The Behavior Analyst 2004, 27, 133-151 No. 2 (Fall)

Individual Behavior, Culture, and Social ChangeSigrid S. Glenn

University of North Texas

The principle of operant selection is examined as a prototype of cultural selection, and the role ofthe social environment is suggested as the critical element in the emergence of cultural phenomena.Operant contingencies are compared to cultural selection contingencies, designated as metacontin-gencies. Both of these types of contingency relations result in evolving lineages of recurrences thatcan become increasingly complex in the number and organization of their elements. In addition toits role in the recurring interlocking behavioral contingencies that constitute cultural organization,operant behavior plays another role in cultures. Although the operants of individuals are functionallyindependent of one another, the behavior of each person may contribute to a cumulative effect thatis relevant to the well-being of many people. Similarly, the outcomes of metacontingencies mayalso contribute to a cumulative effect. The relation between independently evolving operant lineages,or between independently evolving cultural lineages, and their cumulative effect is identified as amacrocontingency. Macrocontingencies do not involve cultural-level selection per se. Effective cul-tural engineering requires identifying the macrocontingencies that produce less than desirable effectsand altering the relevant operant contingencies or metacontingencies to produce change in the cu-mulative effects.Key words: operant contingencies, macrocontingencies, metacontingencies, cultural selection, op-

erant selection

Learned behavior is the substructureof human cultures, and the transmis-sion of learned behavior powers theevolution of human cultures. Humanbehavior produces cumulative changein human environments, and continu-ally changing environments requirecontinuing behavioral adjustments.Successful adjustments can becomeembedded in cultural practices andtransmitted to later generations.

Increasingly complex cultures haveemerged from the interplay among thehuman capacity for learning, the con-tingencies of reinforcement that ac-count for the learned behavior of in-dividuals, and the cultural transmissionof learned behavior all in the forma-tive context of physical features of lo-cal environments. Over a period of lit-tle more than 10,000 years, human cul-tures have evolved from small bands ofhunter-gatherers, presumably showing

I am deeply indebted to Leslie Burkett for hermultiple readings of the manuscript while it wasin preparation. Her critical feedback was invalu-able.

Address correspondence to the author at theDepartment of Behavior Analysis, P.O. Box310919, University of North Texas, Denton,Texas 76205 (e-mail: sglenn@unt.edu).

one another how to produce fire and tofashion simple tools, to huge nation-states in which the integrated activitiesof hundreds of people participate inproducing the fabric used to makeclothing sold as Brand X or to makethe laws by which millions of peoplelive. Decades of education, formal andinformal, are now required to developand maintain the behavioral repertoiresneeded to participate in the vast websof interrelated human behavior thatconstitute modern culture.'Most of the features of modern cul-

tures were not planned. Rather theysimply emerged as a result of the con-tingencies of selection that supportedthe behavior of individuals (cf. John-son, 2001). Systematic planning seemsto begin when cultural practices haveunpredicted, undesired, or belatedlyrecognized suboptimal results. Unin-tended and culturally damaging resultsof ongoing human behavior are firstidentified, then bemoaned and, some-times, finally dealt with. But can they

' The foregoing paragraphs draw on the workof Bonner (1980), Harris (1989), and Diamond(1997).

133

134 SIGRID S. GLENN

be dealt with fast enough to ensure sur-vival?Almost 20 years ago, B. F Skinner

(1987) asked why we were not actingto save the world. His answer was thatthe cultural practices of most of the hu-man race did not include the verbal be-havior required to properly analyze theproblems and plan the changes in theenvironment needed to promote cultur-al (and possibly species) survival. Themissing verbal cultural practice, Skin-ner suggested, was the language of theexperimental analysis of behavior-specifically, the practice of analyzingthe contingencies of reinforcement thatsupport the behavior of members of aculture and predict the results ofchanging those contingencies. He fur-ther suggested that the experimentalanalysis of behavior would support atheory of cultural evolution in the sameway that Darwin's theory of biologicalevolution is supported by the experi-mental science of genetics.

Skinner's (1987) analogy offers astarting point for exploring in this pa-per at least one way of using the lan-guage of the experimental analysis ofbehavior to support an interpretive the-ory of culture. Although Skinner's ul-timate interest was in bringing aboutchanges that would improve the livesof people, the point of his paper wasthat the path to effective action beginswith effective verbal practices. In2004, the 100th anniversary of hisbirth, humans need more than ever alanguage that will help them to analyzetheir problems in ways that guide ef-fective action.

In the sections below, I review theprinciple of operant selection and therole of human social environments inbehavioral contingencies. I then defineculture and cultural practices and con-sider the role of operant behavior inthem. Two types of cultural-level re-lations-macrocontingencies and me-tacontingencies are distinguished,and the role of behavioral contingen-cies in each type is explained as a pre-lude to accomplishing social change.Finally, I return to Skinner's views on

the relation between the verbal practic-es of the experimental analysis of be-havior and a theoretical interpretationof culture.

THE PRINCIPLE OFOPERANT SELECTION

Behavioral principles describe therelations between behavior and envi-ronment that account for the acquisi-tion and maintenance of learned behav-ior. The principle of operant selectionis the bedrock on which other behav-ioral principles rest. Like other scien-tific principles, its simple form masksthe complexity of the universe it de-scribes. The principle of operant selec-tion is sometimes stated as "behavioris a function of its consequences."Such a bald statement makes no men-tion of the different roles played bytime in the selection process describedby the principle. The statement alsoobscures the fact that the words behav-ior and its, which appear to refer to thesame thing, actually refer to differentthings. Let us unpack the statement inorder to lay the groundwork for a sim-ilar unpacking of cultural selection.Figure 1 is a schematic of operant se-lection as it goes on in time.2The schematic shows, first, that the

universe in which behavioral selectionoccurs is localized in the actions andevents outside those actions (environ-ment) of a single organism (let us saya young child learning to stack blocks).Of course, we would not be particular-ly interested in this universe if operantselection affected only this organism'sbehavior. As it turns out, the processappeared early enough in the evolu-tionary history of the earth to predatehumans by millions of years and there-fore is a behavioral characteristic

2 This example is made as simple as possibleand is not meant to suggest that the principlerequires simple responses, that instances are in-stantaneous, that bouts or sequences of respons-es cannot undergo selection, and so on. Thefunction of S I is not discussed because the sche-matic does not show differential control overRI.

BEHAVIOR AND CULTURE 135

Extended Time

Person P --------------------------1 minute-------------------------------------------------------

Situation S ................ . ...................................................... Environment

Cd g Occurrences R R2 R3 R4 R2 R5 I R6 R7 R8 1 R9 1 RIO RI Behavior

H Consequences Ci 1CI C Cl Cl Environment

Operant X

Figure 1. Recurrences of RI that occur with increasing frequency as a result of the operant con-tingency.

shared by many species, perhaps hu-mans most significantly.

In the schematic, three types ofevents are shown to be occurring in re-lation to one another in local time. Onetype of event comprises the actions ofPerson P, whose responses are desig-nated by the letter R followed by anumber. Each numbered R is a differ-ent way of manipulating a block. Tomake the example as simple as possi-ble, all Rs occur during a 1-min periodin a stable Situation S, which is presentwhen each R occurs (blocks are on aflat surface and everything in the play-room remains constant). In that situa-tion, the actions that an observer iden-tifies as RI (e.g., placing a blocksquarely on top of another) are reliablyfollowed by a particular type of stim-ulus change, Cl (a tower appears).That consequence does not follow theother actions occurring in Situation S.Situation S and the consequences rep-resented by Cl are environmental var-iables.The local temporal arrangements be-

tween each occurrence of RI and thesubsequent C I in Situation S areshown in the diagonal arrangements ofthose letters. The recurring temporalrelation between instances of RI (butnot other Rs) and Cl is typically des-ignated as a two-term contingency. Thebehavior identified as RI is shown tooccur several times. Only one other be-havior occurs more than once (R2),and it occurs only twice. Within the 1-min period of observation, RI can beseen to occur with increasing frequen-

cy and by the end of the minute, RIpredominates in Person P's behaviorstream.Assume that R1 has rarely, if ever,

occurred previously in Person P's be-havior stream. The schematic repre-sents the acquisition of an operant wehave labeled Operant X (in this case a"block-stacking" operant). The reli-able temporal relation between RI andCl (together with the absence of sucha temporal relation between other Rsand Cl) is the contingency that is caus-ing an increase in the frequency withwhich RI occurs in Person P's behav-ior stream in a stable environment.

Recall the principle "behavior is afunction of its consequences." Theparticular behavior that is a function ofconsequences in this example is Op-erant X (a lineage of RI responses).The "it" that is followed by a conse-quence is occurrences of RI (eachstacking response). The temporal (andusually causal) relation between occur-rences of R1 and occurrences of Clhas generated a lineage of RI that wehave labeled X. Both the lineage (thechild's block-stacking operant) and theindividual responses (each instance ofstacking) are behavior-specifically,the behavior of Person P. But they aretwo different kinds of things. More-over, the principle itself is content freeand refers to any and all lineages, andtheir component responses, that havebeen generated by contingencies be-tween responses and consequences.The value of the principle as a scien-tific generality depends on the variety

136 SIGRID S. GLENN

of the content that fits the formula andthe variety of organisms whose behav-ior can be predicted, changed, or use-fully interpreted in terms of the prin-ciple.The 1-min slice of a behavior stream

portrayed in Figure 1 shows a stablesituation that is present when all theresponses occur, and it shows a certainkind of environmental change follow-ing some of the responses. Changingthe situation to S2 during the next min-ute and ensuring that no responses arefollowed by CI in Situation S2 can adda level of complexity. Let us say thatin Situation S2, the surface on whichthe blocks rest is not flat. Alternationof S1 and S2 at varying intervals, withC1 occurring after R 1 only in S1, typ-ically results in high frequencies of RIin S 1 and low frequencies of RI in S2.The relation among S1-RI-Cl is thenspecified as a three-term contingency.In this case, the particulars of the three-term contingency are "on flat surfaces(but not on sloping surfaces), placingblocks squarely on top of one anotherresults in towers."

In a more dynamic and completeportrayal of a behavior stream, ele-ments of the situation are usuallychanging moment to moment and mostresponses would be followed by astimulus change of some kind (cf. Ray,Upson, & Henderson, 1977). The re-sulting interplay between environmen-tal changes and moment-to-momentchanges in a behavior stream is iden-tified as Person P's behavior.'

3Specific or particularized operant lineages(and their component occurrences) often areidentified colloquially as behaviors, implicitlysignifying their ontological status as individuallyidentifiable. For example, "The child engaged intwo problem behaviors" could mean he was ob-served repeatedly to hit other kids at lunch andalso to scream when recess ended. On the otherhand, everything one or all organisms do is des-ignated as behavior (a mass noun). Friman(2004) eloquently defended both usages on prac-tical grounds. Here it is suggested that the dif-ferent usages also respect an ontological distinc-tion. Specifically, the mass noun refers to a typeof empirical phenomenon (i.e., activity) and theplural usage suggests localization with respect toa specific person.

The actions that recur in operant lin-eages change over time as the contin-gencies of selection develop andchange. One way the operant lineageschange is that the component actionsbecome more complex. For example, achild may learn to plug cords into wallsockets, and to turn dials such as onthe kitchen timer, and to press hand-kerchiefs when his dad has finishedironing his shirt. In a situation callingfor a pressed shirt, variations of thethree responses may occur in a novelsequence, resulting in a pressed shirt.This coming together of responsesfrom operant lineages learned at dif-ferent times has been termed contin-gency coadduction (Layng & Andron-is, 1984). The relation between thenovel (adduced) response sequenceand the environmental consequencemay increase the likelihood of repeti-tions of the sequence, eventually re-sulting in a new operant lineage in thechild's repertoire. Call it ironing. Theironing lineage is a recurring sequenceof actions, each instance of the se-quence composed of elements origi-nally recruited from earlier acquiredoperant lineages. Because all the com-ponents of the sequence are required ineach occurrence to produce the conse-quence, the sequence of componentsacquires a functional integrity of itsown. The nesting or embedding of oc-currences of one lineage in more com-plex occurrences of another lineage isa highly consistent characteristic of hu-man behavior. An example of suchnesting is shown in Figure 2. Note thatthe lineage is always composed of re-curring events, and the increasing com-plexity is seen in the increasing num-ber of components in the occurrencesof the hierarchical lineages.

ENVIRONMENTAL CONTENTIN OPERANT CONTINGENCIES

The only necessary feature of theenvironmental events in three-termcontingencies is that they be empiricalevents. For everyday purposes, humanshave found it useful to categorize em-

BEHAVIOR AND CULTURE 137

Operant Lineages Elements ofEach Occurrence

(Recurring operant acts)

Press <Enter> Place digit on key

Depress key.....................- ..... . ............. .......... . .......

Open MS Word Locate program icon on desktop

Place cursor on iconPerson A

Press <Enter>

Copy email text into a Word Open MS Word

document Open email

Highlight text to be copied

Click "copy" icon

Switch to page in Word document

Click "paste" iconFigure 2. The components of occurrences in earlier lineages become integrated with the compo-nents of occurrences in later lineages of more complex occurrences.

pirical events in myriad ways: visual orauditory, temporally extended or punc-tate, verbal or nonverbal, social or non-social, and so on. Any single event canbe categorized in many different ways.Lightning as an event in one's environ-ment is visual, punctate, nonverbal,and nonsocial. In a child's environ-ment, my humming a tune can be cat-egorized as auditory, extended, non-verbal, and social. Another way thatempirical events can be categorized isin terms of their temporal relation tothe behavior of a learner or performer.The lightning may be a situation (an-tecedent) in which the behavior of ask-ing "lightning?" results in the conse-quence of "yes." My humming maybe the situation in which the child'shumming results in the sound of ourduet (consequence). Whether the

"yes" or the sound of the duet func-tions like C1 in Figure 1 is an empin-cal question. If they do, we considerthem as belonging to yet another cat-egory: reinforcer.The social environment is defined,

for present purposes, as the behavior ofother people as it relates to the behav-ior of a learner or performer. Yourquestion "What is your name?" is partof my social environment. My answer-ing with my name is a social event inyour environment. For any particularoccurrence of an operant response, thesituation may involve social and non-social events. One or more conse-quences also may be either social ornonsocial. Figure 3 provides examplesof three-term contingencies in whichvarious combinations of social andnonsocial events might function as sit-

138 SIGRID S. GLENN

Situation Response Consequence

Drink machine Deposit coins Cold drink

Five cups on table "One-two-three-four- Teacher nods and smiles

five cups"

Friend says "turn left" Turn left Arrive at destinationi~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...

Friend asks "What time is it?" "Six-fifteen" Friend says "Thanks"

Figure 3. Antecedent (situation) and consequent events may be nonsocial (italicized) or social(boldface) or a combination of social and nonsocial for responses in an operant lineage.

uation and reinforcing consequence inoperant contingencies. Environmentalevents are italicized, and boldfaceidentifies social events.The everyday distinction between

social and nonsocial events has en-couraged some authors to distinguishbetween "social learning" and "indi-vidual learning," the implication beingthat they involve different learningprocesses (e.g., Box, 1984). But suchimplication is not warranted because itconfuses process and content. Alllearning is individual learning (Galef,1988); that is, the locus of learningcomprises the events in the behaviorstream of an individual organism asthey relate to environmental events.The social character of some or all en-vironmental events in behavioral con-tingencies distinguishes the content ofthe environment, not the process bywhich that environment affects behav-ior. The behavioral processes are thesame, whether the environment thatfunctions to select the behavior is so-cial or nonsocial, and those processesare a biological given.The social content in the contingen-

cies that support most of the learningaccomplished by humans is a definingfeature of human cultures. Althoughrudimentary cultures are seen in otherspecies (Bonner, 1980; Harris, 1989),only humans depend almost entirely oncultural transmission of behavioralcontent for the survival of their spe-cies. How could natural selection have

lost so much direct control over thesurvival of one of its creations? Andwhat genetically transmitted humancharacteristics lengthened the leash ofnatural selection so dramatically?

BIOLOGICAL INHERITANCEAND CULTURE

In addition to a distinct combinationof anatomical and physiological char-acteristics such as opposable thumbs,bipedal locomotion, acute hearing andvision, a highly plastic brain, and anintricate and flexible vocal apparatus-the human species has several geneti-cally transmitted behavioral character-istics that appear to have underpinnedthe emergence of culture. The first be-havioral characteristic that made hu-man cultures possible is their sociali-ty-the strong tendency of humans tospend most of their time in close prox-imity to one another. Without this ten-dency, there would be little possibilityfor social events (i.e., the behavior ofconspecifics) to serve as environmentalevents having function with respect tohuman behavior.

Skinner (1984b) identified two otherbehavioral characteristics that togetherconstitute what may be called, some-what vaguely, "learning potential."First, humans are born with a reper-toire of "uncommitted behavior,"which distinguishes them from socialspecies such as ants, whose specific be-havior-environment interactions are

BEHAVIOR AND CULTURE 139

highly constrained by inheritance.Whereas ants, for example, inherit spe-cific behavioral patterns in response tospecific social events, human behaviorbecomes related to any of a wide va-riety of environmental events, depend-ing on the particulars of the social andnonsocial environment in which thehuman lives. To be sure, human actionis constrained by human biology; butin each generation, humans have tolearn all over again what their ances-tors learned-a laborious process in-deed, but one that allows the behaviorof each generation to become adaptedto current environmental events. Andthose environmental events can changefrom generation to generation and havedone so at a steadily increasing pace.By outfitting humans with a largely

uncommitted behavioral repertoire,natural selection gave our species along leash for local behavioral adap-tations. But the uncommitted repertoireof humans would be lethal without thesecond characteristic of human learn-ing potential-the susceptibility of hu-man behavior to operant selection. Al-though this behavioral characteristic isshared by many species, humans ap-pear to be most exquisitely sensitive tobehavioral contingencies of selection(Schwartz, 1974). This characteristicdoes not depend on whether the envi-ronmental events in the behavioral se-lection contingencies are social or non-social, but the preponderance of socialevents in the behavioral contingenciesall but guarantees the emergence ofcultural phenomena.

EMERGENCE OFCULTURAL PHENOMENA

The combination in humans oflearning potential and sociality set thestage for the emergence of culture-anovel kind of phenomenon. Like theword behavior, culture is a mass noun,a category word, and also a word thatrefers to the particulars that are mem-bers of that category (specific cul-tures). As a category of phenomena,we will define culture here as "patterns

of learned behavior transmitted social-ly, as well as the products of that be-havior (objects, technologies, organi-zations, etc.)."

Culture begins with the transmissionof behavioral content, learned by oneorganism during its lifetime, to the rep-ertoires of other organisms. Thus, thelocus of cultural phenomena is su-praorganismic. Unlike learning, whichis localized in repeated temporal rela-tions between the actions of a singleorganism and other empirical events,the locus of cultural things is supraor-ganismic because it involves repeti-tions of the interrelated behavior oftwo or more organisms; one organism'sbehavior functions as the situation orconsequences in the operant contingen-cies accounting for the behavior of theother. Such transmission requires nonew biological trait or behavioral pro-cess, but it does initiate a new kind oflineage: a culturo-behavioral lineage(Glenn, 2003).

Culturo-behavioral lineages extenddeeply into human history, and theyalso occur in rudimentary form amongnonhuman species (Kawamura, 1959).A curious fact about human cultures isthat after anatomically modern humansspent tens of thousands of years in ru-dimentary cultures, human cultures be-came quite complex in little more than10,000 years-an extraordinarily shortperiod of evolutionary time (Harris,1989). This suggests that long beforecultural takeoff, humans had the ana-tomical, physiological, and behavioralcharacteristics (delineated above) thatthey needed for the emergence andevolution of complex cultures. Missingwere the changes in environmentalevents (social and material) that couldenter into operant contingencies acrossgenerations, supporting individual be-havior that differed from generation togeneration.

Cultural Practices andMacrocontingencies

Much of the behavioral content ofindividual human repertoires is similar

140 SIGRID S. GLENN

to the content of many other humans.The term cultural practices refers tosimilar patterns of behavioral content,usually resulting from similarities inenvironments. The term metabehaviorhas been suggested to identify the classof behaviors that constitute a culturalpractice (Mawhinney, 1995). The needfor a term subsuming a supraorganis-mic class of behaviors is recognized,but we will use the term macrobehav-ior here because it is consistent withthe other terminology in this paper.

Cultural practices may be importantor unimportant for the survival of aculture. An example of a practice notlikely to be critical for cultural survivalis hairstyling. Many hairdressers maystyle hair similarly, and this similarityof behavioral content constitutes a cul-tural practice. Important to note is thatsuch similarity does not imply that thepractice is a functional cultural unit. Inother words, the behavior of the vari-ous hairdressers is not necessarilyfunctionally related to the behavior ofany other hairdressers. Individual hair-dressers simply may learn over time tocut certain types of hair in certain waysas a result of the consequent look ofthe product and approval of their pa-trons. The resulting products (hair-styles) consequently look alike. Nei-ther the hairstyles nor the behavior ofthe hairdressers are functionally relatedto one another, even though the behav-ior of each hairdresser interrelates withthe behavior of each of his or her pa-trons. In this case, the similar behaviorof many individuals constitutes a cul-tural practice, but there is no evidenceof cultural transmission and, therefore,no culturo-behavioral lineage exists.On the other hand, there may be a

point of cultural transmission that linksthe behavior of two or more hairdress-ers. For example, Hairdresser A maydemonstrate to other hairdressers away to style hair, and the others mayreproduce the style under the watchfuleye of the originator and later withtheir own patrons. If the hair styled byA is featured in a magazine or seen oncustomers by other hairdressers, some

of them may be able to produce a sim-ilar result for their own patrons. Thesecases involve cultural transmission.Any cultural practice may be made upof independently generated behaviorsand also socially transmitted behaviors.The point of these two examples is thatsimilarity in behavioral content ofmany individuals is sufficient to con-sider the aggregate behavior a culturalpractice, but is not sufficient to assumecultural transmission, and is even lesssufficient to assume a common origin.

Another way of distinguishingamong cultural practices is in terms ofthe complexity of the behavioral con-tent that constitutes the practice. Themacrobehavior that constitutes a spe-cific cultural practice may be straight-forwardly identifiable operants such assmoking cigarettes; or multioperantpatterns of behavior such as stylinghair, driving to work, or recycling; orvery complex patterns of interlockingbehavior of many individuals, such asthat involved in auto manufacturing.Whether comprising simple or com-plex elements, cultural practices allhave two characteristics that are im-portant for the present discussion.First, they involve many people en-gaged in the same repeated actions (be-having individually or in relation toone another) and, second, those actionshave consequences-often several dif-ferent consequences.

Consider the behavior of driving towork. A consequence essential to itscontinuing repetition in an individual'sbehavior stream is arrival at work. Butin most cases, there are other behaviorsthat could result in arriving at work(e.g., carpooling, using mass transpor-tation, bicycling, or walking). The factthat most people drive to work ratherthan getting there some other way sug-gests that additional consequences areinvolved and that they differ for dif-ferent behaviors. Figure 4 shows somelikely consequences of driving to workversus carpooling.

Note that all of the consequencesshown in italics depend only on the be-havior of the individual worker, and

BEHAVIOR AND CULTURE 141

Behavior iConsequences Likely to Consequences Likely Contribution to

Increase Frequency to Decrease Frequency Cumulative Effect

iDriving to- work Most convenient Highest cost More pollution

Takes less time

Carpooling to Lower cost Least convenient Less pollution

Work Takes more time

Figure 4. Effects of behavior that can function as behavioral consequences are in italics and thosethat cannot function as behavioral consequences are in boldface; there is no contingency betweenthe operant and the cumulative effect.

they are experienced within a shorttime after the behavior occurs. Al-though the worker does not control thecost of gasoline, he or she does havecontrol over whether more or less ofhis or her paycheck goes for gasoline,and whether he or she takes more orless time, with more or less conve-nience, in getting to work. Because ofthe correlation between the behaviorand those consequences, those conse-quences have some potential to in-crease or decrease the relative proba-bilities of driving or carpooling. Whenmultiple consequences have conflictingfunctions, the overall effect of the con-sequences may be the algebraic sum-mation of their individual effects (seeSkinner, 1953, pp. 218-223). And theeffect of each of these consequences isrelative. For example, if a worker whodrove to work was transferred to awork site much farther from home, thechange in relative value of gasolinecost versus convenience might makecarpooling more likely than before, es-pecially because the time involved incarpooling versus driving might notdiffer much for the longer drive.The effect of our worker's behavior

on air pollution, shown in boldface inFigure 4, is a very different kind ofconsequence. It is not only a matter ofthe consequence being too small, toodelayed, or too cumulative for it tohave a behavioral function, althoughall of that is true and important (see R.

W. Malott & Suarez, 2004). There issomething else that sets apart the effecton air pollution from the other effectsof the two behaviors. To wit, even if,by some magic, we were able to givethis consequence a powerful function,the consequence itself can be nullifiedby the behavior of other people. Ourworker, mightily motivated to havecleaner air, can carpool for the next 20or 30 years, but if a lot of other peopledo not do the same thing, the air is notgoing to be any cleaner. In otherwords, cleaner air is simply not underour worker's control. That is, as Hardin(1968) succinctly put the matter, thetragedy of the commons. No matterhow much one behaves for the com-mon good, the behavior of others canundo it all. That is the critical differ-ence between the italicized and bold-faced consequences listed in Figure 4.Ulman (1998, p. 209) suggested the

term macrocontingency to define "aset of differing actions (topographies)of different individuals under commonpostcedent control." The term and thedefinition suggest the standard defini-tion of an operant writ large. It couldbe parsed in two ways. Macrocontin-gencies could refer to commonalities inbehavior-consequence relations acrossmany individuals, or it could refer tothe control exercised by the cumulativeeffect of all the topographies on the to-pographies. Ulman makes it clear thatthe "common postcedent control" is

142 SIGRID S. GLENN

Macrocontingencies

Operant Selection

Contingencies

People

P1 ...... Bx (1) -> C ............ Contributes to cumulative effect

P2 ...... Bx (1') -> C .............Contributes to cumulative effect

P3 ...... Bx (1) -> C ..............Contributes to cumulative effect

P4 ...... Bx (1') -> C ............ Contributes to cumulative effect

Pn ...... Bx(1,1') -> Cn ............ Contribute to cumulative effect

(Macrobehavior)* Cumulative effect

Cultural Practice 1Figure 5. Temporally unrelated operants of different people (macrobehavior) that produce behav-ioral consequences and also contribute to a cumulative effect.

the cumulative effects of those differ-ing actions. As mentioned above, how-ever, the cumulative effects cannot bein a contingent relation with the behav-ior of any individual; therefore theycannot control (as operant consequenc-es) the behavior of individuals. And al-though there may be a contingent re-lation between the sum of the topog-raphies and the cumulative effect, thesummed topographies are not part of alineage that can wax or wane togetheras a function of the postcedent. If thepostcedent has any effect at all on anyoperant lineages of individual people,that effect is independent of any effectit may have on operant lineages of oth-er people.4

4Macrocontingencies as here defined can in-volve different topographies of different people,

That being said, the notion of somekind of relation that is bigger than op-erant contingencies seems useful. So Iwill define a macrocontingency as therelation between a cultural practice andthe aggregate sum of consequences ofthe macrobehavior constituting thepractice. Figure 5 shows the relationsin a macrocontingency as here defined.The recurring behavior of each personhas its own effects, and the relation be-tween the behavior and that effect can

the aggregate results of which are a change inthe environment of many people. Todorov, Mor-eira, and Moreira (2004) provide examples ofsuch relations. The aggregate results of the dif-fering topographies in their examples, as in theair pollution example here, cannot have a selec-tive function on those topographies because ofthe poor correlation between the behavior of anyindividual and the aggregate result.

BEHAVIOR AND CULTURE 143

alter the probability of the recurrenceof that individual's behavior (as in Fig-ure 1). For example, if the behavior isdriving to work, then each person'sdriving-to-work operant is a functionof the contingency between driving towork and the operant consequences ofthat behavior. In addition to those in-dividuated consequences, the com-bined behavior of all the people (themacrobehavior) has a cumulative ef-fect. This effect cannot function as abehavioral consequence because it isnot contingent on the behavior of anyindividual driver. It is contingent onthe macrobehavior of the cultural prac-tice.An important feature of macrocon-

tingencies is that their cumulative ef-fects are additive. The more wide-spread a practice, the greater its cu-mulative effects; the greater the cu-mulative effects, the more importantthey are to the well-being of largenumbers of people. Each person con-tributing to the cumulative effect con-tributes in direct proportion to the fre-quency of his or her behavior. It is thecumulative effect of the behavior in acultural practice that constitutes aproblem for the people of a culture. Tocontinue with the example, the drivingbehavior of each individual is as it isbecause of the relative effects of itsmultiple behavioral consequences: ar-riving at work in good time with min-imum difficulty and the money spenton gasoline. These consequences thatmaintain the driving behavior contrib-ute to the probability of driving, butthey are not the culturally relevant cu-mulative effects: gasoline consumptionand associated environmental effects.Further, the behavior is not a problemfor the individuals behaving-rather itis a solution, albeit not an ideal solu-tion, to the problem posed by their dis-tance from work. As in the case of thebehavior of individuals, cultural prac-tices also have multiple consequences.For example, two effects of consumerbehavior are that it helps to create jobsand it contributes to degradation of thephysical environment. Such incompat-

ible effects of cultural practices areeven more difficult to reconcile thansimilar incompatible effects of individ-ual behavior. That is because the mul-tiple cumulative effects of any givencultural practice are likely to be moreadvantageous to some people and moredisadvantageous to others. In the caseof individual behavior, at least the costsand benefits affect the same person.

Discussion of macrocontingencieshas centered on the cumulative effectof many people "doing the samething" (allowing for a broad range oftopographies). The people could beacting individually (e.g., smoking), ortheir behavior could be interrelated(e.g., carpooling). Either way, the sim-ilarity in operant content of many peo-ple is what warrants our calling it acultural practice. Each time the behav-ior occurs, it adds to the cumulative ef-fect. So the cumulative effect dependson the number of times the act occurs,and that number is a function of thenumber of people who engage in theact and the frequency of the behaviorof each person.The relation between any particular

cultural practice and its cumulative ef-fect may be critically important to thewelfare of the people of the culture,and even to the survival of that culture.But a cultural practice (as here definedand as generally, albeit vaguely, un-derstood) cannot participate in a selec-tion process. That is so because a cul-tural practice is a class of acts that arefunctionally independent of one anoth-er. In other words, recurrences of theacts do not participate in a lineage.They are classified as "the same" interms of their form and their effects,but the members of the class are notnecessarily related by descent, which isa defining feature of evolution by se-lection (Hull, Langman, & Glenn,2001). In short, a cultural practice doesnot evolve as a result of cultural selec-tion, but rather as a result of behavioralcontingencies of selection operating onthe behavior of many individuals; as aresult, a different cultural practicecomes to exist. For example, the cul-

144 SIGRID S. GLENN

tural practice of smoking in publicbuildings has been replaced in manyareas by a practice of going outsidepublic buildings to smoke. The changein the cultural practice is a behaviorchange of many individuals, each re-sponsive to his or her own social en-vironment. When there is change in thepractices that constitute a culture, thechange is an emergent side effect ofconcurrent changes in behavioral line-ages of many individuals. The causalmechanism is behavioral selection (i.e.,the principle of reinforcement).

If certain human cultural practices,or their cumulative outcomes, threatenthe safety of the world, then saving theworld will necessarily entail alteringthe operant contingencies that maintainthe behavioral lineages that contributeto those outcomes. Such action re-quires interpretation of complex phe-nomena in the language of the experi-mental analysis of behavior (see Palm-er, 1991). Cultural change will be dif-ficult to accomplish, as Skinner (1987)suggested, when the verbal practices ofthose cultures do not include the lan-guage of the experimental analysis ofbehavior. On the positive side, whenthat language does guide action de-signed to bring about changes in ma-crobehavior, those actions can be quitesuccessful.What about Skinner's (1987) sug-

gestion that the experimental analysisof behavior would support a theory ofcultural evolution in the same way thatDarwin's theory of biological evolu-tion is supported by the experimentalscience of genetics? This suggestionneeds a great deal of clarification be-cause the relations between genes andthe species that carry them are ex-tremely complex, as are the scientificfields of evolutionary biology and ex-perimental genetics. Perhaps most im-portant to the present discussion is theincreasing complexity (of organisms'structure and function as well as eco-logical relations among them) that hascharacterized the evolution of organicphenomena. It appears that human cul-tures, too, have been characterized by

organizational structures and functionsthat have become increasingly com-plex throughout human history. Theconcept of metacontingencies mayhelp us to understand how that com-plexity evolved.

METACONTINGENCIES

A clear distinction between the con-cepts of metacontingencies and macro-contingencies is needed, because earlypapers introducing the concept of me-tacontingencies (Glenn, 1986, 1988)combined terminology suitable for dis-cussion of macrocontingencies (as heredefined) and metacontingencies (ashere defined). The prefix meta- togeth-er with the root contingencies is in-tended to suggest selection contingen-cies that are hierarchically related to,and subsume, behavioral contingen-cies. They represent "a different kindof selection," although "no new be-havioral process" is involved (Skinner,1984a, p. 504). Metacontingencies arenot a matter of an enlarged class of be-havior or more widespread behavioralcontingencies; rather, they are the en-gine of a different kind of selection.The metacontingencies of cultural se-lection emerged only after socialevents become prevalent in the behav-ioral environment of a species that hasthe human combination of physical andbehavioral traits.

The concept of metacontingenciesaddresses evolution by selection whenthe lineages that evolve are not the re-curring acts of individuals (as sche-matized in Figure 1), but rather are re-curring interlocking behavioral contin-gencies (IBCs) that function as an in-tegrated unit and result in an outcomethat affects the probability of future re-currences of the IBCs. Figure 6 is aschematic of the metacontingencies ofcultural selection as it goes on in time.The recurring IBCs comprise operantcontingencies in which the behavior oftwo or more people functions as envi-ronmental events for the behavior ofthe others. The outcomes produced byrecurrences of the IBCs are not the cu-

BEHAVIOR AND CULTURE 145

Extended Time

Persons P1I and P2 -------------------------- I year----------------------------------------------------------------------------

Condition S ....... Environment

Q Occurrences IBCI IBC2 IBC3 IBC4 IBCI IBC5 IBC6 IB IBC7 IBC8 IBC1 IBC9 IBCI IBCI Recurrences

. u Outcomes 0 01 01 01 01 Environment

Cultural Lineage X

Figure 6. Recurrences of IBC 1 that occur with increasing frequency as a result of metacontingency.

mulative effect of the participants be-having individually, but rather the ef-fect of their interrelated behavior. Forexample, Marta and Todd regularlycook meals together. Marta preparesentrees, sauces, and vegetable disheswith Todd serving as helper, and Toddprepares appetizers and desserts withMarta serving as helper. The timing ofeach of their activities is based on whatthey observe the other one doingthroughout meal preparation. The out-come of their interrelated behavior is ameal with perfectly timed courses ofperfectly prepared dishes. The mealcould not be produced by Todd andMarta working in separate kitchens andcombining the results of their individ-ual behavior. Thus, it is not the cu-mulative effect of their individual be-haviors. It is the outcome of their in-terrelated behavior.

Metacontingencies, then, are thecontingencies of cultural selection.They give rise to the organized collec-tions of behavioral contingencies thatconstitute increasingly complex cultur-al-level entities. Let us continue withthe example of the relation betweenTodd's and Marta's IBCs and the re-sulting meals. Variations in the featuresof the IBCs will result in variations inthe outcome, and if the difference inoutcomes perpetuates some patterns ofthe IBCs more than others, cultural-level selection has occurred. Note thatTodd's behavior is a function of behav-ioral contingencies that might includethe taste of the meals cooked, and Mar-ta's behavior is a function of other be-havioral contingencies that might in-clude the taste of the meals cooked.

Those behavioral contingencies arenecessary for the continuation and evo-lution of Todd's and Marta's operants,and thus of the IBCs; but they are notnecessarily sufficient for the IBCs. Theoutcome of the IBCs must be morethan or different than the meals that ei-ther Todd or Marta could produce bythemselves to maintain the recurrencesof the IBCs. It is this "more than" or"different than" that is the source ofcultural evolution and what distin-guishes it from behavioral evolution.

CULTURAL COMPLEXITY

Cultural complexity is the outcomeof cultural selection that results in nest-ed hierarchies of IBCs (Glenn & Mal-ott, in press). For example, Todd andMarta may open a restaurant wherecooking meals is part of a larger pat-tern of recurring IBCs. Figure 7 showsa nesting of IBC relations in increas-ingly complex cultural lineages.Whether the larger pattern continues torecur and evolve depends on the out-comes of cooking but also on the out-comes of other IBCs in the situation.The behavior of other people may be-come part of the larger pattern andcontribute substantially to the outcomethat maintains the continuing recur-rences of the IBC that constitute "thebusiness." Finally, although the IBCsmust continue to recur for the culturallineage to remain in existence, it is notnecessary that Todd's or Marta's be-havior continues to participate. The be-havior of other individuals can replaceone or both of theirs as long as thatbehavior fits well enough into the IBCs

146 SIGRID S. GLENN

IOrganisms Cultural Lineages Some Elements of Each Occurrence

(Recurring Interlocking

Behavioral Contingencies)

'Todd & Marta Cook meals Many behaviors of each person having

function with respect to behaviors of other

Todd, Marta, Operate a restaurant Cook meals

2 waiters, Serve diners

Ilbookkeeper Manage cash flow

Todd, Marta, many Run a franchise chain Operate restaurants

waiters, bookkeepers,,

other personnel

Figure 7. The components of occurrences in earlier IBCs become integrated with the componentsof occurrences in later IBCs of more complex occurrences.

to produce the outcome. Perhaps it isworth noting that such replacement ofone participant's behavior for another'sin a cultural lineage virtually alwayscauses some adjustments in the IBCsand thus always presents both oppor-tunity and threat to the continuing sur-vival of the lineage.

Like the responses in operant contin-gencies, the IBCs in metacontingenciescan result in both automatic outcomesand socially mediated outcomes thatdepend on the features of the automaticoutcome. For example, Todd's andMarta's IBCs at first had automaticoutcomes-meals-that differentiallyperpetuated some variations of theIBCs. Eventually, the IBCs constitut-ing their restaurant were maintained bythe ordering behavior of customers. Asin the case of social reinforcers for in-dividual behavior, the socially mediat-ed relation between the IBCs of therestaurant and the sustaining incomegenerated from customer purchasesprovides a foundation for more com-plex relations.The nested metacontingencies of

cultural selection are the basis for theevolution of cultural complexity aswell as the maintenance (survival) ofevolving organizational lineages. Justas components of one operant lineagebecome embedded in operant lineagesof more complex components (as inFigure 2), components of one lineageof IBCs can become embedded inIBCs of greater complexity (as in Fig-ure 7). These more complex culturalentities are the individually identifiableevolving units we know as organiza-tions: individual companies, their par-ent corporations, schools, school dis-tricts, universities, university depart-ments, government agencies, and soon. Each of these units exists as longas it consists of IBCs that produce anoutcome that can increase the likeli-hood that the IBCs will recur. Theseare all entities that can change orevolve over time or that can disappearas a whole. They are not themselvescultural practices, because each orga-nization is an entity-an evolving lin-eage of IBCs.

Before proceeding to the engineer-

BEHAVIOR AND CULTURE 147

ing of cultural change, let us review thesimilarity in relations that constituteoperant contingencies and metacontin-gencies. Whether or not a behaviorallineage continues to exist and evolvedepends on the automatic or sociallymediated consequences that follow therecurring behavioral instances. Thismeans that behavior change requiresthat consequences be variable; that is,the consequences can either occur ornot occur, or can occur in differingamounts or after differing time lags,depending on the characteristics of thebehavioral instances that produce thechanges in the environment. An ante-cedent event, endogenous or exoge-nous to the behaving organism, can af-fect a given occurrence, but recurrenc-es are a function of consequences. Sobehavior changes the environment, andthe resulting changes may, in turn, alterthe future probability of that kind ofbehavior. Another way of saying this isthat behavioral instances cause envi-ronmental changes, and systematic re-lations between behavior and conse-quences feed back into the ongoingsystem, causing changes in the fre-quency and characteristics of future in-stances (i.e., continuation and adapta-tion of the lineage).The relation between IBCs and their

outcomes has functional parallels tothe complex relations of behavioralcontingencies. The IBCs produce out-comes, variations in instantiations ofIBCs cause differential outcomes, andthe future frequency of the IBCs aswell as their characteristics are a func-tion of the differential relation betweeninstantiations and outcomes. Changesendogenous or exogenous to the IBCsmay result in a variation that producesa different outcome, and that outcomecan increase or decrease the probabilityof recurrences of the IBCs.

ENGINEERING CHANGE

Humans have been engaged in be-havioral engineering since they beganfunctioning as the environment in theoperant contingencies that support the

behavior of other humans. Only in the20th century of the current era, how-ever, has scientific understanding beenbrought to bear on these engineeringpractices. Cultural engineering has notyet found sure scientific footing. Anecessary first step is to understand thephenomena to be engineered. From thepresent perspective, engineering canoccur with respect to two kinds of phe-nomena: macrobehavior and metacon-tingencies.The macrobehavior of cultural prac-

tices can be identified as a problemonly when its cumulative effects arerecognized, and it often takes a longtime to gain understanding of the manyeffects of specific cultural practices.The only way to do something aboutthe cumulative effects of macrobehav-ior is to find ways to alter the behaviorof as many individual participants aspossible. For example, the more indi-viduals who carpool or take publictransportation to work rather than drivealone, the greater is the improvementin air quality (or the slower the wors-ening of air quality). When the numberof participants in a practice is large, achange in the behavior of a small per-centage of them can make an importantdifference. If 10% of the drivers in theU.S. carpooled with two other people,a noticeable reduction in air pollutionmight result. What could bring aboutsuch a change in the behavior of 10%of drivers? Considering that each driv-er's behavior is a function of the op-erant contingencies in effect, we mustconsider the consequences of the be-havior of driving to work versus theconsequences of carpooling, as dis-cussed previously and shown in Figure4. The assumption is made, for pur-poses of discussion, that the effects ofthe behaviors listed in Figure 4 couldfunction as behavioral consequences,with the exception of the effect on airpollution.As matters now stand, attempts by

society to engender alternative macro-behaviors are implemented with littleunderstanding of the potential cumu-lative effects (Nevin, 1998), and little

148 SIGRID S. GLENN

attention is paid to the many operantcontingencies that may be maintainingthe operant behavior of individual par-ticipants in the current practice. Be-cause the macrobehavior of culturalpractices is a function of operant con-tingencies that operate independently,but concurrently and similarly, on thebehavior of many people, behavior an-alysts have rightly called for analysisof the contingencies that maintain thebehavior that constitutes the practice.Mattaini (1995), in particular, has ar-gued that behavior analysts should betrained specifically to focus on behav-ior with cumulative effects that affectthe viability of the culture. When in-terventions are designed to alter the cu-mulative effect of a cultural practice,they must necessarily identify the op-erant contingencies that account for thebehavior of individuals who participatein the practice. The more individualswhose behavior changes, the greater isthe impact on the cumulative effect.This method of cultural interventionentails modifying the operant contin-gencies that are likely to maintain thebehavior of large numbers of people.Biglan (1995) described many of thebehaviors of modern American culturalpractices that result in undesirable cu-mulative effects, and he identifiedmany of the socially mediated behav-ioral contingencies that support thosebehaviors. Other authors (e.g., Gold-stein & Pennypacker, 1998; R. W. Mal-ott, 1998) have offered interpretationsof various specific macrobehaviors andsuggestions regarding intervention.Under the editorial guidance of Rich-ard Rakos, Janet Ellis, and Mark Mat-taini, the journal Behavior and SocialIssues has devoted several issues toanalyses of macrobehaviors with high-ly destructive cumulative effects.

Because much of the operant behav-ior of modern humans is embedded inorganizations that have recurring IBCs,survival of those organizations is, atthe very least, important to those hu-mans. The fact that the organizationsexist at all, however, suggests that theirIBCs were selected by their external

environment and, therefore, are an im-portant part of the larger culture,whether or not alternative organiza-tional structures are considered moredesirable. Engineering, then, can alsooccur with respect to the IBCs in me-tacontingencies.

IBCs can be changed in two waysthat are analogous to the two ways thatspecies characteristics can be altered.The first is by altering the external se-lecting environment and waiting forvariations in the IBCs to produce out-comes suitable to the new selectioncontingencies. This amounts to alteringthe contingencies of selection and let-ting the chips fall where they may. Thesecond way is similar to ascertainingand altering the genetic characteristicsthat are endangering a species' exis-tence given the current selecting envi-ronment. This tactic entails altering thecomponents of the IBCs so that theyare better adapted to the current se-lecting environment. Planned varia-tions of the recurring IBCs can be de-signed to produce outcomes more suit-able to the demands of the external en-vironment.

Engineering change to enhance thesurvival of organizations (recurring ar-rangements of IBCs) requires analysesof current metacontingencies and alsoanalyses of the specific behavioral con-tingencies that affect the outcome ofIBCs. It should be obvious that all ofthe IBCs and the operant contingenciesin complex organizations cannot be an-alyzed. There must be some way todistinguish between those that can beignored and those that must be ad-dressed. M. E. Malott (2003) describedan approach to organizational changethat combines a behavioral systems en-gineering model with metacontingencyanalysis. Her collaborations with thepersonnel in business organizations aswell as in at least one institution ofhigher education (M. E. Malott & Sa-las-Martinez, 2004) demonstrate theimportance-indeed, the necessity-ofisolating the IBCs that fail to meet se-lection contingencies and then identi-fying the operant behavior that must be

BEHAVIOR AND CULTURE 149

altered to bring about the kind ofchanges in IBCs required by the exter-nal environment.

In summary, to bring about changesin the organized IBCs that function asevolving cultural units, it is necessaryto identify the IBCs that contribute toan outcome and to identify the functionof the outcome in sustaining (or not)recurrences of the IBC. Variations canbe made in the IBCs by systematicallymanipulating the behavioral contingen-cies within them, and the variationsmay increase or decrease the probabil-ity of producing an outcome with asustaining function.

RELATION OFMETACONTINGENCIES

TO OPERANT CONTINGENCIESAND MACROCONTINGENCIES

Metacontingencies, like behavioralcontingencies, involve two kinds ofcausality, as can be seen by comparingFigures 1 and 3. First, the recurrencesof IBCs produce outcomes (analogousto consequences produced by recur-rences of operant responses). Second,the outcomes affect the future frequen-cy and other measures of the future re-currences of those IBCs. The contin-gencies of selection in metacontingen-cies are between cultural-level units(IBCs) and their selecting environ-ments. Evolving cultural units are re-curring cycles of IBCs. Like operantsin a repertoire, the recurring entitiesmay become part of increasingly com-plex entities that form a lineage of theirown (see Figure 7). The outcomes pro-duced by a cycle of IBCs can affectfuture cycles of IBCs, just as the con-sequences of a behavioral occurrencecan affect future occurrences of thatbehavior. If one is interested in alteringthe recurrences of IBCs, one can do soby altering the components of IBCs tobetter meet current selection require-ments or by altering the selecting en-vironment. The former strategy wouldbe comparable to genetic alteration andthe latter to artificial selection.The IBCs in metacontingencies, like

the individual behavior in operant con-tingencies, recur in lineages that evolveand change as a function of their se-lecting environments. They are alsoalike in their relation to macrocontin-gencies. Just as the similar operant be-haviors of many people can contributeto a cumulative outcome, the IBCs ofseveral different organizations mayalso contribute to a cumulative out-come, as shown in Figure 8. The be-havioral lineages of the different peo-ple who participate in a cultural prac-tice evolve independently, as do theIBC lineages of the different organi-zations. But both the behaviors and theIBCs may also contribute to a cumu-lative outcome that plays no direct rolein selection but nevertheless may beimportant indicators of the viability ofthe culture.

The organizations in Figure 8 couldbe programs comprising IBCs that pro-duce graduates trained as behavior an-alysts. Each program produces gradu-ates (among other things) whose per-formance contributes directly and in-directly to the selection of the recurringIBCs that produce cohort after cohortof graduates. The IBCs that producebehavior analysis graduates constitutea cultural practice and they have cu-mulative effects, including the numberof people prepared for academic ap-pointments, the number of individualswho can be served by professional be-havior analysts, the amount of federalfunding likely to go to behavior-ana-lytic researchers, and so on. Althoughindividual behavior analysts and theprogram faculty of individual programscan be moved to action by data on thecumulative effects of the summed be-havior or summed IBCs, those effectscannot select any of the individual op-erant lineages or the individual line-ages of IBCs, because there is no lin-eage of recurring entities that producesthose effects. If one is interested in al-tering the cumulative outcomes of acultural practice, one must find a wayto alter the behavioral contingencies ofmacrobehaviors or the metacontingen-cies supporting the IBCs of organized

150 SIGRID S. GLENN

Macrocontingencies

Metacontingencies

Organization 1... IBC (1) 0.Contributes to cumulative effect

Organization 2 ...... IBC (1) -> ......... Contributes to cumulative effect

Organization 3 ...... IBC (1') -> ......... Contributes to cumulative effect

Organization 4 ...... IBC (1) -> ......... Contributes to cumulative effect

Organization.5 ...n... IBC (1,1) -> 0 ......... Contribute to cumulative effect

(Organizational Practice)Cumulative Effects

Cultural Practice XFigure 8. Temporally unrelated IBCs of different organizations (organizational practices) that pro-duce behavioral consequences and also contribute to a cumulative effect.

cultural complexity. The more individ-ual contingencies or organizationalmetacontingencies that are altered, thegreater the potential change in the cu-mulative outcome.The larger the number of organiza-

tions characterized by the same kindsof IBCs, the more likely we are to con-sider those kinds of IBCs a culturalpractice. Statements such as "the cul-tural practices of Japanese businessesare different from the cultural practicesof American businesses" refer to sim-ilarities in the IBCs that characterizeAmerican companies and similaritiesin the IBCs that characterize Japanesecompanies, as well as the differencesbetween the American and the Japa-nese companies.

CONCLUSIONThe distinctions made herein among

behavioral contingencies, macrocontin-gencies, and metacontingencies repre-sent an attempt to clarify the complexways that selection works with respect

to the behavior of individual humansand to organizations of IBCs in whichmuch human behavior is embedded.Cultural practices per se cannot evolve.The constituent members of culturalpractices do evolve, however, whetherthey are the operants of individuals orthe IBCs of organizational entities witha life of their own, above and beyondthe behavior of the particular peoplewho participate in them.

Because cultures are human con-structions, and their increasing com-plexity arises from the increasing com-plexity of the entities that participate inmetacontingenices, it seems highlylikely that humans can alter at leastsome elements of their cultures. Unlesswe understand how cultures arise andevolve, however, it will be difficult tomake wise choices regarding what canbe changed or should be changed.

REFERENCESBiglan, A. (1995). Changing cultural practices:A contextualist framework for intervention re-search. Reno, NV: Context Press.

BEHAVIOR AND CULTURE 151

Bonner, J. T (1980). The evolution of culture inanimals. Princeton, NJ: Princeton UniversityPress.

Box, H. 0. (1984). Primate behavior and socialecology. London: Chapman & Hall.

Diamond, J. (1997). Guns, germs, and steel.New York: Norton.

Friman, P. C. (2004). Up with this I shall notput: 10 reasons why I disagree with Branchand Vollmer on behavior used as a countnoun. The Behavior Analyst, 27, 99-106.

Galef, B. G., Jr. (1988). Imitation in animals:History, definition, and interpretation of datafrom the psychological laboratory. In T. R.Zentall & B. G. Galef, Jr. (Eds.), Social learn-ing: Psychological and biological perspec-tives (pp. 3-28). Hillsdale, NJ: Erlbaum.

Glenn, S. S. (1986). Metacontingencies in Wal-den Two. Behavior Analysis and Social Ac-tion, 5, 2-8.

Glenn, S. S. (1988). Contingencies and meta-contingencies: Toward a synthesis of behavioranalysis and cultural materialism. The Behav-ior Analyst, 11, 161-179.

Glenn, S. S. (2003). Operant contingencies andthe origin of cultures. In K. A. Lattal & P N.Chase (Eds.), Behavior theory and philosophy(pp. 223-242). New York: Klewer Academic/Plenum.

Glenn, S. S., & Malott, M. E. (in press). Com-plexity and evolution. Behavior and Social Is-sues.

Goldstein, M. K., & Pennypacker, H. S. (1998).From candidate to criminal: The contingenciesof corruption in elected public office. Behav-ior and Social Issues, 8, 1-8.

Hardin, G. (1968). The tragedy of the com-mons. Science, 162, 1243-1248.

Harris, M. (1989). Our kind. New York: Harper& Row.

Hull, D. L., Langman, R. E., & Glenn, S. S.(2001). A general account of selection: Biol-ogy, immunology, and behavior. Behavioraland Brain Sciences, 24, 511-573.

Johnson, S. (2001). Emergence. New York: Si-mon & Schuster.

Kawamura, S. (1959). The process of sub-cul-ture propagation among Japanese macaques.Primates, 2, 43-60.

Layng, T V. J., & Andronis, P T. (1984). To-ward a functional analysis of delusionalspeech and hallucinatory behavior. The Be-havior Analyst, 7, 139-156.

Malott, M. E. (2003). Paradox of organization-al change. Reno, NV: Context Press.

Malott, M. E., & Salas-Martinez, M. W. (2004).Strategic organizational change in higher ed-ucation. OBM Newsletter, 18(2), 9-11.

Malott, R. W. (1998). Performance manage-ment and welfare reform: The three-contin-gency model of performance management ap-plied to welfare reform. Behavior and SocialIssues, 8, 109-139.

Malott, R. W., & Suarez, E. A. T. (2004). Prin-ciples of behavior. Upper Saddle River, NJ:Pearson Education.

Mattaini, M. (1995). Teaching cultural design:Shaping new behaviorists. Behavior and So-cial Issues, 5(2), 21-28.

Mawhinney, V. T. (1995). Metabehaviors as dis-criminative stimuli for planned cultural evo-lution. Behavior and Social Issues, 5(1), 35-44.

Nevin, J. A. (1998). Contingencies of welfarereform. Behavior and Social Issues, 8, 101-108.

Palmer, D. C. (1991). A behavioral interpreta-tion of memory. In L. J. Hayes & P. N. Chase(Eds.), Dialogues on verbal behavior (pp.261-279). Reno, NV: Context Press.

Ray, R. D., Upson, J. D., & Henderson, B. J.(1977). A systems approach to behavior: III.Organismic pace and complexity in time-space fields. The Psychological Record, 27,649-682.

Schwartz, B. (1974). On going back to nature:A review of Seligman and Hager's BiologicalBoundaries of Learning. Journal of the Ex-perimental Analysis ofBehavior, 21, 183-198.

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

Skinner, B. F (1984a). Author's response: Someconsequences of selection. Behavioral andBrain Sciences, 7, 502-509.

Skinner, B. F (1984b). The evolution of behav-ior. Journal of the Experimental Analysis ofBehavior, 41, 217-221.

Skinner, B. F (1987). Why we are not acting tosave the world. In Upon further reflection (pp.1-14). Englewood Cliffs, NJ: Prentice Hall.

Todorov, J. C., Moreira, M. B., & Moreira, M.(2004). Metacontingencies: Interlocked andunrelated contingencies. In Contemporarychallenges in the behavioral approach: ABrazilian overview (pp. 221-225). Santo An-dre, Brazil: ESETec Editores Associados.

Ulman, J. D. (1998). Toward a more completescience of human behavior: Behaviorologyplus institutional economics. Behavior andSocial Issues, 8, 195-217.