclasses) they may be maintained by nonsocial

contingencies (C), as when a particular situa-tion occasions a description of contingenciesthat alters someone’s subsequent behavior.

We speak of verbal governance whenever theantecedents are verbal (II), but the differentialcontingencies are most obvious when theymaintain nonverbal behavior (E and F). Again,the higher-order classes are established by so-cial contingencies, but once established suchverbal governance can be maintained either bynonsocial consequences (E), as when someonereaches a destination by following instructionsor makes a repair by following a servicemanual, or by social ones (F), as when some-one complies with a request or follows an or-der. These relations have been distinguishedby the respective terms tracking, for instruc-tion-following based on correspondences be-tween verbal behavior and environmentalevents, and pliance, for instruction-followingbased on social contingencies (Zettle & Hayes,1982).

Verbal governance allows a broad range ofpossible contingencies when verbal anteced-ents set the occasion for verbal behavior (Gand H), probably because verbal units can serveas any of the terms in a three-term contingency(for discussions of the symmetrical status of

Words as Behavior

A. Charles Catania, University of Maryland, Baltimore County

Table 1A classification of verbal and nonverbal contingencies.

Antecedents Behavior Consequences

I. Contingency-Governed BehaviorA. Nonsocial Behavior Nonverbal Nonverbal NonsocialB. Social Behavior Nonverbal Nonverbal SocialC. Verbal Mediation, Nonverbal Verbal Nonsocial

Self-ManagementD. Naming, Labeling, Nonverbal Verbal Social

Description

II. Verbally Governed BehaviorE. Tracking Verbal Nonverbal NonsocialF. Pliance Verbal Nonverbal SocialG. Logic, Calculation, Invention Verbal Verbal NonsocialH. Speaker-Listener Behavior Verbal Verbal Social

2006, 22, 87–88

88 A. CHARLES CATANIA

words as stimuli and as responses, see, for ex-ample, Horne & Lowe, 1996, on naming; andSidman, 1994, and Sidman, Wynne, Maguire,& Barnes, 1989, on equivalence classes). Par-alleling cases C and D, social consequences(H) must create the higher-order verbal classesthat may later enter into contingencies involv-ing nonsocial consequences (G). Social con-sequences, often but not necessarily also ver-bal, are the glue that holds together everydayconversation in both formal and informal set-tings. The contingencies that involve nonso-cial consequences of verbal governance (G),which probably are also prerequisites for thecases that involve verbal behavior occasionedby nonverbal antecedents, are of special inter-est because they are essential features of sci-ence and technology and other varieties of hu-man behavior that have extended the ways inwhich we act upon our environments. It is ofinterest that verbal shaping can operate on ver-bal behavior whether the antecedents are non-verbal or verbal, but social consequences aremore likely to be involved in such shaping thannonsocial ones.

This organization could of course be ex-tended to other dimensions of behavior, suchas the distinction between natural and artifi-cial contingencies or the different functionalproperties of local and higher-order classes orthe relative phylogenic and ontogenic contri-butions to antecedents and responses and rein-forcers. Furthermore, the boundaries betweenthese classifications are not sharp ones. For

example, a case can easily be made to extendthe categories of verbal governance listed astracking and pliance to cases where the relevantbehavior is verbal rather than nonverbal.

The phenomena treated in the article thatfollows are not theoretical. They are all phe-nomena that have been observed in everydayinstances of human verbal behavior and stud-ied in the analysis of verbal behavior. If theaccount is to be regarded as theoretical, it is soonly in its treatment of how these various phe-nomena come together to produce complexoutcomes.

REFERENCES

Horne, P. J., & Lowe, C. F. (1996). On the ori-gins of naming and other symbolic behav-ior. Journal of the Experimental Analysis ofBehavior, 65, 185–241.

Sidman, M. (1994). Equivalence relations andbehavior: A research story. Boston, MA:Authors Cooperative.

Sidman, M., Wynne, C. K., Maguire, R. W., &Barnes, T. (1989). Functional classes andequivalence relations. Journal of the Experi-mental Analysis of Behavior, 52, 261–274.

Zettle, R. D., & Hayes, S. C. (1982). Rule-gov-erned behavior: A potential theoreticalframework for cognitive-behavioral therapy.In Advances in cognitive behavioral re-search and therapy. Volume 1 (pp. 73–118).New York: Academic Press.

The Analysis of Verbal Behavior

89

The human species took a crucial step forwardwhen its vocal musculature came under oper-ant control in the production of speech sounds.Indeed, it is possible that all the distinctiveachievements of the species can be traced tothat one genetic change

—B. F. Skinner (1986, p. 117)

Behavior analysis is concerned with three-term contingencies, or the relations among an-tecedents and behavior and consequences. Torefer to both the antecedents and the conse-quences of words, as in the title of this paper,is implicitly to recognize words as instancesof behavior. A qualification is necessary, how-ever. Once words appear in written documentsor recorded speech, they have become artifactsof past behavior. Even as artifacts they mayeventually engender further behavior in theverbal responses of readers or listeners. It ishard to get away from words. We are sur-rounded by them through most of our waking

hours. Nevertheless it is difficult for us to thinkconsistently about them as behavior, and as aspecies we humans are only beginning to learnabout their functions.

This account considers how the interlock-ing contingencies that act upon verbal behav-ior can create coherent systems of behavior thatvary in the degree to which they are anchoredin social and nonsocial environments. Our fo-cus will be on four major topics: (1) verbalgovernance, in the contingencies, mainly so-cial, that lead not only to the following of in-structions but also to correspondences betweenwhat we do and what we say about what wedo; (2) replication, in the echoic and other pro-cesses that are prerequisites not only for theinitiation and maintenance of the verbal behav-ior of the individual but also for the spread ofverbal behavior throughout social communi-ties; (3) attention to verbal stimuli, in whichthe reinforcing and aversive properties of thesestimuli affect not only whether they will besought out or avoided but also whether theywill become incorporated into one’s own ver-bal behavior; and (4) verbal shaping, in thenatural and artificial contingencies that arrangeconsequences for verbal behavior and therebyraise or lower the probabilities of different ver-bal classes.

An essential feature of the account is thecombination of these processes in the multiple

Antecedents and Consequences of Words

A. Charles Catania, University of Maryland

As instances of behavior, words interact with environments. But they also interact with each other and withother kinds of behavior. Because of the interlocking nature of the contingencies into which words enter,their behavioral properties may become increasingly removed from nonverbal contingencies, and theirrelationship to those contingencies may become distorted by the social contingencies that maintain verbalbehavior. Verbal behavior is an exceedingly efficient way in which one organism can change the behaviorof another. All other functions of verbal behavior derive from this most basic function, sometimes calledverbal governance. Functional verbal antecedents in verbal governance may be extended across time andspace when individuals replicate the verbal behavior of others or their own verbal behavior. Differentialcontact with different verbal antecedents may follow from differential attention to verbal stimuli correlatedwith consequential events. Once in place, verbal behavior can be shaped by (usually social) consequences.Because these four verbal processes (verbal governance, replication, differential attention, and verbal shap-ing) share common stimulus and response terms, they produce interlocking contingencies in which exten-sive classes of behavior come to be dominated by verbal antecedents. Very different consequences followfrom verbal behavior depending on whether it is anchored to environmental events, as in scientific verbalpractices, or becomes independent of it, as in religious fundamentalism.

Key words: words, verbal governance, replication and selection, attention, shaping, social contingen-cies, scientific practice, religious fundamentalism.

This manuscript is an extension of arguments advancedin Catania (2003) and is based on presentations at the firstmeeting of the European Association of Behavior Analy-sis (Parma, Italy, July 2003) and at international sessionsof the annual meeting of the Association for BehaviorAnalysis (Boston, May 2004).

Address correspondence to A. Charles Catania, 10545Rivulet Row, Columbia, MD 21044-2420; phone: 410-455-3002; fax: 410-455-1055; e-mail: [email protected].

2006, 22, 89–100


causation of verbal and nonverbal behavior.The account is not a theory. Each of the fourtopics just listed involves common and readilydemonstrable properties of verbal behavior. Forexample, we observe verbal governance andreplication and differential attention and ver-bal shaping whenever we see individuals re-sponding to simple requests or repeating some-thing said or attending to one speaker ratherthan another or changing a topic of conversa-tion upon changes in the reactions of an audi-ence. To enumerate such effects is simply todescribe how verbal behavior works. The ac-count suggests how complex human behaviorcan sometimes be effectively interpreted interms of combinations of and interactionsamong these known properties.

It may also be useful to point out that theaccount does not appeal to meanings or ideasor other contents of words. The range of phe-nomena that can be addressed in terms of theverbal processes considered here suggests thatfor the purposes of analyzing and interpretingsome functions of verbal behavior those termsmight be expendable.

THE NATURE AND ORIGINS OF VERBAL BEHAVIOR

Verbal behavior includes any behavior inwhich the responses are words, without regardto whether the words are spoken, written orsigned. It involves both speaker behaviorshaped by its effects on the behavior of listen-ers and listener behavior shaped by its effectson the behavior of speakers. The functionalunits of verbal behavior (such as phonemes andwords and grammatical markers) are main-tained by the practices of verbal communities.

The most basic consequence of verbal be-havior is that through it speakers change thebehavior of listeners. In other words, verbalbehavior is a way to get people to do things; itis “effective only through the mediation ofother persons” (Skinner, 1957, p. 2). Some-times its effects are nonverbal, as when we asksomeone to do something; sometimes its ef-fects are verbal, as when we change what some-one has to say about something. This functionis primary because other functions gain theirsignificance only through it. Transmitting in-formation and describing feelings are functionsof language, but these secondary functionsmatter only if they sometimes make a differ-ence by changing the behavior of others. For

example, we transmit information or conveyour thoughts because, as a consequence, oth-ers may act upon them, and we describe ourfeelings and express our emotions because, asa consequence, others may behave differentlytoward us. The thoughts or feelings or emo-tions do not travel from the speaker to the lis-tener. Only the words do.

Speakers can lead others to say things as wellas to do things, and whether an instance of ver-bal behavior functions as an instruction doesnot depend on its grammatical form. Giving adefinition or stating a fact instructs with respectto verbal behavior just as giving an order in-structs with respect to nonverbal, and all utter-ances are, usually in multiple senses, ways oftelling someone else what to do.

Verbal behavior allows individuals to actupon stimuli available to others but not to them-selves, as when one monkey’s vocal call al-lows another to escape from a predator it hadnot seen (Seyfarth, Cheney, & Marler, 1980).But here too the irreducible function is that thebehavior of the second monkey is changed bythe vocal behavior of the first. An account ofthe ways in which such contingencies mightevolve in complexity is beyond the scope ofthis paper, but consider the potential impact onreproductive success of just a single utterance,corresponding in its effect to that of the con-temporary word “stop,” that reduces the likeli-hood that an out-of-reach offspring will wan-der off into danger. From such a start, a mini-mal vocabulary with relatively simple effects(keeping together during movement, coordi-nating aggression or flight, etc.) could evolveover millennia into a richly differentiated rep-ertory involving the behavior of mothers andoffspring, mates, and other subgroups of homi-nid social units (for a more detailed account,see Catania, 2003).

Verbal behavior can emerge only in organ-isms whose behavior is or can become sensi-tive to social contingencies. Any account of theorigins and evolution of language must be con-sistent with selection as it operates at three dif-ferent levels: at the level of phylogeny, as popu-lations of individuals (and their genes) are se-lected by evolutionary contingencies; at thelevel of ontogeny, as populations of responsesare selected by their consequences over the life-time of individual organisms; and at the levelof cultural practices or memes, as social con-tingencies select behavior as it is passed on

91ANTECEDENTS AND CONSEQUENCES OF WORDS

from one organism to another (Darwin, 1859,1871; Dawkins, 1976, 1982; Skinner, 1981).The evolution of the particular anatomical andphysiological features of human vocalization,such as the structure of the vocal tract and thefunctional organization of respiration and ar-ticulation in operant classes, was no doubt cru-cial, but human languages could arise onlywhen verbal behavior began to be replicatednot only in individuals during their own life-times as speakers and listeners but also amongthe different speakers and listeners of socialgroups within and across successive genera-tions (Catania, 1985, 1991, 1994, 2001).

This account will appeal only occasionallyto Skinner’s functional verbal units in VerbalBehavior (Skinner, 1957), but as in that bookit will give considerable emphasis to the mul-tiple causation of verbal behavior. A given ver-bal response is jointly determined by manyvariables, including but not limited to nonver-bal discriminative stimuli, earlier verbal re-sponses, prior reinforcing or aversive conse-quences of related responses, the nature of theaudience, and the condition of the speaker. Inhis treatment of multiple causation, Skinnerdescribed cases in which novel behavioremerges from the novel combination of vari-ables as “a different type of multiple control,in which functional relations, established sepa-rately, combine possibly for the first time upona given occasion” (Skinner, 1957, p. 229). Forexample, two or more newly learned wordsmay appear together for the first time in achild’s sentence that the child has never heardor uttered before. The coming together of ex-isting responses in novel combinations to pro-duce new behavior is sometimes called adduc-tion.

Skinner also gave special consideration tothe hierarchical organization of verbal behav-ior, as when smaller verbal units such as pho-nemes or letters combine in larger units calledwords and words combine in still larger unitscalled sentences. It was implicit in his accountthat the contingencies operating on operantclasses at one level of analysis need not be con-sistent with those operating on the higher-or-der classes at another level. For example, thecontingencies that determine the thematic fea-tures of an utterance or its grammatical formmay be different from those that determine itspronunciation. A higher-order class is one thatincludes within it other classes that can them-

selves function as operant classes, as whengeneralized imitation includes specific imita-tions each of which can be separately reinforcedas a subclass. The emergence of novel instancesis one criterion for distinguishing betweenhigher-order classes and sets of specific cases.But when different contingencies operate for ahigher-order class than for the subclasseswithin it, behavior can seem to be insensitiveto one or the other set of contingencies(Catania, 1995); opportunities can be missedby accounts that do not explicitly consider thedifferent contingencies that can operate at dif-ferent levels (e.g., Hayes, Barnes-Holmes, &Roche, 2001).

We are now ready to consider the cases ofverbal governance, replication, attention toverbal stimuli, and the shaping of verbal be-havior. Each of these is likely to be alreadyfamiliar, so each will be outlined only brieflyin preparation for a discussion of their interac-tions.

1. Verbal Governance

Verbally governed behavior is behavior, ei-ther verbal or nonverbal, determined by ver-bal antecedents (it has been called rule-gov-erned behavior, but definitions of rules areproblematic because they are sometimes basedon structural criteria and sometimes on func-tional ones). Such behavior is maintained notso much by consequences arranged for particu-lar responses given particular verbal stimuli,but rather by social contingencies that gener-ate higher-order classes of behavior character-ized by correspondences between verbal ante-cedents and subsequent behavior. Higher-or-der classes of behavior are held together bythe common contingencies shared by theirmembers, just as the various topographies of arat’s food-reinforced lever pressing (e.g., leftpaw, right paw, both paws) are held togetherby the common contingencies according towhich they produce food. Verbal behavior in-cludes other higher-order classes along withverbal governance; one example is naming, asin Horne & Lowe (1996).

Within higher-order classes, relations be-tween responses and their consequences at thehigher-order level need not be compatible withthose at the more local level of the subclasses.In the case of verbal governance, social con-tingencies maintain the higher-order class, but


other contingencies act on specific instances.In the military, for example, social contingen-cies maintaining obedience may conflict withnonsocial contingencies prevailing on thebattlefield, as when the aversive nonsocial con-sequences of advancing under fire oppose theaversive social consequences of retreat. Whensocial consequences prevail, the componentresponses remain consistent with the higher-order class (orders are obeyed) and the behav-ior is called verbally governed; it is relativelyinsensitive to the nonsocial contingencies.When nonsocial consequences dominate (re-treat occurs against orders), the behavior iscalled contingency-governed or contingency-shaped. The former usage emphasizes main-taining contingencies and the latter emphasizesorigins (cf. Skinner, 1969).

The vocabulary of verbal governance andcontingency governance is convenient, but itis important to recognize that contingenciesoperate at both levels, the higher-order and thelocal levels, and that the distinction betweenhigher-order and local levels is orthogonal tothat between verbal and nonverbal behavior.In other words, verbal behavior is defined bycertain social contingencies, but such contin-gencies can operate either on higher-orderclasses or locally on specific subclasses or onboth. Verbally governed behavior is often de-termined more strongly by higher-order socialcontingencies than by more local (often non-social) contingencies, and therefore is often lesslikely than contingency-governed behavior tochange when the local contingencies change.We do not ordinarily tell people to do what theywould do even without being told.

Verbal antecedents in verbal governance aremost obvious in the form called instructions,as when we are told to do or say something.But “I’m thirsty,” a declarative, may have thesame function as “May I have a drink of wa-ter?,” an imperative. Furthermore, verbal an-tecedents may specify contingencies, as whenwe are told what will happen if we do or saysomething, or they may specify behavior, aswhen we are simply told what to do.

Verbal governance, like any other operantclass, may be conditional on other events. Hu-mans learn to follow instructions from someindividuals but not others. Humans often talkto themselves, usually silently but occasion-ally, as when following complex instructions,out loud. Thus, verbal antecedents may lead to

other verbal behavior, as when implications andcourses of action are derived from somethingsaid. And because humans can often distinguishbetween what they have been told and whatthey have arrived at without being told, themost effective verbal antecedents may be thosethat they generate themselves. In such cases,they may fail to recognize the remote originsof what they generated, in the verbal behaviorof others that initiated their own self-talk.

Local contingencies may also generate ver-bal behavior that in turn produces nonverbalbehavior consistent with natural consequences(as when making calculations based on mea-surements). Behavior then may become sensi-tive to its consequences because an effectiveverbal account of contingencies and relatedperformances has been formulated. But suchcases still qualify as instances of verbal gover-nance, and the indirect sensitivity to contin-gencies, mediated by verbal behavior, remainsfunctionally different from that of contingency-governed behavior (Shimoff & Catania, 1998;Skinner, 1969).

Ironically, one class of human behavior morelikely to be locally than verbally governed isverbal behavior itself. Our everyday languagedoes not include an effective vocabulary deal-ing with the functional properties of our ownverbal behavior, so we rarely talk about thevariables that determine it. In other words,many properties of verbal behavior are typi-cally not verbally governed. That may be whythe shaping of human verbal behavior is ofteneasier than the shaping of human nonverbalbehavior. But verbal shaping is discussed fur-ther below. For the moment it is sufficient todisplay verbal governance paradigmatically:Verbal discriminative stimuli set the occasionfor verbal and nonverbal responses, and theseresponses may have consequences:

Verbal Governance

VERBAL SD’s VERBAL/NONVERBAL R’s

VERBAL/NONVERBAL R’s CONSEQUENCES

2. Replication

We tend to repeat what we and others say.But this replication is not mere reproduction.The child who echoes what a parent says du-


plicates not the acoustic properties of theparent’s utterance but rather a particular se-quence of phonemes the properties of whichdepend on the arbitrary vocal practices of aparticular verbal community. The differentphonetic units of different human languagesare acquired by speakers over long individualhistories in which complex articulations areshaped by the differential consequences ofvocalization. Included among these conse-quences are the degree of correspondence be-tween the sounds individuals produce them-selves and those produced by others and espe-cially those produced by caregivers. The latterare likely to become significant because theyhave been correlated with significant interac-tions between the individual and caregivers orothers. The role of such correspondences asreinforcers that shape the various details of theverbal behavior of young speakers during theacquisition of language has been extensivelytreated elsewhere (e.g., Catania, 1998, pp. 241–246; Palmer, 1996, 1998; Risley, 1977; Skin-ner, 1957, p. 58).

But we must deal with more than speech.Based on a long and complex history in whichwe learn the relations among spoken and writ-ten stimuli and responses, we say that wordsare the same whether they are in the auditoryor visual mode. We repeat what others say orcopy what they have written, but we also speakwhat has been written or copy down what hasbeen said. For the present purposes, these areall instances of replication, even though theysometimes carry across different modalities.

The main relevance of replication viewed inthis way is that these practices allow the ef-fects of verbal stimuli to be extended over timeand space as the verbal behavior produced bysome individuals is passed on to others. Suchreplication is not merely a prerequisite for lan-guage development. It remains importantthroughout life, as when we repeat somethingwe have heard or jot down a note about it.

Replication is crucial to the maintenance ofverbal behavior both in individuals and in ver-bal communities: once some individuals beginrepeating what they or others say, verbal be-havior may be maintained by cultural contin-gencies and therefore survive across genera-tions (cf. Blackmore, 1999). On the other hand,evolutionary contingencies are likely to favorsystems in which replication can serve addi-tional functions over those in which replica-

tion does not. Some effects of replication arefairly simple and straightforward. For example,if a single utterance does not produce charac-teristic effects on the behavior of a listener, oneor more repetitions may do so; the summationof the effects of repeated stimuli is a familiarphenomenon. But if the replicated verbal re-sponse also participates in verbal governance,the listener’s replication of the speaker’s ver-bal behavior extends the influence of thespeaker, as when instructions are passed alongfrom one member to another in a large group.Furthermore, the listener’s repetition may cre-ate conditions under which instructions maybe followed in the speaker’s absence, later andelsewhere, in effect transferring governancefrom the speaker’s verbal behavior to thelistener’s replication of it, as when we repeatto ourselves the details of a task that someonehas asked us to complete. The continuing rep-lication of the listener’s own verbal behaviormay therefore create conditions under whichthe effects of verbal governance become ex-tended over time and space, even in the ab-sence of the speaker who originally producedthe verbal behavior (cf. Jaynes, 1976).

Replication has to come first, but once inplace powerful contingencies can maintain it.The effects are potentially far reaching, thoughtheir paradigmatic display is simple: verbaldiscriminative stimuli set the occasion for ver-bal responses, and those responses can func-tion in turn as verbal discriminative stimuli:

Replication of Verbal Behavior

VERBAL SD’s VERBAL R’sVERBAL R’s VERBAL SD’s

3. Attention to Verbal Stimuli

Some discriminative stimuli are correlatedwith the delivery of reinforcers and others withperiods of extinction. Only the former are likelyto acquire reinforcing functions of their own;when they do so they are called conditional orconditioned reinforcers. The effectiveness ofdiscriminative stimuli as occasions for behav-ior depends on whether the organism attendsto them, which in turn depends on their statusas conditional reinforcers (Dinsmoor, 1983,1995). Looking at or attending to a stimulus isreinforced when the stimulus is correlated withreinforcers but not otherwise. An informative


stimulus is also a discriminative stimulus, inthe sense that different contingencies operatewhen the stimulus is present than when it isabsent. It is not necessarily a conditional rein-forcer, however, so informativeness alone is notsufficient to maintain attention to discrimina-tive stimuli. A discriminative stimulus corre-lated solely with extinction or with aversiveevents will not function as a conditional rein-forcer and therefore may not maintain atten-tion. Experiments on attention typically makethe behavior of attending explicit by introduc-ing an observing response, a response that pro-duces discriminative stimuli that would other-wise be unavailable, and then record changesin the rate of this response as a function of thecorrelations of discriminative stimuli with vari-ous reinforcing or aversive contingencies(Kelleher, Riddle, & Cook, 1962).

In other words, organisms do not attend tostimuli because they are informative. Instead,they attend to informative stimuli only if thosestimuli are conditional reinforcers. This find-ing undercuts the appeal to information pro-cessing as a primary cognitive process. It alsohas implications for what happens when stimuliare verbal, because it follows that the effec-tiveness of a message depends more on whetherits content is reinforcing or aversive than onwhether it is correct or complete. The phenom-enon has long been recognized in folklore, asin accounts of the unhappy treatment of mes-sengers who bring bad news. In fact, whatneeds explanation is that humans attend at allto bad news or that sometimes they reach con-clusions even when the answer is not what theywanted to hear. It is presumably relevant thatbad news does sometimes allow effectiveavoidance behavior and that many stimuli arecorrelated with sufficient reinforcers that theycan maintain attention even when they are alsocorrelated with occasional aversive events.

Attention to a verbal stimulus is a prerequi-site for its replication or for governance by thatstimulus, and the consequences of either vari-ety of behavior may change the likelihood ofsubsequent responses to that and similar ver-bal stimuli. We repeat to ourselves what wehave heard or read, and once we have done soour own verbal behavior is likely to summatewith other verbal antecedents that participatein the verbal governance of our behavior. Ifthese verbal stimuli were reinforcing in the firstplace, then behavior producing contact with

more or similar instances will be strengthened,in turn providing even more related initiatinginstances. It is no surprise, for example, thatpeople turn to news media that are biased to-ward political views they already hold.

The responses that provide access to verbalstimuli may themselves be either verbal or non-verbal. For example, at a newsstand we canget a newspaper either by asking for one or bypicking one up. In the paradigmatic display ofattention to verbal stimuli, verbal or nonverbalresponses produce verbal discriminativestimuli, and these in turn produce other verbalor nonverbal responses:

Differential Attention to Verbal Stimuli

VERBAL/NONVERBAL R’s VERBAL SD’s

VERBAL SD’s VERBAL/NONVERBAL R’s

4. Verbal Shaping

We are most familiar with examples of shap-ing from the laboratory, but natural contingen-cies may produce shaping, as when differen-tial attention by a parent inadvertently shapesan infant’s annoying cries. Shaping is effec-tive because behavior is variable. Reinforce-ment of a response produces a spectrum of re-sponses that differ from the reinforced responsealong dimensions such as topography and mag-nitude. Verbal shaping is of special interestbecause verbal behavior can also vary alongsemantic and syntactic dimensions. Demonstra-tions of the shaping of verbal behavior havean extensive history (e.g., Greenspoon, 1955;Lovaas, 1964; Rosenfeld & Baer, 1970).

Once verbal behavior has been shaped, itmay participate in other verbal functions, suchas verbal governance. When verbal governanceinvolves an individual’s own verbal behavior,it can make a difference whether that verbalbehavior has been shaped or instructed. Thedifference has been studied in the context ofhuman performances maintained by schedulesof reinforcement (e.g., Catania, Lowe, &Horne, 1990; Catania, Matthews, & Shimoff,1982; Catania, Shimoff, & Matthews, 1989;Shimoff & Catania, 1998), where nonverbalbehavior is sometimes more easily changed bythe shaping of relevant verbal behavior than


by changes in the contingencies arranged forthe nonverbal behavior itself.

Just as verbal governance by the instructionsgiven by others can be conditional on othervariables, so also for verbal governance byone’s own shaped verbal behavior. For ex-ample, it makes a difference whether what isshaped describes behavior or describes contin-gencies operating for that behavior (Catania,Shimoff, & Matthews, 1989). Descriptions ofwhat is relevant behavior in a given environ-ment are not equivalent to descriptions of howthat environment works.

The shaping of verbal behavior is a potenttechnique for changing human behavior. Apractical implication is that shaping whatpeople say about their own behavior may be amore effective way to change their behaviorthan either shaping their behavior directly ortelling them what to do, perhaps because ver-bal communities arrange contingencies for cor-respondences between saying and doing, as inthose for lying versus telling the truth or forkeeping versus breaking promises (e.g., Baer,Detrich, & Weninger, 1988; Risley & Hart,1968).

The fact of verbal shaping implies that ver-bal behavior itself is sensitive to local contin-gencies even when it governs other behavior.The greater effectiveness of verbal shapingrelative to instructions may depend in part onthe speaker’s failure to discriminate the sourcesof the verbal behavior. In other words, evenspeakers who accurately discriminate amongvarious sources of verbal behavior when fol-lowing instructions usually call their ownshaped verbal behavior self-generated. Tochange beliefs is to change verbal behavior,but it matters whether we have been told whatto say or have come to say it in other ways (cf.Rosenfarb, Newland, Brannon, & Howey,1992).

Audiences provide discriminative stimulithat set the occasions on which verbal behav-ior may have consequences and provide rein-forcers that shape verbal behavior. Differentaudiences set the occasion for different verbalclasses. Verbal shaping has been studied inexperimental contexts, but anyone who hasobserved drifts in the content of conversationas attention to particular topics or speakerspicks up or flags has seen it in a natural set-ting. Verbal shaping is difficult to track in natu-ral environments because a wide range of so-

cial reinforcers enters into it (e.g., eye contact,changes in facial expression or posture, con-tinued verbal behavior) and their effectivenesschanges over time (e.g., a comment that worksearly in a conversation may be totally irrelevantlater on).

Verbal behavior is typically shaped by so-cial contingencies, but nonsocial contingenciesmay also be effective, as when an engineer’scalculations lead to successive changes in thespecifications of a project or as when a writeredits successive drafts of a manuscript. Suchinstances, however, clearly depend on an ex-tensive social history of behaving verbally, soit could be argued that the interactions in thebehavior of the individual as both writer andreader are functionally similar to the interac-tions that occur when writer and reader are dif-ferent individuals.

Shaping typically involves quantitativechanges along one or more dimensions of anorganism’s behavior, but sometimes it producesqualitative changes, analogous to a horsechanging from one gait to another as it increasesits running speed (cf. Catania & Harnad, 1988,p. 476). Similarly, the gradual changes pro-duced by selection relative to a populationmean in phylogeny have been contrasted withmore abrupt changes (sometimes called salta-tions) produced by large-scale changes in theenvironment. Analogies in the shaping of ver-bal behavior by natural contingencies might bethose dramatic changes called religious con-versions or epiphanies, when an accumulationof small changes is followed suddenly by ma-jor shifts over a range of verbal and nonverbalclasses. In any case, the paradigmatic displayof verbal shaping illustrates not only that ver-bal responses have consequences but also thatthe new verbal responses shaped by these con-sequences may then function as verbal dis-criminative stimuli:

Verbal Shaping

VERBAL R’s CONSEQUENCESNEW VERBAL R’S VERBAL SD’s

THE COHERENCE OF VERBAL CLASSES

We have just considered verbal shaping, andit is relevant to note that some of the most in-teresting reinforcers of verbal behavior are


themselves verbal: an answer to a question, anacknowledgment, a continuation of a line ofthought, and so on. But if some verbal conse-quences are more effective reinforcers thanothers, it follows that some command moreattention than others. Verbal stimuli are dis-criminative stimuli, and we attend to them, aswe attend to nonverbal ones, not on the basisof the information they carry but rather as afunction of their correlation with reinforcers.Having attended to them we may replicatethem, but in the course of successive replica-tions they may be subject to further shaping.And the newly shaped verbal behavior may alsobegin to participate in verbal governance. Inthis brief compass we have already touched oneach of the four verbal properties that have beenconsidered. Here they are again, but with aslight difference. These paradigms do not in-clude the nonverbal components.

Verbal Governance

VERBAL SD’s VERBAL R’sVERBAL R’s CONSEQUENCES

Replication of Verbal Behavior

VERBAL SD’s VERBAL R’sVERBAL R’s VERBAL SD’s

Differential Attention to Verbal Stimuli

VERBAL R’s VERBAL SD’sVERBAL SD’s VERBAL R’s

Verbal Shaping

VERBAL R’s CONSEQUENCESNEW VERBAL R’S VERBAL SD’s

The point is that the same sorts of terms en-ter into each of the categories. It has been cru-cial to distinguish among verbal and nonver-bal antecedents and consequences in the courseof this account, because verbal units can serveas any term in the three-term contingency. Simi-lar relations can be generated for nonverbalstimuli and responses, but they cannot soreadily exchange their positions (we do notworry about a pigeon greening in the presenceof pecks or a rat toning in the presence of leverpresses). A verbal discriminative stimulus thatparticipates in verbal governance can be repli-

cated, it can command differential attention,and its replications can be shaped by their con-sequences. Verbal behavior that has been rep-licated can participate in verbal governance, itcan command differential attention, and it canbe shaped by its consequences. A verbal stimu-lus that commands differential attention can bereplicated, it can participate in verbal gover-nance, and its replications can be shaped bytheir consequences. A verbal response that isshaped is replicated in the course of its shap-ing, and it can function as a verbal stimulusthat commands differential attention and thatparticipates in verbal governance. But if theconsequences that maintain these aspects ofverbal behavior are primarily social, then ver-bal communities can create and maintain ver-bal classes that have ever diminishing contactwith nonverbal contingencies. In other words,they can create verbal worlds that become in-creasingly autonomous.

The individual who generates varied verbalstimuli some of which are more potent rein-forcers than others may be producing condi-tions for automatic verbal shaping. Thus, dif-ferential attention to verbal stimuli may leadto self-generated verbal behavior that is some-what independent of current environments andthat may become a pervasive feature of anindividual’s behavior across a range of differ-ent situations. Skinner (1957) provides ex-amples in the areas of poetry and other literaryverbal behavior. As contact with current ante-cedents and contingencies becomes more tenu-ous (it is after all difficult to disprove claimsof heaven and hell, parallel universes and alienabduction) and the circle of speakers and lis-teners becomes more limited, higher-order ver-bal contingencies may generate idiosyncraticverbal interactions that are maintained in thebehavior of isolated individuals or smallgroups, as even the social consequences avail-able in the broader verbal community becomeless relevant. When such concentrated effectsshow up in the behavior of individuals, theyare sometimes called interests or obsessions;when they extend across groups, they are some-times called fads or cults. Though such behav-ior has its own internal coherence, it may ap-pear incoherent to those with other verbal his-tories. In such cases, verbal behavior seems tostand between the individual and the contin-gencies of the nonverbal environment.

Figure 1 illustrates these interactions. A flow


Fig. 1. Schematic display of potential interactions among four verbal processes: verbal governance, verbal replica-tion, differential attention to verbal stimuli, and verbal shaping. These processes may operate not only as differentspeakers and listeners interact with each other but also within the verbal behavior of an individual speaker/listener.

chart representing a sample behavior streamin which the relations among the various ver-bal stimuli and responses of a small number ofspeakers and listeners were labeled accordingto our several categories might have been at-tempted, but it could not have done justice tothe different degrees in which particular stimulior responses might share multiple functions ormight vary in the magnitudes of their contri-butions to each verbal relation. This figure mustsuffice. It shows that each process interacts withevery other one as well as with itself, and thebottom frame suggests that such verbal func-tions can have their effects not only on the ver-bal behavior that passes between differentmembers of a verbal community but also onthe verbal behavior that is maintained in therepertory of the individual speaker or listener.

The parts of Figure 1 do not merely fit welltogether; they also become components ofcycles within which they strengthen and buildupon each other. Omit any one and the cyclicityis significantly attenuated. Even if the socialreinforcers that participate in these functionsare ordinarily small, they operate on behaviorday in and day out over weeks and months andyears. We know how much behavior we canshape in just a few minutes with a nonverbalorganism, so should we be surprised at the ef-fects of interlocking verbal contingencies op-erating over human lifetimes? These are pow-erful variables, and it is easy to imagine howthey could sometimes produce verbal behav-ior that is dull and highly stereotyped and some-times verbal behavior that is flexible and highlycreative. Verbal governance and verbal repli-


cation and differential attention to verbalstimuli and verbal shaping make their ownseparate contributions, but their synergisticeffects produce complex behavior that is muchmore than just the sum of its parts.

With regard to nonverbal contingencies, ver-bal behavior need not always get in the way. Ittakes social contingencies to create higher-or-der classes of verbal behavior occasioned bynonverbal antecedents, but such classes, onceestablished, may be maintained by nonsocialcontingencies, as when a situation occasions adescription of contingencies that alterssomeone’s subsequent nonverbal behavior.Similarly, governance by verbal antecedents isestablished by social contingencies, but onceestablished it can be maintained either by non-social consequences, as when someone makesa repair by following a service manual, or bysocial ones, as when someone complies with arequest. Although social consequences holdeveryday conversation together, the contingen-cies that involve nonsocial consequences ofverbal governance are essential features of sci-ence and technology and other extensions ofthe ways in which we act upon our environ-ments. Its anchoring to nonverbal environmentsthrough data is the special advantage of scien-tific verbal behavior.

CONTINGENCIES IN POLITICS AND RELIGION

Those who become engrossed in some kindsof artificial verbal worlds (e.g., those of DonQuixote or Frodo Baggins or Harry Potter) mayget themselves into trouble by neglecting othercontingencies, but they might not otherwisemake much trouble for others. That is not in-evitably the case, however. For example, anindividual with a history of replicating verbalbehavior in the recitation of religious texts maylater differentially attend to a speaker whoseverbal behavior is consistent with those texts.That speaker may then shape new verbal be-havior and in the course of doing so may alsobecome a potent source of verbal governance.Is it too far-fetched to suggest that such an in-dividual might be induced to hijack a commer-cial airliner and fly it into an office tower or amilitary structure?

In fundamentalism, the word is the last word.It cannot exist without verbal control. It ismaintained through the shaping of verbal be-havior that is consistent with the replication of

sacred texts that command differential atten-tion and that participate in verbal governance.Given what we know about verbal governance,acts in the name of religion that some call ter-rorism and others call martyrdom should comeas no surprise. Given what we know about ver-bal replication, the political endorsement ofreligious speech should come as no surprise.Given what we know about differential atten-tion to verbal stimuli, the suppression of theteaching of evolution or of Darwin’s theory ofnatural selection by those who endorse the lit-eral truth of the biblical story of creation shouldcome as no surprise. Given what we knowabout verbal shaping, the vast diversity of hu-man religions should come as no surprise.

Once speakers could instruct the verbal be-havior of listeners who could in turn instructnonverbal behavior, the prerequisites for hu-man political and religious institutions werefirmly in place. The invention of writing, per-haps initially a matter of record-keeping,moved verbal governance further from the be-havior of individual speakers. Human behav-ior throughout the world has been and still isheavily influenced by records of long-past ver-bal behavior. We need only list a few amongmany: The Analects of Confucius; The OldTestament; Tao Te Ching; The New Testament;The Brahma-Sutras; The Koran; The AdiGranth; Bhagavad-Gita; The Book of Mormon.Even within these examples, disputes may rageover which version of a text is legitimate (thestatus of the King James version of the Bibleprovides one obvious example).

In “Science and Human Behavior,” Skinner(1953) treated the areas of government andreligion as separate topics. Both, however, en-tail strong elements of verbal governance. It isuseful to distinguish the kinds of consequencesthat governmental institutions can bring to bearfrom those available to religious institutions,but whether reinforcers are promised and pun-ishers are threatened in this life or in an after-life, both the promises and the threats dependon a verbal history. Exhortations to behaviorin the name of patriotism are not so very dif-ferent from exhortations in the name of a de-ity. In their reliance on verbal governance, thetwo sorts of institutions have much in common.Again we need only list a few documentsamong many: the Code of Hammurabi; MeinKampf; the Communist Manifesto; the Napo-leonic Code; the Code of Bushido; the Talmud;


the Little Red Book of Chairman Mao; theDeclaration of Independence; the Charter ofthe United Nations. The details of the socialcontingencies that maintain or have maintainedverbal governance are somewhat less obviousin these cases than in instances of religiousverbal governance, and those details differ inimportant ways from one instance to another.Perhaps a case can be made that some of thosedetails are more likely than others to be con-sistent with the survival of cultures.

We have seen that verbal behavior can betightly determined by nonverbal environmen-tal contingencies, as in scientific practices, orloosely determined, as in social practices suchas literature and religion. Like government,religion has its roots in social control. Religiousbehavior provides compelling examples of thephenomena reviewed here, in verbal gover-nance as demonstrated by the following of re-ligious precepts, in the replication of verbalbehavior through recitations of sacred scrip-ture, in differential attention to prescribed andproscribed texts, and in the shaping of verbalbehavior by religious leaders in both informalinteractions with constituents and in formalones such as confessions or inquisitions.Heaven and hell, like angels and devils, arehuman verbal creations. But remember: Any-thing that threatens them and their interlock-ing verbal contingencies is sometimes calledblasphemy or heresy.

REFERENCES

Baer, R. A., Detrich, R., & Weninger, J. M.(1988). On the functional role of the ver-balization in correspondence training pro-cedures. Journal of Applied Behavior Analy-sis, 21, 345–356.

Blackmore, S. (1999). The meme machine.New York: Oxford.

Catania, A. C. (1985). Rule-governed behav-ior and the origins of language. In C.Bradshaw (Ed.), Behavior analysis and con-temporary psychology (pp. 135–156).Hillsdale, NJ: Erlbaum.

Catania, A. C. (1991). The phylogeny and on-togeny of language function. In M. Studdert-Kennedy (Ed.), Biological and behavioraldeterminants of language development (pp.263–285). Hillsdale, NJ: Erlbaum.

Catania, A. C. (1994). The natural and artifi-cial selection of verbal behavior. In K. Ono

(Ed.), Behavior analysis of language andcognition (pp. 31–49). Reno, NV: Context.

Catania, A. C. (1995). Higher-order behaviorclasses: Contingencies, beliefs, and verbalbehavior. Journal of Behavior Therapy andExperimental Psychiatry, 26, 191–200.

Catania, A. C. (1998). Learning (4th ed.). Up-per Saddle River, NJ: Prentice-Hall.

Catania, A. C. (2001). Three varieties of selec-tion and their implications for the originsof language. In G. Györi (Ed.), Languageevolution: Biological, linguistic and philo-sophical perspectives (pp. 55–71). NewYork/Bern: Peter Lang.

Catania, A. C. (2003). Verbal governance, ver-bal shaping, and attention to verbal stimuli.In P. N. Chase (Ed.), Behavior theory andphilosophy (pp. 301–321). New York:Kluwer/Academic Press.

Catania, A. C., & Harnad, S. (Eds.). (1988).The selection of behavior: The operant be-haviorism of B. F. Skinner. New York: Cam-bridge University Press.

Catania, A. C., Lowe, C. F., & Horne, P. (1990).Nonverbal behavior correlated with theshaped verbal behavior of children. TheAnalysis of Verbal Behavior, 8, 43–55.

Catania, A. C., Matthews, B. A., & Shimoff,E. (1982). Instructed versus shaped humanverbal behavior: Interactions with nonver-bal responding. Journal of the Experimen-tal Analysis of Behavior, 38, 233–248.

Catania, A. C., Shimoff, E., & Matthews, B.A. (1989). An experimental analysis of rule-governed behavior. In S. C. Hayes (Ed.),Rule-governed behavior: Cognition, contin-gencies, and instructional control (pp. 119–150). New York: Plenum.

Darwin, C. (1859). On the origin of species bymeans of natural selection. London: JohnMurray.

Darwin, C. (1871). The descent of man. Lon-don: John Murray.

Dawkins, R. (1976). The selfish gene. NewYork: Oxford University Press.

Dawkins, R. (1982). The extended phenotype.San Francisco: Freeman.

Dinsmoor, J. A. (1983). Observing and condi-tioned reinforcement. Behavioral and BrainSciences, 6, 693–728.

Dinsmoor, J. A. (1995). Stimulus control. TheBehavior Analyst, 18, 51–68, 253–269.

Greenspoon, J. (1955). The reinforcing effectof two spoken sounds on the frequency of


two responses. American Journal of Psy-chology, 68, 409–416.

Hayes, S. C., Barnes-Holmes, D., & Roche, B.(2001). Relational frame theory: A post-Skinnerian account of human language andcognition. New York: Kluwer Academic/Plenum.

Horne, P. J., & Lowe, C. F. (1996). On the ori-gins of naming and other symbolic behav-ior. Journal of the Experimental Analysis ofBehavior, 65, 185–241.

Jaynes, J. (1976). The origin of consciousnessin the breakdown of the bicameral mind.Boston, MA: Houghton Mifflin.

Kelleher, R. T., Riddle, W. C., & Cook, L.(1962). Observing responses in pigeons.Journal of the Experimental Analysis ofBehavior, 5, 3–13.

Lovaas, O. I. (1964). Cue properties of words:The control of operant responding by rateand content of verbal operants. Child De-velopment, 35, 245–256.

Palmer, D. C. (1996). Achieving parity: Therole of automatic reinforcement. Journal ofthe Experimental Analysis of Behavior, 65,289-290.

Palmer, D. C. (1998). The speaker as listener:The interpretation of structural regularitiesin verbal behavior. The Analysis of VerbalBehavior, 15, 3–16.

Risley, T. R. (1977). The development andmaintenance of language: An operantmodel. In B. C. Etzel, J. M. LeBlanc & D.M. Baer (Eds.), New developments in be-havioral research (pp. 81–101). Hillsdale,NJ: Erlbaum.

Risley, T. R., & Hart, B. (1968). Developingcorrespondence between the non-verbal and

verbal behavior of preschool children. Jour-nal of Applied Behavior Analysis, 1, 267–281.

Rosenfarb, I. S., Newland, M. C., Brannon, S.E., & Howey, D. S. (1992). Effects of self-generated rules on the development ofschedule-controlled behavior. Journal of theExperimental Analysis of Behavior, 58, 107–121.

Rosenfeld, H. M., & Baer, D. M. (1970). Un-biased and unnoticed verbal conditioning:The double agent robot procedure. Journalof the Experimental Analysis of Behavior,14, 99–107.

Seyfarth, R. M., Cheney, D. L., & Marler, P.(1980). Vervet monkey alarm calls: Seman-tic communication in a free-ranging primate.Animal Behaviour, 28, 1070–1094.

Shimoff, E., & Catania, A. C. (1998). The ver-bal governance of behavior. In K. A. Lattal& M. Perone (Eds.), Handbook of researchmethods in human operant behavior (pp.371–404). New York: Plenum.

Skinner, B. F. (1953). Science and human be-havior. New York: Macmillan.

Skinner, B. F. (1957). Verbal behavior. NewYork: Appleton-Century-Crofts.

Skinner, B. F. (1969). An operant analysis ofproblem solving. In B. F. Skinner, Contin-gencies of reinforcement (pp. 133–157).New York: Appleton-Century-Crofts.

Skinner, B. F. (1981). Selection by conse-quences. Science, 213, 501–504.

Skinner, B. F. (1986). The evolution of verbalbehavior. Journal of the ExperimentalAnalysis of Behavior, 45, 115–122.

words as behavior - university of maryland, baltimore countycatania/vb/acc avb words and...

Documents