Understanding Observational Learning. an Interbehavioral Approach

Understanding Observational Learning: AnInterbehavioral Approach

Mitch J. Fryling, The Chicago School of Professional Psychology

Cristin Johnston and Linda J. Hayes, University of Nevada, Reno

Observational learning is an important area in the field of psychology and behavior science moregenerally. Given this, it is essential that behavior analysts articulate a sound theory of how behaviorchange occurs through observation. This paper begins with an overview of seminal research in the areaof observational learning, followed by a consideration of common behavior analytic conceptualizationsof these findings. The interbehavioral perspective is then outlined, shedding light on some difficultieswith the existing behavior analytic approaches. The implications of embracing the interbehavioralperspective for understanding the most complex sorts of behavior, including those involved inobservational learning are considered.

Key words: observational learning, interbehaviorism, interbehavioral psychology, stimulus substitution,rule-governed behavior

Research in observational learning repre-sents a critical development in the history ofpsychology. Indeed, the research and schol-arly work conducted by Bandura and col-leagues set the occasion for the socialcognitive perspective of learning (Bandura,1986), which seemed to challenge thepossibility that all behavior could be ac-counted for by respondent and operantprocesses alone. Toward this, the socialcognitive perspective focused more explicitlyon both modeling and cognition, and theirrole in understanding behavior. Meanwhile,behavior analysts have continued to contendthat observational learning can be explainedthrough processes of generalized imitation,conditioned reinforcement, and rule-gov-erned behavior (e.g., Catania, 2007; Pear,2001; Pierce & Cheney, 2008). However,these contentions become increasingly diffi-cult when we take a closer look at thepsychological event of interest in observa-tional learning. Further, while behavioranalysts have continued to conduct researchin the area of observational learning, rela-tively little progress has been made towarddeveloping a theoretical understanding of

this work. The primary aim of the currentpaper is to consider the general findings ofthe observational learning research within athoroughly naturalistic, behavioral perspec-tive. Of course, verbal processes play animportant role in understanding observation-al learning, and thus, they are given bothgeneral and specific treatment throughout. Inpursuing this work, J. R. Kantor’s philosophyof interbehaviorism and scientific system ofinterbehavioral psychology are reviewed.The potential benefits of embracing theinterbehavioral perspective with respect tounderstanding observational learning andcomplex behavior more generally are con-sidered.

OBSERVATIONAL LEARNING

In the 1960s and 70s Albert Bandura andhis colleagues became well known for theirsocial psychology research in the area ofobservational learning. Indeed, several of theearly experiments in this area are very wellknown, and considered hallmarks in the fieldof psychology and behavior science (e.g.,Bandura & McDonald, 1963; Bandura, Ross,& Ross, 1963). These studies were pursuedfor a variety of reasons; partially to under-mine the value of common psychoanalytic(Bandura & Huston, 1961; Bandura, Ross, etal., 1963) and developmental theories (Ban-dura & McDonald, 1963), and also toevaluate the role of observation as a primary

Correspondence concerning this article shouldbe addressed to Mitch Fryling, The ChicagoSchool of Professional Psychology; 617 W. 7thSt., 8th Floor, Los Angeles, CA 90017. (e-mail:[email protected]).

Cristin Johnston is affiliated with SpectrumCenter, Oakland, CA.

The Analysis of Verbal Behavior 2011, 27, 191–203

191

determinant of behavior change. Early stud-ies examined the role of modeling1 on theacquisition of aggression (Bandura, Ross, &Ross, 1963) and moral judgment (Bandura &McDonald, 1963), for example, and provideda foundation upon which the social cognitivetheory was built. Importantly, this theory isoften considered to extend beyond behavioraltheories, questioning the possibility thatbehaviorism alone could provide a compre-hensive understanding of learning. Given theimportance of this research, we will nowprovide a brief overview of some of thegeneral findings of studies on observationallearning. It is important to note that ourreview is admittedly less than comprehen-sive, and that our primary aim is to describesome common themes within this literature.

The Role of Modeling

An early and longstanding aim of theobservational learning literature is to under-stand the role of modeling in behaviorchange (e.g., Bandura & Huston, 1961;Bandura & McDonald, 1963; Bandura, Ross,& Ross, 1961). For example, an early studyexamined how the incidental behaviors of anexperimenter might be acquired in thecontext of learning another task (Bandura &Huston). The important conclusion of thesestudies is that behavior change can and doesoccur through observation, even when suchobservation is incidental, occurring in thecontext of other activities. While this findingseems rather simple, it has significantimplications for how we conceptualize learn-ing. As we will discuss in the comingparagraphs, this general finding may presentspecific conceptual challenges for behavioraltheories of learning.

The role of consequences. Specific em-phasis was also placed on the role ofconsequences in the observational learningliterature (e.g., Bandura, 1965; Bandura,Grusec, & Menlove, 1966; Bandura &McDonald, 1963, Bandura, Ross, & Ross,1963). Experiments that added to our under-standing of the role of consequences gener-

ally compared behavior change betweenchildren who either observed a model whowas rewarded, a model who was punished, ora control condition (e.g., observing non-aggressive play or observing no consequenc-es). Generally, less behavior change isobserved when a child observes a modelbeing punished (e.g., Bandura, Ross, & Ross,1963).2

Interestingly, there is often no differencebetween conditions involving rewards andconditions involving no consequences at all.For example, Bandura and McDonald (1963)compared the effects of three differentvariables on the acquisition of moral judg-ment responses. In this study, the threevariables involved three different groups ofadult/child dyads: group one involved boththe model and child’s target judgments bereinforced, group two involved the model’sbehavior being reinforced but not the child’s,and group three involved no model and onlychild reinforcement. Importantly, in themodel/child groups trials alternated betweenthe model and the child. Groups one and twodemonstrated more behavior change thangroup three at a 1–3 week post-treatmentassessment. Thus, the researchers concludedthat modeling was the significant factorinvolved in the acquisition of the moraljudgment repertoire.3 Other experiments alsofound no difference between the reward andno consequence groups, while the modelpunished group continued to yield differentresults (e.g., Bandura, 1965).

Along similar lines, other studies seemedto raise questions about the potentiallydetrimental effects of incentives on theacquisition of behavior. For example, at thebeginning of one experiment (Bandura,Grusec, & Menlove, 1966) half of theparticipants were placed into an incentivecondition where they were told that they

1 The term modeling is used synonymously withobservation and demonstration in this context. Inother words, when something has been modeledthe individual has observed a demonstration of theresponse and factors surrounding it.

2 See Greer et al., 2004 for a description ofrelated studies on peer tutoring, where it was theobservation of corrections, and not simply ofreinforcement, that resulted in observationallearning.

3 Of note, the researchers acknowledged thepossibility that their positive statements may nothave been the most optimal reinforcers, and thus,it is possible that the modeling plus reinforcementcondition would have been superior had morepowerful reinforcers been used (Bandura &McDonald, 1963, p. 281).

192 MITCH J. FRYLING et al.

would be given candy treats for correctlydemonstrating what they learned after watch-ing a movie. More specifically, after watch-ing a film, children in both conditions wereasked to demonstrate what they observed onthe movie. Generally, the researchers foundthat children in the incentive condition didslightly worse than those in the no incentivecondition, raising questions about the bene-fits of incentives on learning (see Bandura, etal., p. 505).4

At this point we must note that the termsreward, reinforcement, and operant condi-tioning are used rather loosely within thisliterature. From a behavior analytic perspec-tive, a stimulus change can only be classifiedas a reinforcer if it increases the futurefrequency of the class of behavior it wasmade contingent upon (e.g., Cooper, Heron,& Heward, 2007). Given this, the majority ofstimulus changes called ‘‘rewards’’ or ‘‘re-inforcers’’ in the observational learningliterature do not technically meet the criteriato be classified as reinforcers, or as beinginvolved in the process of reinforcement oroperant conditioning in general. Neverthe-less, we can say that consequences seem toplay some role in observational learning.Again, there are studies suggesting that thereare no differences between observation withreinforcement and observation with no con-sequence at all, leaving us more confidentthat if consequences have a role, aversiveconsequences seem to play a large part.Given these important concerns, however,these findings need to be interpreted withcaution.

The Role of Verbal Behavior

As this line of researched progressed,increasing attention was paid to the role ofcognitive factors, often described with theterms coding and rehearsal. Generally,coding can be thought of as describing whatis observed in some way, whereas rehearsalcan be thought of as practicing what wasobserved. For example, Bandura, Grusec, &Menlove. (1966) examined the effects ofdescribing the activity of the model (‘‘cod-

ing’’) on the acquisition of observed behav-ior. Of specific interest, this study was fueledby motivation to discredit behavior analystswho failed to account for ‘‘delayed repro-duction of modeling behavior’’ (p. 499),which was assumed to necessarily involvesome sort of cognitive activity. In this studythree groups of children all viewed a video;one group was asked to ‘‘verbalize everyaction of the model as it is being performed’’(p. 501), the second group to ‘‘count 1 and a2, and a 3, and a 4, and a 5’’ (p. 501)repeatedly while watching the video, and athird group observed without any instruction.The researchers found that those individualswho verbally described every action of themodel were the most successful when testedfor behavior change at a later time. Impor-tantly, this study highlights the early recog-nition of ‘‘cognitive’’ factors in observation-al learning.

In an effort to elaborate upon this sort ofresearch, Bandura and Jeffrey (1973) exam-ined the role of ‘‘coding and rehearsal’’ onthe acquisition of observed behavior. Theresearchers found that participants who‘‘symbolically coded’’ (i.e., developed num-ber or letter coding systems) the model’sactions, and also immediately rehearsed (i.e.,practiced) those codes had the best outcomes.Neither coding without symbolic rehearsal orsymbolic rehearsal without coding was foundto be sufficient. Put differently, developing acoded description of the models actions andpracticing that description were both found tobe important factors in the acquisition ofobserved behavior. Interestingly, physicallypracticing (‘‘motor rehearsal’’) the observedbehavior was found to be less important. Thisseemed to support a growing distinctionbetween different aspects of an individual’srepertoire and the various processes thatcontribute to their existence (see below).

Learning and performance. Related to therole of verbal behavior, Bandura and col-leagues began to notice a difference betweenthe observers imitative performance at a latertime compared to their ability to describewhat was observed when asked. The abilityto describe what was observed was viewed asa measure of learning, while engaging in theobserved behavior at a later time was viewedas performance. For example, Bandura, Ross,& Ross (1963) found that children in both the

4 The idea that rewards distract individuals fromlearning seems to be related to the concerns raisedby Alfie Kohn (1999).

OBSERVATIONAL LEARNING 193

aggressive-reward (participants observed amodel be rewarded for engaging in asequence of responses) and aggressive-pun-ished (participants observed a model bepunished for engaging in a sequence ofresponses) groups were able to describe theobserved sequences of behavior, despitedifferences in imitative behavior change.Similarly, Bandura (1965) found that differ-ences between group measures on imitationof observed behavior were removed on an‘‘acquisition index,’’ where children weretold they would get a reward for telling theexperimenter what the model did. Thesefindings further highlighted the role of verbalbehavior in the process of learning fromobservation, including the various ways inwhich such learning from observation mightbe measured. That is, one way of measuringlearning from observation is through imita-tion of the observed response at a later time,while another is through descriptions of theobserved behavior. As these repertoiresseemed to be influenced by different factors,Bandura and colleagues began to distinguishbetween them more and more.

Theoretical Developments

Throughout the above studies Bandura andcolleagues began to articulate a theoreticalmodel of observational learning. Fueled byfindings that individuals might be able todescribe observed behavior at a later time,even if they did not actually engage in thebehavior themselves during a testing condi-tion (e.g., Bandura, 1965; Bandura, Ross, &Ross, 1963), Bandura and colleagues beganto distinguish between learning and perfor-mance (also see Greer, Singer-Dudek, &Gautreaux, 2006). Specifically, Bandura andcolleagues noted that verbal processes weremore likely to influence learning,5 whereasconsequences were more likely to influencethe extent to which the individual’s behaviorchanged through observation (i.e., that theyactually engaged in the observed behavior).Indeed, theoretical accounts of observationallearning highlight this distinction (e.g.,

Bandura & Jeffrey, 1973; Greer, Singer-Dudek, & Gautreaux, 2006).

Bandura and colleagues assumed thatlearning from observation occurred via aninput-output, cognitive model. Specifically,Bandura and Jeffrey (1973) described fourprocesses that account for learning fromobservation: attentional, retention, motorreproduction, and motivational. Banduraand Jeffery (1973) say, ‘‘Within this frame-work acquisition of modeled patterns isprimarily controlled by attention and reten-tion processes. Whereas performance ofobservationally learned responses is regulat-ed by motor reproduction and incentiveprocesses’’ (p. 122).

Attentional processes were described ascognitive abilities that ‘‘regulate sensoryregistration of modeled actions’’ and reten-tion processes were those that took ‘‘transi-tory influences and converted to enduringinternal guides for memory representation’’(Bandura & Jeffery, 1973, p. 122). Motorreproduction processes are those that movecomponent actions stored in memory intoovert action resembling that of the modeledbehaviors. Finally, motivational processesdetermine whether or not those behaviorsemerge as overt action.

According to the authors, this model notonly explains how a modeled response can beimitated immediately after it is observed, butcan also explain how this behavior can bereproduced later under many different cir-cumstances. Bandura and Jeffrey (1973)conclude, ‘‘After modeled activities havebeen transformed into images and readilyutilizable verbal symbols, these memorycodes can function as guides for subsequentreproduction’’ (p. 123). The authors alsoconcluded that participants who engage intransforming modeled actions into eitherdescriptive words or visual images achievehigher levels of observational learning thanthose who did not.

As a result of these and other experiments,Bandura theorized that observational learn-ing was an integral part of human develop-ment, which accounted for the developmentof the personality (Bandura & Walters,1963), as well as social and antisocialbehaviors in children (Bandura, 1973). Im-portantly, this research shows that humanscan learn without directly experiencing the

5 In this literature the term learning is used todescribe the individual’s ability to describeobserved behavior at a later time.


consequences of their own actions. Thus, ifbehavior analysts aim to develop a compre-hensive account of learning it must includean adequate description of these instances. Inparticular, behavior analysts must account forthe acquisition of novel behavior in theabsence of contingent reinforcement for theindividual engaging in those responses, andalso articulate the role of verbal behavior inobservational learning.

In summary, the studies conducted byBandura and colleagues seemed to questionthe role of rewards on the behavior of theobserver. Importantly, Bandura believed thatreinforcement history alone was not suffi-cient, and that the observation of a modelwas the most critical factor. Moreover,learning from observation was viewed to bea result of other processes, of which ‘‘verbalcoding’’ was one. These general findingsseemed to devalue the comprehensiveness ofthe behavioral position, and set the stage forthe social cognitive perspective. However, itis crucial that we reiterate the fact thatBandura and colleagues often misused theterms reinforcer and reinforcement, and thus,it is difficult to draw valid conclusions aboutthe role of consequences from this line ofresearch. What can be said is that observa-tional learning is an important area forbehavior science to consider.

Bandura found limitations with the operantinterpretation of behavior, albeit a less thanthoroughly informed understanding of it.Observational learning seems to defy tradi-tional discriminative stimulus—response—reinforcer analyses, even when more con-temporary concepts (e.g., the motivatingoperation) are considered. Specifically, novelresponses occur in observational learningmodels, responses that have obviously neverbeen reinforced. Added to this, delayedresponding is common, and such respondingpresents conceptual challenges to traditionalbehavioral concepts (e.g., Bandura, Grusec,& Menlove, 1966, p. 499). As mentionedearlier, it is perhaps not surprising thatBandura’s work may be considered by someto be an extension or move beyond thebehavioral position. The limitations of Ban-dura’s work not withstanding, Bandura andcolleagues raised several important issuesregarding the role of observation and verbalbehavior in behavior change processes.

Still, Bandura’s model relies upon theexistence of hypothetical entities that do notexist in the spatiotemporal event matrixcomprising the natural world. In other words,Bandura’s theoretical constructs are notderived from events, and as such cannot befound and thereby can never actually bestudied (see Kantor, 1957; Smith, 2007).Rather, they are inferences derived from athoroughly mentalistic, dualistic worldview.Behavior analysts have long held thatembracing such constructs can only distractworkers from a scientific analysis (e.g.,Skinner, 1953). It isn’t surprising, then, thatbehavior analysts have proposed an alterna-tive conceptualization of observational learn-ing. In the following section we provide anoverview of the behavior analytic position onobservational learning.

THE BEHAVIOR ANALYTIC POSITION

The behavior analytic account of observa-tional learning rests squarely upon theprocess of generalized imitation (Baer, Pe-terson, & Sherman, 1967; Baer & Sherman,1964; Pierce & Cheney, 2008). This is afamiliar process, where the organism is askedto imitate several responses of the model(e.g., ‘‘do this’’ while the model is touchingtheir nose), and after multiple exemplarshave been successfully trained, the organismis asked to engage in a response which hasnever been modeled before. Generalizedimitation is said to occur when the organismengages in a response that has never beenmodeled or reinforced in the past; that is,when imitation has ‘‘generalized’’ to newbehaviors. Furthermore, it is assumed that thesocial community shapes up delays inimitative responses, and thus, it is said that‘‘all instances of modeling and imitationinvolve the absence of the Sd’’ (Pierce &Cheney, 2008, p. 252). For example, a childmight watch their favorite TV show, and at amuch later time repeat a phrase from theshow, perhaps while sitting in the car, andtheir parent might say ‘‘yes, that’s what youheard on TV!’’. In other words, the organismis said to learn to imitate observed behaviorin the absence of any particular stimulus, andperhaps at a much later point in time. In thissense, the organism may be said to ‘‘emit’’


behaviors, which typically fall under thepurview of generalized imitation.

Importantly, conditioned reinforcementhypotheses are also central to the behavioranalytic conceptualization of observationallearning and imitation in general. In thissense, behaviors that closely resemble theobserved behavior of models are presumed tohave a history of reinforcement, and thus,behaving in a manner which is similar to themodel may become conditioned reinforceritself. This sort of conceptualization seems tobe particularly helpful toward the behavioranalytic understanding of delayed imitation(see Gladstone & Cooley, 1975; Rosales-Ruiz & Baer, 1997).

Behavior analysts have also provided anaccount of the verbal coding that is said toparticipate in observational learning. Forexample, behavior analysts propose thatindividuals derive self-rules when they ob-serve their environment (e.g., Hayes, Barnes-Holmes, & Roche, 2001; Hayes, Zettle, &Rosenfarb, 1989; Poppen, 1989). It is as-sumed that society teaches the organism totact (Skinner, 1957) relationships in theirenvironment, and that these descriptionsexert tremendous control over behavior.Indeed, it is suggested that a large amountof rule-following behavior is reinforcedthroughout the organisms lifetime, and whencombined with a history of tact repertoiresbeing reinforced, individuals both deriveself-rules (i.e., tact if-then relations in theirenvironment) and subsequently engage in agreat deal of rule-following with respect tothose rules.

For example, a child might observe ateacher praising another child for accuratelymatching a Spanish flashcard to the corre-sponding English flashcard (‘‘Good jobmatching perro with dog!’’). Two days later,the child who observed the incident may beasked to ‘‘match same’’ when given thatsame Spanish flashcard, and correctly place iton the corresponding English flashcard.From the behavior analytic perspective itmay be assumed that the child already has ageneralized imitative repertoire, so they areimitating the child they observed at a laterpoint in time (see conditioned reinforcementhypotheses above). Furthermore, the childmay or may not have tacted the observedrelationship when it occurred (rule-stating),

and engaged in rule-following behavior whenshe interacted with the card at a later time.Both of these possibilities are consistent withthe behavior analytic position. Importantly,the behavior analytic position does notrequire the individual to engage in rule-stating and following for observational learn-ing to occur. Related to the latter, a recentseries of studies conducted by Greer andcolleagues seems to support the notion thatobservational learning may occur withoutrule-following. For example, individualshave acquired the ability to learn new wordsthrough experiences that do not involveobserving consequences of another, andstimuli have been conditioned as reinforcersthrough the observation of others interactingwith them, both of which do not requireanalyses of rule-governed behavior (seeGreer & Ross, 2008, Greer & Speckman,2009).

It must be noted that many of these issuesare at the center of current controversy,debate, and development in the field ofbehavior analysis. For example, the perspec-tives of joint control (e.g., Lowenkron, 1998)naming (Horne & Lowe, 1996), relationalframe theory (Hayes, Barnes-Holmes, &Roche, 2001), and verbal behavior develop-ment (e.g., Greer & Ross, 2008; Greer &Speckman, 2009) all seem to account for thetype of phenomena we have commented onherein. Given the importance of these issues,this is a good sign. We primarily mention thisto acknowledge the current fact that there isnot a behavior analytic position on many ofthese issues. Nevertheless, missteps mayoccur while we are on our journey to accountfor such phenomena, missteps that couldhave more or less dangerous implications forbehavior analysis as an enterprise. It is ourperspective that the interbehavioral positionmay be a rather useful foundation forworkers as we continue on this journey (seeMorris, Higgins, & Bickel, 1982).

Generally speaking, the behavior analyticconceptualization of observational learningrelies on generalized imitation, conditionedreinforcement, and a range of verbal pro-cesses, depending on ones theoretical prefer-ence. These processes seem to account forthe fact that imitative responses which havenever been reinforced occur at a later time,and also for the role of verbal behavior in


observational learning. The fact that there area number of different perspectives on manyof these issues may be considered a sign ofprogress and growth within behavior analy-sis, but at the same time highlights the needfor further system building in this area. In thefollowing sections we take a closer look atthe behavior analytic position through thelens of interbehavioral psychology. Beforedoing so, we briefly introduce the reader tothe interbehavioral position, as it is relativelyless familiar to most behavior analysts.

THE INTERBEHAVIORIAL POSITION

From the perspective of interbehavioralpsychology the event of interest is always athoroughly naturalistic, psychological event.Specifically, this event is always the stimulusfunction (sf ) r Rresponse function (rf)interaction (Kantor, 1958). Moreover, thisinteraction always participates in a multi-factored, interrelated field. This field isconceptualized by the following formula:PE 5 C (k, sf, rf, hi, st, md); where PE is thepsychological event, C is the interrelation-ship of all of the participating factors, k is theunique organization of all factors, sf is thestimulus function, rf is the response function,hi is the interbehavioral history, st is settingfactors, and md is the medium of contact.Importantly, this is one event, one interbe-havioral field. When one factor is changedthe entire field is altered. This is to say noneof the above factors are viewed as indepen-dent, dependent, or having causal status.Rather, all of the factors are equal partici-pants in the one, integrated whole (see Smith,2006).

Of particular relevance to our discussionof observational learning and complex be-havior in general is the explicit distinctionbetween stimulus objects and stimulus func-tions made within Kantor’s system (e.g.,Kantor, 1924, pp. 47–48; Parrott, 1983a,1983b, 1986). In other words, the stimulatingaction of stimulus objects is differentiatedfrom the formal properties of those objects inKantor’s system. Kantor has suggested thatthe borrowing of the terms stimulus andresponse from biology, where stimulus andresponse functions are at least relativelymore determined by their structural proper-ties, has perhaps contributed to the failure to

distinguish between object and functionalproperties in the domain of psychology(Kantor, 1958, p. 68). For example, inKantor’s system a picture as a stimulusobject would be explicitly distinguished fromits psychological functions, such that ac-counting for seeing something in the absenceof the thing seen (as when looking at apicture ‘‘reminds you’’ of the time or place itwas taken) is not difficult (see Parrott, 1983a,1983b, 1986; Skinner, 1974). The process bywhich this happens is central to understand-ing complex behavior, including those thattypically fall within the purview of observa-tional learning, and we will now describe thisprocess in more detail.

Kantor suggested that association condi-tions are fundamental psychological process-es (1921, 1924). The term association is usedhere to refer to spatiotemporal relationships;that is, to relationships among various factorsthat occur in the environment together inspace and time. To be clear, these factors areassociated in the environment, and not withinthe organism. Further, it is not the organismwho is associating; rather, the environment iswhere all associating takes place. Associationconditions may involve stimuli and respons-es, stimuli and stimuli, settings and stimuli,settings and reactions, settings and settings,and reactions and reactions (including im-plicit and nonimplicit variations thereof;Kantor, 1924, pp. 321–322).

Stimulus Substitution

Stimulus substitution is the outcome of ahistory of an organism interacting withvarious association conditions (Kantor,1924, 1958; Parrott, 1983a, 1983b, 1986).That is, given an organisms history ofinteracting with spatiotemporal relationships(A-coffee shopr R B-Peter), stimulus ob-jects may have the stimulational properties ofother objects, even when those other objectsare no longer physically present. This is howyou might see Peter when you enter a coffeeshop you frequented with him, even when heisn’t physically there. In this example,stimulus A (coffee shop) and B (Peter)occurred together in space and time, and anorganism interacted with that relationship,such that B becomes A (B[A]) and A becomesB(A[B]), psychologically speaking (see


Hayes, 1992a). This process is of particularimportance to understanding complex behav-ior of various sorts. Furthermore, this is howinterbehaviorists are able to conceptualizethe past and present as one, avoiding bothmentalistic and reductionistic practiceswhich place the past within the organism inone way or another (see Hayes, 1992b).

Added to this, through processes of gener-alization, stimuli that share physical features ofthose that participated in spatiotemporal asso-ciation conditions may also develop substitutestimulus functions. For example, a coffee shopthat is physically similar to the coffee shop youwent to with your friend Peter might alsosubstitute for Peter. Specifically, you might seePeter in the presence of a coffee shop that isphysically similar to the shop you frequentedwith him. That is to say, substitute stimulusfunctions also generalize to stimuli which havenever actually participated in spatiotemporalassociation conditions, but which are physical-ly similar to stimuli which have, and therebyinvolve similar stimulus functions. This type ofprocess may become particularly subtle, and islikely to be involved in a range of complexbehaviors, including imagining and dreaming.

At this point it is important to address onepotential misunderstanding with the interbe-havioral perspective, specifically with respectto association conditions and the developmentof substitute stimulus functions.6 We aresuggesting that all stimuli which occur togetherin space and time, and which the organisminteracts with, may develop substitute stimulusfunctions of one another. That is, it is possiblefor all stimuli to develop substitute stimulusfunctions of any other stimulus, given theappropriate interbehavioral history. Indeed, asan individual’s interbehavioral history be-comes more and more elaborate, one mightimagine how all stimuli could develop substi-tute stimulus functions of all other stimuli,such that everything might become one,psychologically speaking. However, recall thatthe stimulus functionr Rresponse functioninteraction is always a participant in anexceptionally unique, complex, multifactoredfield. Indeed, Kantor stated ‘‘Each interaction

is always absolutely specific. What the react-ing organism and the stimulus object do ineach interaction constitutes a distinctly uniquerelational happening’’ (1977, p. 38). Thus,while a specific stimulus object may indeedsubstitute for a wide range of things given anappropriate interbehavioral history, specificsubstitute stimulus functions are always actu-alized (or not) in a unique interbehavioral field.For example, a glass of sangria mightsubstitute for a particular friend in a specificmultifactored field (you might see your friendand remember drinking sangria together),whereas that same glass of sangria mightsubstitute for the music of a live band in adifferent multifactored field (you might hearthe music that was playing at a restaurantwhere you drank sangria in the past). As thisexample demonstrates, while there may be awide range of potential substitute stimulusfunctions for every stimulus object, in each andevery specific psychological event, particularsubstitute stimulus functions are actualized.

Thus far we have briefly introduced someimportant features of interbehavioral psy-chology, which we find to be particularlyrelevant to our understanding of observation-al learning. From the interbehavioral per-spective, individuals observe (i.e., interactwith) spatiotemporal association conditionsin the environment (e.g., a child putting scrappaper in the recycling bin and this beingfollowed by praise), such that at a later timethe stimulus objects involved might substi-tute for the prior observation (e.g., the scrappaper might have the stimulus functions ofpraise in the previous observation). In otherwords, the scrap paper develops the stimula-tional properties of the observed relations; itsubstitutes for them. Psychologically speak-ing, the scrap paper is those relations (seeHayes, 1992a, 1992b).

The role of verbal behavior must also beconsidered in the context of our analysis thusfar. Generally speaking, one outcome ofinteracting with an observed relationship isbeing able to describe it. In other words,describing an observed relationship requiresthe organism to interact with it, and thus,descriptions are a particularly strong indica-tion that the relations assumed to be observedhave indeed actually been contacted. How-ever, from our perspective verbal behavior,including rules more generally, does not

6 For example, some have criticized interbeha-viorism for its ‘‘loose form of associationism’’(e.g., Hayes, Barnes-Holmes, & Roche, 2001,p. 8).


explain observational learning. This is to say,whether or not the organism describes theobserved relationship does not explain be-havior change at a later time; however, notsurprisingly, it is likely to be correlated withit, as it assures the organism has interactedwith the observed relation. Moreover, to theextent that rule-statements substitute for ahistory of reinforcement, they may furtherenhance any learning by observation. Impor-tantly, in this sense verbal behavior does not‘‘mediate’’ responding. Its participation inthe process of observational learning, how-ever, seems to be worth considering. In doingso, it is important that verbal behavior not begiven any causal or special sort of status.Observational learning certainly can, anddoes occur in the absence of verbal behavior,as is the case in animal research within thisarea (e.g., Biederman, Robertson, & Vana-yan, 1986; Meyers, 1970; Reiss, 1972).

Our contention that verbal behavior not begiven any causal status within the conceptu-alization of observational learning may seemto be at odds with a number of popularperspectives in behavior analysis. For exam-ple, a growing body of research on naming(e.g., Miguel, Petursdottir, Carr, & Michael,2008), joint control (e.g., Lowenkron, 1998),and generalized imitation (e.g., Horne &Erjavec, 2007) seems to support the idea thatverbal behavior is mediational. Again, asstated above, we do not deny that verbalbehavior is likely to be helpful in a number ofcircumstances, but caution against giving itany sort of special status. That is, verbalbehavior may, but importantly also may not,participate in learning from observation. Inthis sense, verbal behavior need not beconsidered ‘‘meditational.’’ Our perspectiveon this matter seems to be both parsimoniousand comprehensive. That is, it does notemploy any unnecessary assumptions orconstructs, and accounts for observationallearning that occurs with and without verbalbehavior.7

We hope we have made it clear thatobservational learning isn’t puzzling from aninterbehavioral perspective. Stimulus substi-tution offers a straight forward, naturalistic,and parsimonious way to conceptualizecomplex processes, including those involvedin observational learning. Importantly, theinterbehavioral perspective also avoids someshortcomings found with the behavior ana-lytic interpretation of observational learning.In the following section we outline andaddress these issues specifically.

Review of the Behavior Analytic Perspective

As described earlier, the behavior analyticconceptualization of observational learningrests on the processes of generalized imitation,conditioned reinforcement, rule-governed be-havior, and verbal processes more generally.From our perspective these analyses fail tofully articulate the nature of stimulation in thepsychological event. Again, from the interbe-havioral perspective the psychological eventis always the stimulus functionr Rresponsefunction interaction. The generalized imita-tion analysis leaves us questioning the natureof the stimulus interacted with. In other words,it is not clear what the stimulus is. Thisproblem is further underscored by the sugges-tion that generalized imitation involves re-sponding in the absence of a discriminativestimulus (Pierce & Cheney, 2008, p. 252).Given our assumption that psychologicalevents always involve sfr Rrf interactions,as participants in multifactored fields, thisaccount is problematic. The process ofderiving and following self-rules leaves us ina similar situation. Again, we are leftquestioning the nature of the stimulus inter-acted with. That is, it unclear what theorganism is interacting with when he/shederives a self-rule, and similarly, when he/she follows such a rule. Again, given ourassumptions about the psychological event,both of these analyses require further consid-eration of the stimulus involved.

Added to the concerns described above,behavior analytic conceptualizations also failto explicitly articulate the location of thestimulus. In other words, it is unclear wherethe stimulus interacted with is located.Failing to fully describe the nature andlocation of the stimulus leaves the door open

7 A number of socially significant behaviorsinvolve language, and we are not questioning theinterest in it for the purposes of understandinghow to promote such behaviors (e.g., categoriza-tion). However, we are arguing that language notbe given special status in the conceptualization ofobservational learning.


for common mentalistic explanations tothrive. In the case of generalized imitationwe find ourselves saying that the response is‘‘in the repertoire’’ of the organism, becausethe stimulus is private, covert, or biologicalin nature (also see Hayes & Fryling, 2009).Alternatively, the organism may be said to‘‘derive’’ or ‘‘relate’’ with respect to partic-ipating verbal processes. In other words, weeither avoid attempting to specify the stim-ulus, place it within the organism, or,alternatively, suggest that it is available onlyto those involved in other scientific disci-plines, namely biology.8 In each of thesecases, we fail to provide a thoroughlypsychological account of the event we areinterested in, leaving our job unfinished. Ashas been the case throughout history, whereour work is left unfinished, both dualistic andreductionistic workers are quick to completethe job. While it may be argued that much ofthe contemporary work in the area ofcomplex behavior does in fact avoid manyof the concerns we have described, a failureto be explicit about these important issuescan only result in long-term confusion, and apossible resurfacing of mentalistic thinking.

CONCLUSION

The behavior analytic community contin-ues to be interested in the important process-es involved in observational learning (e.g.,Alvero & Austin, 2004; Bruzek & Thomp-son, 2007; Greer & Singer-Dudek, 2008;Greer, Singer-Dudek, Longano, & Zrino,2008; Moore & Fisher, 2007; Ramirez &Rehfeldt, 2009; Rehfeldt, Latimore, & Stro-mer, 2003). Added to this, there are someinteresting reasons to believe that thisprocess has important clinical value whencompared to other procedures (see Hayes,

Kohlenberg, & Melancohn, 1989). What isneeded is a thoroughly naturalistic concep-tualization of observational learning, one thatavoids all mentalism (i.e., no intermediatesteps within the organism). As we havedescribed, the interbehavioral perspectiveoffers us just that, a clear, consistent, andthoroughly naturalistic conceptualization ofobservational learning. Moreover, it is onethat does not require any additional con-structs to explain complex processes, remain-ing comprehensive all the while.

It is our perspective that the positiondescribed in this paper may be integratedwith contemporary research and scholarshipin behavior analysis. This is especially sowhen we make clear distinctions betweeninvestigative constructs and events, as isadvocated by interbehaviorists (see Fryling& Hayes, 2009; Kantor, 1957; Smith, 2007).Kantor (1958) has suggested that investiga-tive constructs are acceptable within thecontext of the investigative subsystem ofscience, but that these constructs should notbe confused with the constructions of thesubject matter and philosophy more general-ly. That is, the constructs we employ tounderstand various interrelations among fac-tors participating in psychological eventsshould never be confused to be representa-tions of the subject matter as a whole, asbeing explanatory of one another, or ashaving more or less causal status. Forexample, both operant and respondent pro-cesses can be conceptualized within the moreglobal processes of association and subse-quent outcomes of stimulus substitution.Contemporary research in behavior analysisrequires us to emphasize specific aspects tothe interbehavioral position, particularly withrespect to the role of the context (uniquemultifactored fields), and the actualization ofspecific substitute stimulus functions. In thisregard, the research on relational respondingis particularly stimulating. In this line ofresearch a multitude of historical associationconditions are manipulated in unique ways,under various contextual conditions, and thedevelopment or ‘‘emergence’’ of a widerange of events is then tested. When theseinteresting outcomes are conceptualized asunique sorts of substitute stimulation, oper-ating in historical, multifactored fields, theirexplanations remain wholly consistent and

8 Here, it is important to note that even whenbiological factors are observed (and indeed, theyincreasingly are) they are never observed to beengaging in the psychological event of interest.That is to say, we can never observe the brain orany biological component of the organism engag-ing in the behavior we are most interested in (seeKantor, 1947). Confusions between what ismeasured and what ones says they measuring arecommon in science (see Kantor, 1957; Smith,2007), and are especially likely when there is afailure to fully articulate the boundary conditionsbetween individual scientific disciplines.


naturalistic. We think most contemporaryresearch and scholarship in behavior analysiscan and should be integrated with theinterbehavioral perspective. Importantly,such integration might serve to coordinatethe efforts of various workers in the field, andultimately maximize on our productivity as ascientific enterprise.

The limitations of Bandura’s work notwithstanding, the process of learning fromobservation is interesting and relevant to acomprehensive analysis of behavior. Indeed, ifone values such comprehensiveness, our mostbasic concepts and principles must be relevantto, and provide an account of observationallearning. Moreover, this comprehensiveness isonly valuable when it is achieved within thecontext of validity (internal consistency) andsignificance (external consistency within thegreater field of the sciences; see Clayton,Hayes, & Swain, 2005; Kantor, 1958). Theinterbehavioral perspective is particularlyvaluable in this regard. Kantor’s conceptuali-zation of the psychological event, with all ofits fullness, provides an avenue by which themost complex sorts of behavior, includingthose involved in observational learning,might be fully integrated into a natural scienceapproach to the analysis of behavior.

REFERENCES

Alvero, A. M., & Austin, J. (2004). Theeffects of conducting behavioral observa-tions on the behavior of the observer.Journal of Applied Behavior Analysis, 37,457–468.

Baer, D. M., Peterson, R. F., & Sherman, J.A. (1967). The development of imitationby reinforcing similarity to a model.Journal of the Experimental Analysis ofBehavior, 10, 405–416.

Baer, D. M., & Sherman, J. A. (1964).Reinforcement control of generalizedimitation in young children. Journal ofExperimental Child Psychology, 1, 37–49.

Bandura, A. (1965). Influence of models’reinforcement contingencies on the acqui-sition of imitative responses. Journal ofPersonality and Social Psychology, 1,589–595.

Bandura, A. (1973). Aggression: A sociallearning analysis. Oxford, England: Pren-tice-Hall.

Bandura, A. (1986). Social foundations ofthought and action: A social cognitivetheory. Englewood Cliffs, NJ: Prentice-Hall.

Bandura, A., Grusec, J. E., & Menlove, F. L.(1966). Observational learning as a func-tion of symbolization and incentive set.Child Development, 37, 499–506.

Bandura, A., & Huston, A. C. (1961).Identification as a process of incidentallearning. Journal of Abnormal and SocialPsychology, 63, 311–318.

Bandura, A., & Jeffrey, R. W. (1973). Role ofsymbolic coding and rehearsal processes inobservational learning. Journal of Person-ality and Social Psychology, 26, 122–130.

Bandura, A., & McDonald, F. J. (1963).Influence of social reinforcement and thebehavior of models in shaping children’sjudgment. The Journal of Abnormal andSocial Psychology, 67, 274–281.

Bandura, A., Ross, D., & Ross, S. A. (1961).Transmission of aggression through imi-tation of aggressive models. Journal ofAbnormal and Social Psychology, 63,575–582.

Bandura, A., Ross, D., & Ross, S. A. (1963).Vicarious reinforcement and imitativelearning. Journal of Abnormal and SocialPsychology, 67, 601–607.

Bandura, A., & Walters, R. H. (1963). Sociallearning and personality development.New York: Holt Rinehart and Winston.

Biederman, G. B., Robertson, H. A., &Vanayan, M. (1986). Observational learn-ing of Two visual discriminations bypigeons: A within-subjects design. Jour-nal of the Experimental Analysis ofBehavior, 46, 45–49.

Bruzek, J. L., & Thompson, R. H. (2007).Antecedent effects of observing peer play.Journal of Applied Behavior Analysis, 40,327–331.

Catania, A. C. (2007). Learning, interim 4thedition. Cornwall-on-Hudson, NY: SloanPublishing.

Clayton, M. C., Hayes, L. J., & Swain, M. A.(2005). The nature and value of scientificsystem building: The case of interbeha-viorism. The Psychological Record, 55,335–359.

Cooper, J. O, Heron, T. E., & Heward, W. L.(2007). Applied behavior analysis (2nded.). Upper Saddle River, NJ: Pearson.


Fryling, M. J., & Hayes, L. J. (2009).Psychological constructs and events: Analliance with Smith. The PsychologicalRecord, 59, 133–142.

Gladstone, B. W., & Cooley, J. (1975).Behavioral similarity as a reinforcer forpreschool children. Journal of the Exper-imental Analysis of Behavior, 23, 357–368.

Greer, R. D., Keohane, D., Meincke, K.,Gautreaux, G., Pereira, J. A., Chavez-Brown, M., & Yuan, L. (2004). Keyinstructional components of effective peertutoring for tutors, tutees, and peerobservers. In D. J. Moran & R. W. Malott(Eds.), Evidence-based educational meth-ods (pp. 295–334). New York: Elsevier/Academic Press.

Greer, R. D., & Ross, D. E. (2008). Verbalbehavior analysis: Inducing and expand-ing complex communication in childrenwith severe language delays. Boston:Allyn & Bacon.

Greer, R. D., & Singer-Dudek, J. (2008). Theemergence of conditioned reinforcementfrom observation. Journal of the Experi-mental Analysis of Behavior, 89, 15–29.

Greer, R. D., Singer-Dudek, J., & Gautreaux,G. (2006). Observational learning. Inter-national Journal of Psychology, 41(6),486–499.

Greer, R. D., Singer-Dudek, J., Longano, J.,& Zrino, M. (2008). The emergence ofpraise as conditioned reinforcement as afunction of observation in preschool andschool age children. Revista Mexicana dePsicologia, 25(1), 5–26.

Greer, R. D., & Speckman, J. (2009). Theintegration of speaker and listener re-sponses: A theory of verbal development.The Psychological Record, 59, 449–488.

Hayes, L. J. (1992a). Equivalence as process.In S. C. Hayes & L. J. Hayes (Eds.),Understanding verbal relations (pp. 97–108). Reno, NV: Context Press.

Hayes, L. J. (1992b). The psychologicalpresent. The Behavior Analyst, 15, 139–145.

Hayes, L. J., & Fryling, M. J. (2009).Overcoming the pseudo-problem of pri-vacy in the analysis of behavior. Behavior& Philosophy, 37, 39–57.

Hayes, S. C., Barnes-Holmes, D., & Roche,B. (Eds.) (2001). Relational Frame

Theory: A post-Skinnerian account ofhuman language and cognition. NewYork: Plenum.

Hayes, S. C., Kohlenberg, B. S., & Melan-con, S. M. (1989). Avoiding and alteringrule-control as a strategy of clinicalintervention. In S. C. Hayes (Ed.), Rule-governed behavior: Cognition, contingen-cies, and instructional control (pp. 359–385). New York: Plenum.

Hayes, S. C., Zettle, R. D., & Rosenfarb, I.(1989). Rule following. In S. C. Hayes(Ed.), Rule-governed behavior: Cognition,contingencies, and instructional control(pp. 191–220). New York: Plenum.

Horne, P. J., & Erjavec, M. (2007). Doinfants show generalized imitation ofgestures? Journal of the ExperimentalAnalysis of Behavior, 87, 63–87.

Horne, P. J., & Lowe, C. F. (1996). On theorigins of naming and other symbolicbehavior. Journal of the ExperimentalAnalysis of Behavior, 65, 185–241.

Kantor, J. R. (1921). Association as afundamental process of objective psychol-ogy. The Psychological Review, 28(6),385–424.

Kantor, J. R. (1924). Principles of psychol-ogy (I). Chicago: The Principia Press.

Kantor, J. R. (1947). Problems of physiolog-ical psychology. Chicago: The PrincipiaPress.

Kantor, J. R. (1957). Constructs and events inpsychology: Philosophy: Banished andrecalled. The Psychological Record, 7,55–60.

Kantor, J. R. (1958). Interbehavioral psy-chology. Chicago: The Principia Press.

Kantor, J. R. (1977). Psychological linguis-tics. Chicago: The Principia Press.

Kohn, A. (1999). Punished by rewards: Thetrouble with gold stars, incentive plans,A’s, praise, and other bribes. New York:Houghton Mifflin.

Lowenkron, B. (1998). Some logical func-tions of joint control. Journal of theExperimental Analysis of Behavior, 69,327–354.

Meyers, W. A. (1970). Observational learn-ing in monkeys. Journal of the Experi-mental Analysis of Behavior, 14, 225–235.

Miguel, C. F., Petursdottir, A. I., Carr, J. E.,& Michael, J. (2008). The role of namingin stimulus categorization by preschool


children. Journal of the ExperimentalAnalysis of Behavior, 89, 383–405.

Moore, J. W., & Fisher, W. W. (2007). Theeffects of videotape modeling on staffAcquisition of functional analysis meth-odology. Journal of Applied BehaviorAnalysis, 40, 197–202.

Morris, E. K., Higgins, S. T., & Bickel, W.K. (1982). The influence of Kantor’sinterbehavioral psychology on behavioranalysis. The Behavior Analyst, 5, 158–173.

Parrott, L. J. (1983a). Similarities anddifferences between Skinner’s radicalbehaviorism and Kantor’s interbehavior-ism. Mexican Journal of Behavior Anal-ysis, 9(2), 95–115.

Parrott, L. J. (1983b). Systemic foundationsfor the concept of ‘‘private events.’’ In N.W. Smith, P. T. Mountjoy, & D. H. Ruben(Eds.), Reassessment in psychology: Theinterbehavioral alternative (pp. 251–268).Lanham, MD: University Press of Amer-ica.

Parrott, L. J. (1986). On the role ofpostulation in the analysis of inapparentevents. In H. W. Reese & L. J. Parrott(Eds.), Behavior science: Philosophical,methodological, and empirical advances(pp. 35–60). Hillsdale, NJ: L. ErlbaumAssociates.

Pear, J. J. (2001). The science of learning.Philadelphia, PA: Psychology Press.

Pierce, D. W., & Cheney, C. D. (2008).Behavior analysis and learning (4th ed.).New York: Psychology Press.

Poppen, R. L. (1989). Some clinical impli-cations of rule-governed behavior. In S. C.Hayes (Ed.), Rule-governed behavior:

Cognition, contingencies, and instruction-al control (pp. 325–357). New York:Plenum.

Ramirez, J., & Rehfeldt, R. A. (2009).Observational learning and the emergenceof symmetry relations in teaching Spanishvocabulary words to typically developingchildren. Journal of Applied BehaviorAnalysis, 42, 801–805.

Rehfeldt, R. A., Latimore, D., & Stromer, R.(2003). Observational learning andthe formation of classes of readingskills by individuals with autism andother developmental disabilities. Researchin Developmental Disabilities, 24, 333–358.

Reiss, D. (1972). Vicarious conditionedacceleration: Successful observationallearning of an aversive Pavolovian stim-ulus contingency. Journal of the Experi-mental Analysis of Behavior, 18, 181–186.

Rosales-Ruiz, J., & Baer, D. M. (1997).Behavioral cusps: A developmental andpragmatic concept for behavior analysis.Journal of Applied Behavior Analysis, 30,533–544.

Skinner, B. F. (1953). Science and humanbehavior. New York: Free Press.

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

Skinner, B. F. (1974). About behaviorism.New York: Knopf.

Smith, N. W. (2006). The interbehavioralfield. In B. D. Midgley & E. K. Morris(Eds.), Modern perspectives on J. R.Kantor and interbehaviorism (pp. 87–110). Reno, NV: Context Press.

Smith, N. W. (2007). Events and constructs.The Psychological Record, 57, 169–186.


Understanding Observational Learning. an Interbehavioral Approach

Documents

Transcript of Understanding Observational Learning. an Interbehavioral Approach