Jeab, AC y Neurociencias

8/6/2019 Jeab, AC y Neurociencias

1/18

RELATIONS AMONG FUNCTIONAL SYSTEMS IN BEHAVIOR ANALYSIS

TRAVIS THOMPSON

DEPARTMENT OF PEDIATRICS UNIVERSITY OFMINNESOTA SCHOOL OF MEDICINE

This paper proposes that an organisms integrated repertoire of operant behavior has the status ofa biological system, similar to other biological systems, like the nervous, cardiovascular, or immunesystems. Evidence from a number of sources indicates that the distinctions between biological andbehavioral events is often misleading, engendering counterproductive explanatory controversy. A gooddeal of what is viewed as biological (often thought to be inaccessible or hypothetical) can becomepublicly measurable variables using currently available and developing technologies. Moreover, suchendogenous variables can serve as establishing operations, discriminative stimuli, conjoint mediatingevents, and maintaining consequences within a functional analysis of behavior and need not lead toreductionistic explanation. I suggest that explanatory misunderstandings often arise from conflatingdifferent levels of analysis and that behavior analysis can extend its reach by identifying variablesoperating within a functional analysis that also serve functions in other biological systems.

Key words: functional systems, scientific explanation, neuroscience, reductionism

_______________________________________________________________________________

Since its inception, the field of behavioranalysis has been concerned primarily with variables external to the organism that in-fluence its behavior. Endogenous factors havelargely been considered private, inaccessible,and in some cases, hypothetical (Skinner,1938), a view that persists today (e.g., Barnes-Holmes, 2005; Faux, 2002). These distinctionsare contrary to the epistemology of a functionalanalysis of behavior, which attempts to identifythe functions of variables in relation to observablebehavior, not their physical locus or ease ofaccessibility to public scrutiny. Dividing anorganisms world into behavioral and biolog-ical factors has created counterproductiveexplanatory problems, often presented asa conflict between reductionism and explana-tion based on publicly accessible external variables (Moore, 2002). The main purposeof this paper is to suggest that an organismsintegrated repertoire of operant behaviorhas the status of a biological system, similarto other systems, like the nervous, cardiovas-cular, or immune systems. This collectivesystem of functional behavioral units (see

Thompson & Zeiler, 1986) provides the majormechanism by which organisms interact with,and act upon their physical and social envir-onments. A second purpose is to presentevidence that the distinctions between biolog-ical and behavioral events are often mislead-ing, since a good deal of what is often viewedas biological (often believed to be inaccessibleor hypothetical) can be made publicly measur-able using currently available and developingtechnologies. Finally, I suggest that misunder-standings arise from conflating different levelsof analysis and their associated causal relation-ships.

The integrated repertoire of behavioralunits (operants) that have been acquired andmaintained under the functional control ofmotivational or establishing operations, dis-criminative stimuli, mediating events conjointwith reinforced responses, and consequences(mostly exogenous, some endogenous), func-tions as a biological system. Embedded withinoperants are respondents that contributediscriminative stimuli altering the internalmilieu, thereby moderating the expression ofoperant behavior. Individual operants areconstituent members of a behavioral reper-toire making up that system. In evolutionaryterms, the adaptability of these functionalbehavioral systems and the propensity to

construct more complex and higher-orderdispositional repertoires accounts for our viability and species uniqueness. These func-tional behavioral systems perform two impor-

The author is indebted to William McIlvane, RichardMeisch, and Jay Moore for their thoughtful comments andsuggestions on an earlier draft of this manuscript.

Reprints can be obtained from Travis Thompson, PhD,

Professor, Autism Program, Department of Pediatrics,MMC 486 Mayo, University of Minnesota School ofMedicine, 420 Delaware, Minneapolis, MN 55455 or (e-mail: [email protected]).

doi: 10.1901/jeab.2007.21-06

JOURNAL OF THE EXPERIMENTAL ANALYSIS OF BEHAVIOR 2007, 87, 423440 NUMBER 3 (MAY)

423


2/18

tant tasks. First, they make it possible toregulate essential physiological and biochem-ical housekeeping tasks, such as food acquisi-tion, intake and energy balance, liquid con-sumption which regulates cooling, blood

pressure and excretion of wastes via thekidney, and reproductionabilities all sharedwith other organisms. Second, our humannessis a product of our ability to acquire and uselanguage, to represent things, concepts,thoughts, or internal stimuli (e.g., emotions) with abstract symbols. As a result, we developenduring idiosyncratic functional behavioralclusters that contribute to individual unique-ness (see Lubinski & Thompson, 1986), andallow us to function collectively with other

people.The field of behavior analysis has opportu-nities to expand its reach greatly by exploring variables that interdigitate or overlap withthose of other biological systems. In so doing,one must be clear about the relative level ofanalysis that is the basis for causal attributions.Inclusion of endogenous components withina functional analysis of more complex humanbehavioral repertoires may provide insightsinto the ontogeny of larger combinations ofinterrelated operants that make us uniqueindividuals, as well as enabling us to functioncollectively as human beings.

Biological Systems

Skinner (1972) wrote, A comprehensive setof causal relations stated with the greatestpossible precision is the best contribution which we, as students of behavior, can makein the co-operative venture of giving a fullaccount of the organism as a biological sys-tem (pp. 269270). A biological system isa functionally related group of interacting,interrelated, or interdependent elementsforming a complex whole that serves anorganismic function, such as reproduction orendocrine activity. Most biological systemsconsist largely of endogenous components,such as the heart and blood vessels, or thebrain with its neurons and nerve fibers, butsuch systems also are responsive to externalinputs, such as energy (food), oxygen, tem-perature changes, visual, auditory, tactile, and

other external stimulation. Each system oper-ates according to a set of internally consistentand empirically derived principles that havebeen established, largely independently of one

another. However, most biological systemsinteract with one another; for example,the vascular system distributes hormones pro-duced by the endocrine system, and endo-crine hormones, in turn, can influence vascu-

lar function. The nervous system makes itpossible for the respiratory system to introduceoxygen into the body, and oxygenated bloodpermits brain cells to function. The coordinat-ed interactions among various biological sys-tems make it possible for an organism tofunction as an integrated whole and permitsurvival.

A Functional Behavioral System

Within the field of behavior analysis, func-

tional behavioral units have historically beentreated as if they were a closed system. Theterms, concepts, and defining operations referback to variables within the system, rather thanto components of other biological systems. Justas other biological systems are necessarilyinteractive, one might more appropriatelythink of functional behavioral units as a typeof dynamical system (Bertalanffy, 1968; White,Harrison, & Mottershead, 1992) interactingwith other such units. When a variable in one

system also functions as a variable in a secondsystem, it becomes useful to explore therelationships among systems, as suggested byMach (1914/1959) and Mayr (1982). Onesystem need not be reduced to the other, butexploration of relationships among systemswill likely enhance our understanding of eachof the individual systems.

In science we are concerned with the degreeto which phenomena or events within, say, twosystems are concordant, not whether they areidentical, since absolute isomorphism acrosslevels of analysis is rare in biology. My purposein this paper is not to argue that endogenousevents (e.g., neurochemical receptor binding)are equivalent in every respect with exogenousoperations (e.g., food deprivation), but toexamine the degree to which their behavioralproperties are concordant.

The functional units of behavior proposedin Skinners Behavior of Organisms(1938) weredefined by a complex set of relationships thatincluded four components: (1) motivational

events, (2) discriminative stimulus events, (3)responses, and (4) consequences. An operantis an abstraction referring to the relationamong these variables that is demonstrated

424 TRAVIS THOMPSON


3/18

when the probability of the response compo-nent of those relationships is changed bymanipulating aspects of the other components(i.e., establishing operations, discriminativestimuli, or consequences). Some of the vari-

ables that are components of functionalbehavioral units can be located beneath theskin, as Skinner (1945) noted many years ago.In the present paper I suggest that endoge-nous components of operant behavior can beeffectively incorporated into a functional anal-ysis, including interactions with other biolog-ical systems, to provide a more completeaccount.

Interactions among Systems

In his early writing, Skinner (1938) empha-sized procedures for identifying regularitiesamong the directly observable environmental variables and changes in the probability ofbehavior of which they were a function. Although he made reference to physiologythroughout his writing (Morris, Lazo, & Smith,2004), only in his later works did Skinnerembrace the idea that objectively measurableevents obtained at a different level of analysiscould have the status of familiar external

variables within a functional analysis. Skinnerdevoted most of his chapter, The NervousSystem and Behavior, to reasons for rejectingreductionism; however it is noteworthy that heconcluded the chapter with a quotation fromErnst Machs (1914/1959) Analysis of Sensa-tions:

It often happens that the development of twodifferent fields of science goes on side by sidefor long periods, without either of themexercising influence on the other. On occa-

sion, again, they may come into closer contact, when it is noticed that unexpected light isthrown on the doctrines of one by thedoctrines of another (as cited in Skinner,1938, p. 432).

Mach suggested that overly enthusiasticefforts to combine the two sister fieldsoften fail, and the disciplines eventually gotheir separate ways. Salient in the presentcontext was Machs final comment quoted bySkinner:

the temporary relation between them [thefields] brings about a transformation of ourconceptions, clarifying them and permittingtheir application over a wider field than that

for which they were originally formed. (as citedin Skinner, 1938, p. 432)

Near the end of his book Skinner wrote, Iam not overlooking the advance that is madein the unification of knowledge when terms atone level of analysis are defined (explained) ata lower level (Skinner, 1938, p. 428). Skinnertook pains, as have later theorists (e.g., Moore,2001; Schnaitter, 1984), to argue that endog-enous events are not independent initiators ofaction; rather, they should be understood asplaying a role within the functional analysis ofbehavior. Endogenous events appropriatelyenter into a functional analysis when theyhave concordant functional properties as withthose of light falling on cells in the retina ordelivery of water to the tongue of a water-deprived rat following lever pressing.

Theoretical reduction is a popular notion inthe biological sciences, but in practice actualreduction across levels of analysis is rare. Marr(1977) pointed out that homogenous reduction,a characteristic of all sciences, occurs whenthere is a common terminology, such asresponse, stimulus, reinforcement, and moregenerally, a common set of variables enteringinto functional relations as economical de-

scriptions of nature. This, necessarily, is theform of reduction that has occurred within thefield of behavior analysis. In contrast, inheterogeneous reduction, terms or concepts thatappear within a lower-level explanatorytheory (e.g., ion flux, long- term potentiation,neurochemical receptor) are lacking in thedomain to be explained (behavior analysis)(see, e.g., Nagel, 1961). That renders establish-ing reduction of one to the other extremelydifficult, which is true of the fields of behavior

analysis and neuroscience. Batterman (2003)argues that in such cases it is more appropriateto speak of intertheoretic relations rather thanreduction (cf. Sklar, 1967). In this borderlandbetween theories and levels of analysis thegreatest richness is often found. Instead ofseeking to derive logically valid reductionsentences, a concept used by philosophers ofscience (Carnap, 1938), my goal is to empir-ically develop bridge hypotheses that containelements from each level of analysis, thereby

permitting probabilistic assertions about therelation of one to the other. To demandidentity across levels of analysis is unrealisticand some would say logically impossible

FUNCTIONAL SYSTEMS IN BEHAVIOR ANALYSIS 425


4/18

(Feyerabend, 1962). For example, if we ask whether the properties of heat, light, andsound are functionally equivalent in everyrespect as discriminative stimuli, the answer would be no. However, the manner in

which each variable functions in relation tobehavior is highly concordant, though vari-ables differ qualitatively (e.g., visual after-images) and quantitatively (e.g., delay fromstimulus onset until detection). My purpose inthis paper is not to argue that endogenousevents (e.g., hormones binding to receptors)are equivalent in every respect with exogenousoperations (e.g., food deprivation), but toexplore the degree to which their behavioralproperties are concordant and therefore heu-

ristically and explanatorily useful. Although many biologists assume theoreti-cal reductionism is their ultimate goal, thehistorian and philosopher of biology, ErnstMayr (1982), argued that it is not reductionthat should be sought in biological explana-tion, but rather an understanding of theinteractions among biological systems. In thispaper the components making up functionalunits of behavior are advanced as one kind ofbiological system, some aspects of which are inthe external environment and others withinthe organism. Ill treat such examples as: theeffects of establishing operations and history,immediate discriminative stimulus events,characteristics of reinforced mediating eventsand responses, and their consequences. Butthe question is not, Where is a variablephysically located? The more relevant ques-tion is, What role does a variable play ina functional analysis of behavior? In succeed-ing sections of this paper, the role of endog-enous variables within a functional analysis of

behavior will be illustrated.

ENDOGENOUS COMPONENTS OFOPERANT BEHAVIOR

Motivational Operations

Many organismic states are induced bycircumstances that have impinged upon a per-son that increase or decrease the efficacy ofreinforcing events and change the probabilityof behavior maintained by those events, called

setting events or establishing operations. In TheBehavior of Organisms (1938), Skinner referredto those events as motivational or emotionaloperations. Bijou and Baer (1978) wrote, A

setting event influences an interactional se-quence (of behavior and consequences) byaltering the strengths and characteristics of theparticular stimulus and response functionsinvolved in an interaction (p. 26). As they

used the term, an experimenter or practition-er did not necessarily control a setting event.Michael (1982, 1993) provided a formal defi-nition of Establishing Operations (EOs). In2000 he elaborated on the EO concept,stating,

The two effects of an EO are an alteration inthe reinforcing effectiveness of some stimulus,object, or event (the reinforcer-establishingeffect) and an alteration in the currentfrequency of all behavior that has been

reinforced by that stimulus, object or event(the evocative effect). (Michael, 2000, p. 403)

Some transient states that change responseprobabilities and reinforcer efficacy cannot belinked to identifiable external environmentalsetting events or establishing operations. Hor-monal changes associated with menstrual cyclecan change the value of positive and negativereinforcers (e.g., Carr & Smith, 1995) but haveno identifiable environmental antecedent.Food is a highly effective reinforcer for people with Prader Willi Syndrome (PWS) who aretypically obese, of short stature, and with mildintellectual disability (Holm et al., 1993;Prader, Labhart, & Willi, 1956). Deletion ordisruption of a gene or several genes on theproximal long arm of the paternal chromo-some 15 or inactivation of that region inindividuals who have two maternal copies ofthe chromosome lead to the cluster of physicaland behavioral features making up PWS(Prader-Willi syndrome, 2006). People with

PWS have a nearly insatiable appetite, withsome affected individuals eating until theirstomachs become severely distended withpossible perforation (Schrander-Stumpel etal., 2004; Wharton, Wang, Graeme-Cook,Briggs, & Cole, 1997). They have abnormallyhigh levels of a hormone, ghrelin, in the bloodstream that produces overeating (DelParigi etal., 2002). The abnormal levels result from a 2-to 3-fold increase in ghrelin-producing cells inthe stomach and small intestine (Choe et al.,

2005). Elevated ghrelin binding to its recep-tors appears to constitute a sufficient establish-ing operation to account for the excessivereinforcing effectiveness of food. Self-reports

426 TRAVIS THOMPSON


5/18

by people with PWS of chronically feelinghungry (e.g., tacting the intereoceptive state),engaging in food-seeking and food consump-tion are the behavioral consequences of thatantecedent peptide-receptor condition.

Neurochemical establishing operations: Sleepdeprivation is associated with serotonin andcorticosteroid changes in the brain (Anderson,Martins, DAlmeida, Bignotto, & Tufit, 2005;Farooqui, Brock, & Zhou, 1996; Meerlo,Koehl, van der Borght, & Turek, 2002). Ina series of studies, Craig Kennedy and coworkers have examined the relationship be-tween sleep deprivation, brain serotonin andits receptors, and avoidance behavior. Com-pared with baseline, rats shock-avoidance

response rates increased across a range ofresponseshock intervals when sleep deprived,compared with non-sleep-deprived conditions(Kennedy, Meyer, Werts, & Cushing, 2000).Kennedy and coworkers subsequently exam-ined lever pressing maintained under a multi-ple fixed-ratio 30 fixed-interval 60-s (mult FR30 FI 60 s) schedule of appetitive reinforce-ment in rats exposed to 24, 48, or 96 hr ofsleep deprivation and control conditions. Appetitively maintained lever pressing de-

creased at 96 hr of sleep deprivation but didnot change with lesser degrees of sleepdeprivation (Kennedy, 2002). Sleep depriva-tion and serotonin depletion via p-chlorophe-nylalanine (PCPA) administration resulted inan increase in avoidance responding that wasreversed by administration of tryptophan,a serotonin precursor. In a subsequent study,rats were trained to lever press under a free-operant avoidance schedule. Rats were ex-posed to 48 hr of sleep deprivation or equiv-alent control conditions, and then adminis-tered a range of dosages of a drug thatinhibited serotonin neurons. These conditionsalso resulted in increased avoidance respond-ing. These effects were reversed by preinjec-tion of the drug that enhanced serotoninneuronal activity (Harvey et al., 2004).

Similar effects of sleep deprivation appear tooccur in people as well as laboratory animals.Kennedy and Meyer (1996) studied individuals with developmental disabilities whose daytimeproblem behavior was negatively reinforced by

escape or avoidance from instruction, espe-cially involving task demands. They examinedthe students behavior over several weeks usinganalogue functional analysis methods and

found that increased rates of negatively re-inforced problem behavior followed nightsduring which the students experienced theleast amount of sleep. This result does notprove sleep deprivation produced serotonin

depletion in these youth, which in turnlowered the threshold for aversively main-tained behavior, but the finding is consistentwith that possibility. If this analysis is correct, 5-HT1A receptor-binding, often induced bysleep deprivation, may constitute an establish-ing operation that increases the reinforcingproperties of aversive stimuli that are termi-nated by operant responding. Serotonin hasa wide array of other effects as well; however,depletion of this neurochemical appears to

mediate sensitivity to negatively reinforcedbehavior.

Endogenous Discriminative Stimuli

Neuroreceptor mediated stimuli. Experimentalevidence for powerful stimulus control byinteroceptive stimuli began emerging in the1960s, when nonhuman laboratory studiesdemonstrated convincingly that such stimulicould serve discriminative stimulus functions,much as lights or tones in a traditional operant

paradigm. In a prescient study, Schuster andBrady (1971) persuasively demonstrated in-teroceptive stimulus control over operantbehavior. They brought a lever-pressing oper-ant of rhesus monkeys under the stimuluscontrol of the intravenous infusion of epi-nephrine. Responses produced food reinforce-ment under a fixed-ratio schedule followingepinephrine infusion; but following a salineinfusion, lever pressing produced no pro-grammed consequences. The interoceptivestimulus events presumably related to theeffects of epinephrine binding at alpha andbeta adrenergic receptors (e.g., increasedheart rate and blood pressure) set the occasionfor lever pressing reinforced by food.

In later studies other researchers demon-strated that a wide variety of drugs producinginteroceptive states could reliably serve dis-criminative stimulus functions (Thompson &Pickens, 1971). Many drugs used to treatmental health problems (as well as those thatare addictive) bind to the same chemical

receptors in the brain as do naturally occur-ring neurotransmitters and can also demon-strate discriminative stimulus properties (seeSociety for stimulus properties of drugs, 2006).



6/18

Laboratory studies indicate that animals canreliably respond discriminatively to effects ofdrugs that mimic normal brain chemicaltransmitter function. Not only do animalsrespond discriminatively to consequences of

a drugs binding to one brain chemicalreceptor type versus another (e.g., dopamine versus GABA), they also can distinguishbetween effects of dosages of the same drugand corresponding internal stimulus states.

The ability to experimentally manipulateinteroceptive stimuli pharmacologically thatcan mimic those produced by naturally occur-ring neurochemical events lends credence tothe notion that endogenous brain chemicalchanges can also serve as discriminative

stimuli. Emotional states, such as anxiety anddepression, have discriminative properties thatfor most people lead us to emit such tacts asanxious, or angry or sad as the namefor the respective internal states. Those in-ternal events not only have discriminativeproperties, but they also can modulate theevents that could serve as reinforcers. A person with depression may respond to an invitationfrom a friend to attend a movie by saying, Nothanks, I dont feel like going to a movietonight. The person with depression isperforming a complex conditional discrimina-tion. The friends verbal invitation is theproximal discriminative stimulus event, butresponding to that stimulus is conditionalupon the unique interoceptive discriminativestimuli associated with the depressive neuro-chemical state. When she is experiencing thisparticular internal state, and a friend invitesher to attend a movie, the verbal response thatcomes to highest strength is to decline theinvitation. The depressed individuals utter-

ance is also based on the discriminativeproperties of the diminished state of strengthof her own behavior.

Conjoint Mediating Events

As Skinner noted in Science and HumanBehavior, The line between public and privateis not fixed. The boundary shifts with everydiscovery of a technique making private eventspublic (Skinner, 1953, p. 282). Some of thoseevents are conjoint with observable external

behavioral changes. The durability of changesin external response probabilities may dependon mediating events that are strengthenedconjointly with reinforced responding.

Operant responding that leads to a reinfor-cing event can be called an effector event oractivity. Neurochemical and microstructuralbrain changes associated with reinforced re-sponding cannot properly be called effector

activities because they are not muscular orglandular activities. Instead they are calledconjoint mediating events that occur in conjunc-tion with reinforced effector events. Conjointmediating events occur within the brain sub-sequent to establishing operations and dis-criminative stimuli and before reinforcedresponding. Changes in brain microstructure(e.g., synapse formation) following reinforcedresponding can be increased via reinforce-ment and diminished by extinction as illus-

trated in the following example. Rats trainedto perform a reaching task followed by positivereinforcement not only displayed increasedpercent correct responding of the motorperformance but also developed significantlymore synapses per neuron than controlswithin layer V of the caudal forelimb area. Inthe absence of reinforcement following exe-cution of the reaching response, the numberof synapses per cell declined (Kleim et al.,2002). The rats reaching responses are effec-

tor events and the synaptic changes constituteconjointly strengthened mediating events.Having increased the number of synapses instructures involved in emitting a given re-sponse increases the probability of recurrenceof that response, which thereby lays thefoundation for later learning and fluency.Reinforcement not only increased the proba-bility of recurrence of members of a specificoperant response class, but also had retroac-tive effects on the neurochemical and/ormicrostructural neural events that precededoccurrence of those responses.

Constraint induced therapy. Shaping newresponses assumes the organism has an intactnervous system sufficient to enable respond-ing. In 1980 Edward Taub conducted ground-breaking research that he believed would leadto a new treatment for people who wereparalyzed by a stroke or other nervous systemdamage. The experimental work involvedsevering a nerve to one arm of a rhesusmonkey, and then restraining the animals

other (undamaged) arm during recovery. Hehypothesized that by differentially reinforcinguse of the dysfunctional limb by successiveapproximations, eventually some or most of

428 TRAVIS THOMPSON


7/18

the use of the dysfunctional limb wouldreturn. That could only be accomplished bylimiting the animals use of its functional limb.Once sensory input ceased, it was commonlythought that the parts of the cortex that

normally received that nerves input wouldregress. Around the same time, Kaas, Merze-nich, and Killackey (1983) conducted researchand reviewed evidence that the brain of adultmonkeys undergoes reorganization followingloss of sensory input from one limb, such thatother brain areas take over the functions ofthose structures that were no longer con-nected to the dysfunctional limb. Their work was consistent with Taubs hypothesis. Taubsubsequently demonstrated that through re-

peated reinforcement of use of the dysfunc-tional limb, partial recovery was possible.This work led Taub et al. (1993) to develop

Constraint Induced Movement Therapy, in whichpatients with stroke, cerebral palsy, and otherneuromotor disorders substantially recoveredif they arduously practiced using their dys-functional limb following injury. He demon-strated that some of the long term, evenpermanent, motor impairment following ner-vous system damage is due to a combination of

motivational and learning factors resulting ina learned non-use rather than being theresult of the CNS damage per se. A functionalanalysis of his rehabilitative work with monkeysand later with human patients was difficult torationalize unless one assumed it was possiblethat repeated use of the dysfunctional limb would produce significant changes in thebrainand these are now known to occur(see, e.g., Pons et al., 1991).

Intensive Early Behavior Therapy (IEBT) forchildren with autism is the only knownpsychological or behavioral treatment thatproduces sustained changes in a severe behav-ior disorder following treatment. Lovaass(1987) demonstration that approximately halfof the children with autism diagnoses developsocial and language skills that are sustainedfollowing treatment, and permits them tofunction intellectually in the average range, iscompelling testimony to the power of behavioranalytic principles and procedures. However,half of the children in Lovaass and later

studies fail to show such improvement. Re-cently Sallows and Graupner (2005) foundthat children with lower baseline IQs, lack ofjoint attention, and little or no motor or verbal

imitation (at 2 to 3 years of age) and withautism diagnoses showed limited improvementin core autism symptoms over 4 years of 30+ hrper week of IEBT. That raises the question ofwhy IEBT is effective with some children with

autism but not with others. Whatever isdifferent about the brain functioning ofchildren with autism must be consistent withthe IEBT evidence, namely that half of thechildren dramatically improve and half showlimited gains using what appear to be the samebehavior therapy techniques.

Bauman and Kemper (1994) found thatbrain tissue samples from young children withautism had an abnormally large number oftightly packed immature neurons in limbic

and prefrontal areas as well as cerebellarvermis, with few synapses. By adolescence or young adulthood many neurons had disap-peared, but those that remained had richdendritic arborization and were synapticallyconnected. This suggests that some of theneurons in those structures had the potentialto form synapses, but presumably because theyhad inadequate input during development,neuronal pruning led to their regression.

I recently suggested that children with autism

diagnoses who were responsive to IEBT had thepotential to form synapses in brain areas knownto be dysfunctional in autism, whereas thosewho profited far less from IEBT had abnormal-ities in the same structures, but were notamenable to correction by forming new synap-ses (Thompson, 2005). IEBT involves repetitivetraining that can be accomplished only byenlisting the active engagement of thosespecific brain structures. For example, a childcannot anticipate the events that are likely tooccur during a sequential social interaction without the active engagement of cells in theorbitofrontal cortex. By repeatedly requiringthe child to perform such discriminative tasks,new synapses will very likely be formed in thatbrain area. The more those skills are practiced,the more functionally consolidated or effectivethose synapses become, which is very likely whythe effects are permanent. In the absence ofpractice, as was common in the past when young children with autism were placed insequestered residential or ineffective school

settings, synapses failed to form, and theneurons that would normally have served thatfunction regressed through disuse. Thosestructures became permanently nonfunctional.



8/18

Much research shows that synapse formationcan be activity-dependent (Cohen-Cory, 2002;Greenough & Churchill, 2003; Jontes & Smith,2000). Errors in brain development in somehuman disabilities are associated with activity-

dependent retraction of dendritic spines,which may be insufficient (Boulanger & Shatz,2004). Dendritic spines are the locus oncertain brain neurons where synapses areformed. Nonfunctional dendritic spines areexcessively long, and fail to convey synapticcurrents from the head of the spine to thedendrite (and cell body) to which the extra-long spine neck is attached. In normal braintissue, dendritic spines change continuouslythrough typical use and disuse. The spines

contract with mechanisms similar to muscle:actin and myosin. Early disuse could reduceexpression of the molecules needed for spinecontraction, which is required to make thespine functional. By identifying the genes thatproduce the proper amount of substraterequired for dendritic spine contractions thatare only turned on by performing specifictasks, one has a more complete account of howsynapse formation does or does not occur, andof the role of functional behavioral units in

that process.To the degree that functional behavioranalysis procedures promote discriminativeresponding that can only be accomplished bycausing release of neurotransmitters in cells instructures that are synaptically deficient, onesystem, the four-component operant (estab-lishing operations, discriminative stimuli, conjoint mediating events, and maintaining con-sequences), interacts with another functionalsystem (synaptogenesis). The promotive effectof training does not imply that the four-component operant is reduced to synaptogen-esis. Rather, it suggests that a conjoint medi-ating event (synapse formation and consolida-tion) becomes a component of the four-element operant. The point of contact be-tween the two systems is the activity-dependentrelease of neurotransmitters that promotessynapse formation.

One need not refer to correlated brainchanges (synaptic-reinforced mediating events)in order to conduct IEBT with young children

with autism. However, failing to do so leaves theformer question unanswered: Why do somechildren greatly benefit from IEBT and othersdo not? Those who do not benefit equally

from IEBT may have damage to some of thesame brain structures, but the damage mayhave been produced by a different mechanism,one that is not amenable to amelioration byforming and consolidating new synapses. Axons

could have been misrouted to the wrongstructures, or receptors necessary for formingsynapses could be damaged, as happens due totoxin exposure or some genetic defects (e.g., inFetal Alcohol Syndrome, GABA receptor genedeletion in Prader Willi Syndrome). Thispossibility does not imply that children whoare minimally responsive to IEBT should re-ceive no intervention. However, it suggests that we must understand better the mechanismsunderlying their disability and determine which

intervention strategies are most likely to ame-liorate those limitations.

Maintaining Events

Drug dependence. Throughout the first halfof the 20th century, the medical communityassumed that drug addiction was a purelyphysiological or biochemical process. Pharma-cologists and other medical researchers com-monly believed that addiction could be char-

acterized in terms of the transaction betweena drug molecule and the tissue upon which itimpinged. In the early 1960s that conceptual-ization changed when experiments in labora-tory animals indicated that opiate drugs couldserve as reinforcers for operant behaviorverymuch like more familiar reinforcers such asfood or water (Thompson & Schuster, 1964; Weeks, 1962). Reinforcement schedules, stim-ulus control procedures, and other typicalcontingency manipulations affected drug-maintained behavior in ways that were largelyindistinguishable from behavior maintained byother consequences (Pickens & Thompson,1968; Schuster & Thompson, 1969). Theseand other subsequent studies suggested thatunderstanding addiction required explicationof the relationships among a drugs biochem-ical and physiological effects and their in-teraction with reinforcement contingencies.Performances in these animal models arehighly predictive of which newly developeddrugs are mostly likely to have addictive

properties in people (Ator & Griffiths, 2003;Higgins, Heil, & Lussier, 2004; Meisch 2001; Vanderschuren & Everitt, 2005). As a conse-quence, operant drug self-administration pro-

430 TRAVIS THOMPSON


9/18

cedures are now required by the U.S. Foodand Drug Administration and are used world wide to screen for the abuse-liability of newlydeveloped drugs.

Over the intervening years numerous stud-

ies have demonstrated that when most drugsare self-administered by laboratory animals,those agents bind to specific brain receptorsthat mediate the reinforcing effect of the drug.If those receptors are blocked by an antago-nist, the drug no longer serves as a reinforcer,and previously drug-maintained operant be-havior extinguishes (Yokel, 1987). But addic-tion does not reside solely in drug-receptorbinding. Whether drug administration servesas a maintaining event also depends on the

nature of the contingency relationship be-tween responses and their consequences.Spealman (1979) trained squirrel monkeys toself-administer cocaine by pressing a leverunder a variable-interval (VI) 3-min scheduleof reinforcement. Concurrently, the monkeyscould terminate the stimulus indicating theavailability of self-administered cocaine bypressing a second lever where under a FI 3-min schedule, a 1-min timeout was presented.The monkeys reliably terminated the oppor-

tunity to self-administer cocaine under thefixed-interval schedule and continued to self-administer cocaine under a variable-intervalschedule at other times. The maintenance ofbehavior by schedule-controlled drug injec-tions and by termination of access to druginjections indicated that whether and in whatway a drug would control behavior dependedon the contingency relationships, even withthe same highly addictive drug.

Self-injury in developmental disabilities. Self-injurious behavior by people with develop-mental disabilities shares features in commonwith behavior maintained by drug self-admin-istration. People who engage in self-injury doso for various reasons. Functional analysis andassessment research has demonstrated thatself-injury most often serves a demand-avoid-ance function or leads to access to caregiverattention (Khang, Iwata, & Lewin, 2002). Forsome people with severe and unremitting self-injury, there is evidence that such behavior ismaintained at least in part by the neurochem-

ical consequences of self-inflicted pain (Sand-man, Hetrick, Taylor, & Chicz-DeMet, 1997),much as drug-maintained behavior is rein-forced through the mediating brain chemical

actions of the drug that has been self-admin-istered. Among the consequences of self-inflicted painful stimulation (e.g., followingself-injury) is release of beta endorphin, ananalgesic peptide that binds to the brains

opiate receptors, the same receptors to whichmorphine and heroin bind. Cataldo andHarris (1982) hypothesized that some self-injury may be maintained by endogenousopioids released following self injury. Thomp-son, Hackenberg, Cerutti, Baker, and Axtell(1994) and Sandman et al. demonstrated that when people with chronic self-injury weretreated with the opiate antagonist drug nal-trexone, between 30 and 50% of individualsstopped or markedly reduced self-injury. Sand-

man and colleagues obtained blood samples within 5 min following an episode of self-injury and determined the blood levels of betaendorphin relative to baselines when no self-injury occurred. They then treated the sameindividuals with naltrexone and found a 0.67correlation between the amount of betaendorphin increase following self-injury andthe amount of reduction in self-injury pro-duced by naltrexone. It appears that for suchindividuals, a portion of the maintaining

consequence for self-injury is binding of betaendorphin to the brains opiate receptors,which can be blocked by naltrexone, much ascocaine self administration can be reduced byadministering specific dopamine receptor an-tagonists.

Self-injury is usually multiply determined. Intreating a 14-year-old boy with significant self-injury, Symons, Fox, and Thompson (1998)conducted an analysis of the effects of naltrex-one alone and combined with behavioralinterventions. The latter included augmenta-tive communication training for requestingescape from nonpreferred tasks, augmentativecommunication training for requesting teach-er attention, and augmentative communica-tion training alone. Using an ABCBC design,they demonstrated that while naltrexone alone(Condition B) reduced self-injury by half,adding augmentative communication intervention (Condition C) reduced self-injury by95100% in the classroom during the Cintervention phase. This result suggests that

the youths self-injury was maintained bya combination of positive and negative socialreinforcement and release and binding of betaendorphin to his brains opiate receptors.



10/18

DISCUSSION

Causal Analysis

Mach argued that scientific explanation iscondensed indirect description, by which

he meant causal explanation involved not onlydescribing what was observed but providing anaccount of the observed events in terms ofmore general principles derived from otherobservations (see discussion by Marr, 2003).The term cause as used in this article refers toan event, process, or microstructural changeshown to be at least probabilistically related toa given effect. This meaning of cause is in thetradition of Reichenbachs (1956) and Hitch-cocks (1993) probabilistic causality theories

and implies a temporal directional arrow ofcausality. Dretske (2004) proposed distinguish-ing triggering from structuring causes. Accord-ing to this formulation, structuring causes areevents that, in turn, cause a trigger to produceits effect. Structuring causes are higher-orderdispositions (see below) that make triggeringcauses possible. Establishing operations andreinforcing events are structuring causes thatmake reinforcement possible and increase theprobability of recurrence of operants. Discrim-inative stimuli are triggering causes in thisformulation.

Food deprivation is an establishing opera-tion that increases the reinforcing value offood. General causal claims, such as fooddeprivation causes food to serve as a reinforc-er, refer to repeatable general properties ofcauses and effects. Singular causal statementsrefer to specific events that have spatiotempo-ral locations, such as food deprivation causedfood to serve as a reinforcer for lever pressingby Rat 287 in operant chamber four. There isalso a difference between thecause and acause,a distinction that originated with Mill (1843/1986). Food deprivation has widespread phys-iological and biochemical effects, includingchanges in blood sugar, glucagons, amylin,ghrelin, cholysystokinin, and neuropeptide Y(e.g., Beck, 2006; Moran, 2006). It is accurateto say that food deprivation is a cause of thereinforcing value of food, but if it could beexperimentally shown that effects of fooddeprivation on the reinforcing value of food

were uniquely mediated by one or somecombination of these biochemical changes,then it would be more parsimonious to say thatthe magnitude of the reinforcing properties of

food are regulated (i.e., caused) by thespecificbiochemical events that had been identified.In the present article I have employeda probabilistic analysis of causes, withoutmaking claims that they are the cause of

specific effects. I have been primarily con-cerned about the degree to which endogenousevents demonstrate probabilistic causes con-cordant with those of established environmen-tal variables within a functional analysis ofbehavior.

Dispositional Analysis

Ryle (1949) contended that mental statescould be analyzed as dispositions, that is, a pro-pensity or predisposition to behave in a partic-

ular way under specified circumstances (Cross,2005). He warned against what he calledcategory mistakes, that is, ascribing a prop-erty to a state that it could not possibly have.Ryle meant that a mental (i.e., a non-physicalevent) could not cause a behavioral event (aphysical event), and language and conceptsapplicable to one are not appropriate for theother. We say I thought very hard about it orIm drawing a blank, metaphorically apply-ing terms from the physical world to mental

activities. Ryle argued that applying physicalis-tic language to phenomenological events leadsto illusory consequences, a problem thatplagues much of psychology (Holth, 2001).

Although Ryle (1949) suggested that oneneed not look for causal or mechanisticexplanations of the dispositions, later writershave largely disagreed (Mellor, 1974; Mum-ford, 1998). The causal basis of an objectsdisposition is something like a microstructuralproperty of the object that is causally re-sponsible, under specifiable conditions, forthe manifestation of the disposition (Prior,Pargetter, & Jackson, 1982, p. 251). Accordingto this reasoning, a statement such as, A glass vase breaks when struck by a hard objectbecause of its irregular silicon dioxide molecularstructure, would provide a more adequatedispositional account than A glass breaksbecause it is fragile. The latter accountappears tautological, though many wouldcontend it is not because it can be reducedto further dispositional properties or mecha-

nistic explanans (e.g., Mellor, 1974; Mumford,1998; Prior et al., 1982). A tautologicaldispositional explanation becomes scientifical-ly meaningful when the explanatory disposi-

432 TRAVIS THOMPSON


11/18

tion can be linked to other lawful relationshipsof a more general sort. Vanderbeeken and Weber (2002) argued that dispositional ex-planations can be useful in behavior analysis. As with the glass vase example, it would be

unacceptably circular to assert that a rats leverpressing increased in frequency because it wasreinforced, unless the concept of reinforce-ment can be tied to a broader set of general-izations and abstract concepts (e.g., scheduleeffects, matching, contrast effects, reinforce-ments role in the abstract construct oper-ant) (see Meehl, 1950).

A lay person might ask, concerning a personshe is observing, Why is that man reportingthat he sees and hears people talking who no

one else sees, and appears to be talking topeople who are not apparent to others? Aclinical psychologist might reply, Because hehas schizophrenia. The dispositional causalstatement that people who have schizophreniahave the disposition to hallucinate is notsimply reification, because it refers to one ofa complex set of interrelated dispositions thatcomprise the properties of the conditioncalled schizophrenia.

A more satisfactory dispositional account

calls for a description of conditions that giverise to the disposition to hallucinate andengage in psychotic verbalizations. Differentaccounts may be provided, such as from datasuggesting hallucinatory behavior is underimmediate interoceptive stimulus control(Leibman & Salzinger, 1998; Salzinger, 1973)and that it is possible to reduce psychotic verbalizations by extinction and differentialreinforcement of appropriate speech (Ayllon& Haughton, 1963; Leiberman, Teigen, Pat-terson, & Baker, 1973) or, alternatively, thathallucinations arise from hyperactive dopami-nergic signal transduction (Sedval & Farde,1995). The former refers to a structuring causeof the disposition whereas the latter refers totriggering neurochemical causes responsiblefor the current disposition to hallucinate. Within the field of behavior analysis, theformer types of dispositional accounts havepredominated, though at times there may beadvantages of considering endogenous causesas well.

Once we know that a system has a certaindisposition, we know how this system willbehave in certain situations due to thepresence of an underlying set of properties

that functions as a set of internal causal factors(i.e., an organisms state resulting from a re-inforcement history or a neurochemical con-dition). Dispositions do not tell us whichcausal basis is to be found in a system, they

tell us that there is some sort of causal basisthat, together with certain triggering situa-tions, will bring about typical behavior (Van-derbeeken & Weber, 2002).

Dispositional Misattribution

As Meehl (1993/2006) pointed out, explan-atory problems can arise when we propose anaccount at a given dispositional level that maynot apply to that level, but is appropriate toanother dispositional level. A common exam-

ple is a statement such as genes causebehavior. Genes cannot cause behavior.Genes produce proteins. Proteins are compo-nents of higher-level dispositions that mayeventually influence, but not cause behavior.Genes may be necessary prerequisites forparticular behavioral events to occur, but theirexpression depends stochastically on experi-ence and other epigenetic factors.

At times similar mistakes are made inattributing dispositional causes of behavior to

external environmental events. Staff members working in group homes or classrooms mayask, Why do our residents (or students) withintellectual disabilities engage in seeminglyunpredictable behavioral outbursts (e.g., ag-gressive, disruptive, or self-injurious behav-ior)? Therapists and teachers collect dataindicating that behavior outbursts seldom ifever occur on some days, but do occur withhigh frequency on others. They then conducta functional analysis of putative proximalcontrolling variables and find that on days with frequent outbursts, behavior problemsserve as task-avoidance responses terminatingdemands (Kennedy et al., 2000). On days withfew outbursts, similar demands fail to produceoutbursts. If we then further ask why the ratesof such avoidance behavior fluctuate from dayto day in apparently unpredictable ways, wefind that the threshold for demand avoidancecovaries with the number of hours per sleepthe previous night. Evidence from laboratoryanimal studies suggest sleep deprivation may

lead to depletion of brain serotonin (Harvey etal., 2004), which in turn lowers the avoidancethreshold. The assumption that a functionalanalysis of proximal environmental controlling



12/18

variables provides an adequate account over-looks the possibility that the origins of thedisposition in question (propensity to engagein avoidance behavior) are to be found ata different dispositional level, neurochemistry.

The initial point of intervention might involverectifying the sleep abnormality by administer-ing a serotonergic medication like trazadone.Once that has been accomplished, a function-al-analysis-based social intervention will morelikely be effective on the remaining occasionson which behavioral outbursts occur.

Figure 1 presents a hypothetical causalchain of events leading to signs and symptomsof autism that illustrates levels of analysis andorders of dispositions. In thinking about

causes (i.e., what causes what) one should beclear about the level of analysis or order ofdisposition being discussed. Obviously, fromthis hypothetical analysis, there are potentiallyseveral levels at which intervention mightoccur to prevent or ameliorate the develop-ment of autism symptoms, from gene therapyto pharmacological interventions targeted atdendritic spine activity to Early IntensiveBehavior Therapy (IEBT). Actions taken ata given dispositional level cannot logically

change higher-order dispositions (i.e., IEBTcannot cause new genes to be produced), butcan affect lower-order dispositions (e.g., selec-tive synapse formation).

Much as the statement genes cause behav-ior involves a dispositional misattribution,behavior analysts commit similar mistakes when they assert that autism is learnedthrough differential reinforcement of autisticbehavior (Drash & Tudor, 2004). Autism is nota set of behaviors to be learned. Specificoperants such as hand flapping and tantrumscould likely be acquired through differentialreinforcement. However, autism is a complexstate, expressed in an array of specific featurespresent at various levels. These include objec-tively measurable dysfunction in five or sixbrain areas, serotonin abnormalities, frequenterrors on human Chromosomes 2, 7, 13, or 15,lack of eye contact apparent shortly after birth,lack of social initiations from infancy (e.g.,gestural, tactile), inability to speak so as tocommunicate effectively, excessive and quali-

tatively unusual stereotyped behavior withinthe first year of life, intolerance for changes inexternal stimulus conditions leading to tan-trumssome or all of these characteristics

detectable by 12 to 18 months of age. X(differential reinforcement of hand flapping)could not cause Y (lack of amygdala activationwhen looking at faces). Differential reinforce-ment of hand flapping cannot logically cause

the brain abnormalities or the covaryingcluster of behavioral features seen in autism.

Thinking more clearly about elements insuch a causal sequence is necessary to preventour making unfortunate category mistakes,and can enlighten us about the relativeeffectiveness of different foci as well astechniques of intervention. It is illogical tosuggest that administering an atypical antipsy-chotic drug (Roeder, 1995) or repeatedlystroking the childs skin with a brush as part

of Sensory Integration Therapy (Kranowitz,2005) would replace or overcome the con-sequences of the missing synapses or substitutefor appropriate and necessary learning experi-ences. However, given a condition of limitedsynaptogenesis, if we have reason to believethat intensive early behavior therapy couldpromote synapse formation as indicated inFigure 1, then such would seem a rationalstrategy.

Superordinate Dispositions

The field of behavior analysis has provideddetailed accounts of lower-level behavioraldispositions, and to some degree, higher-orderdispositions, though with limited specificity.For example, we have some understanding ofthe establishing operations, stimulus control,response characteristics, and consequencesthat might lead a child to raise her hand inthe classroom. But we would be hard pressedto prescribe the necessary and sufficientconditions to create a needy or a resilientchild, terms that could be translated intoobjective behavioral descriptions. We may havereasonable hunches, but these hunches wouldmerely be plausible accounts, not the types offunctional analyses that have been providedfor first- and second-order dispositions.Higher-level dispositions are likely to beacquired through relational learning processes(e.g., Sidman, 1994). Understanding the con-ditions giving rise to superordinate categoricaldispositional clusters has practical as well as

theoretical implications. McIlvane and Dube(2003) have suggested that generalized stimu-lus control by more complex and formallydissimilar stimuli depends on the agreement

434 TRAVIS THOMPSON


13/18

Fig. 1. Hypothetical causal chain leading to signs and symptoms of autism, illustrating orders of dispositions. Autism

symptoms and features constitute a first-order disposition. Defective dendritic spines reducing synaptogenesis impair theability to learn typical language and social skills, a second-order disposition. Lack of substrate for activity-dependentdendritic spine contraction is a third-order disposition. Lack of protein expressed in specific brain structures is a fourth-order disposition. Absence of a gene to produce those proteins is a fifth-order disposition. Early language and sociallearning experience is an event that impacts synapse formation, a second-order disposition.



14/18

between putatively relevant aspects of thosestimuli, as perceived by the person impactedby those stimuli. They refer to this relationshipas stimuluscontrol topography coherence. McIlvane(2003) has suggested that providing an ade-

quate account of these processes may requirenew behavioral principles. Hayes and co-work-ers (Hayes & Barnes-Holmes, 2001) haveproposed a related approach (RelationalFrame Theory) in an attempt to account for ways in which apparently dissimilar stimulibecome linked into elaborate verbal responseclasses.

Information making it possible for psy-chotherapists to dissect a clients superordi-nate relational stimulusresponse classes may

be instrumental in resolving personal prob-lems. A client who has repeatedly lost jobsbecause of confrontations with supervisorsmay be aided by strategies for devolvinga stimulus class such as authority figures,with their presumed properties, and establish-ing more relevant stimulusresponse classes.Attempting to change racial and ethnic stereo-types by differentially reinforcing young chil-dren for saying positive things about theirpeers from different backgrounds may have

little impact; while perhaps laudable, sucha procedure may not be directed at theappropriate dispositional level. Resolving ra-cial and ethnic discriminatory behavior maydepend on better understanding how suchsuperordinate classes (e.g., Black, Caucasian,and Latino) are established, and how they maybe changed to incorporate more sociallyrelevant personal attributes.

CONCLUSION

I have suggested in this paper that theconcatenated repertoire of operant behavioralunits (with embedded respondents) comprisesa functional biological system comparable toother biological systems. These behavioraldispositions are hierarchically organized be-ginning with individual response classes thatare combined to create increasingly complexrepertoires that distinguish human symbolicbehavior. There is nothing inherent in a func-tional analysis of behavior that requires all of

the variables to be located external to the skin.Indeed, as is true of other biological systems,there may be advantages of studying interac-tions between and among functional behav-

ioral repertoires and other systems (e.g.,nervous, endocrine, and immune systems).The four basic components of operantsmotivational operations, antecedent stimuli,responses and their associated mediating

events, and reinforcing consequencesmayinclude endogenous conditions as well asevents occurring outside the skin. In someinstances substantial overlap between behav-ioral processes and endogenous physiologicaland biochemical events may occur, but a com-plete reduction of one to the other may likelyprove difficult, if not impossible. However,providing more information regarding theseries of events occurring within the nervoussystem between presentation of a discrimina-

tive stimulus and emission of a response canplay an important role in completing theaccount (Skinner, 1989).

The proposed analysis addresses severalproblems. For behavior analysts, the proposedapproach makes it largely unnecessary to ask whether the cause of a specific instance orclass of behavior is environmental or biologi-cal. This distinction is not helpful; rather theapproach presented provides a systematic wayof formulating experimental questions regard-

ing the role of endogenous variables withina functional analysis. For example, what is thebehavioral mechanism by which an antidepres-sant medication changes the behavior andpsychological functioning of a person witha major depressive disorder? Does it reducethe negatively reinforcing value of socialinteractions as well as enhance the value ofputative positive reinforcers, or does it changethe interoceptive discriminative stimuli inher-ent in a depressive state? The proposedstrategy assists us in thinking about the properlevel at which to intervene to produce a giveneffect, or how to best measure a treatmenteffect, or, more generally, allows us to examinea problem of interest with greater care.

For neuroscientists and neurogeneticists,the proposed analysis assists in thinking moreclearly about the causal role of genes vis a vizbehavior, as well as neuropathological condi-tions expressed as behavioral or developmen-tal disorders. The approach may, for example,facilitate identification of endophenotypes. Got-

tesman and Shields (1967) and Gottesmanand Gould (2003) argued that identifyingspecific physical features (often physiologicalor neurochemical) that uniquely covary with

436 TRAVIS THOMPSON


15/18

a clinical syndrome is more likely to beproductive in unraveling genepsychopathol-ogy relationships than focusing on clinicalsymptoms (e.g., thought disorder) that areoften influenced by experience. The use of

behavior-analytic strategies could extend therange of highly reproducible phenotypic fea-tures that could be included in endopheno-typic research strategies. For example, thedegree to which discriminative responding tofacial expression could be established in very young children could be a predictor variablein heritability studies of autism spectrumdisorders.

The biologist Waddington (1942) coined theterm epigenetics, referring to modifiable herita-

ble changes in gene function that occur withouta change in DNA sequence (genotype). Thesechanges may be induced spontaneously, inresponse to environmental or experientialfactors, or to the presence of a particular allele.The degree to which modifiable epigeneticchanges occur depends on the underlyingmechanism. The approach suggested herecould provide neuroscientists with more sophis-ticated approaches to exploring activity-depen-dent neural processes. Behavior-analytic labo-

ratory procedures may make it possible toexplore specific experience-dependent epige-netic influences that may alter gene expressionin specific brain areas. Early differential re-inforcement of discriminative responding tovisual images of hand and arm movements maypromote gene expression in the mirror neuronsystem, for example.

Finally, clear thinking about dispositionscausally linked to individual performancesand larger behavioral classes may help avoid

errors of causal attribution: what Ryle calledcategory mistakes and what Meehl described aserrors of dispositional levels. Avoiding suchmistakes can also affirmatively guide moreproductive analyses and intervention strate-gies.

REFERENCES

Anderson, M. L., Martins, P. J., DAlmeida, V., Bignotto,M., & Tufik, S. (2005). Endocrinological and cate-cholaminergic alterations during sleep deprivation

and recovery in male rats. Journal of Sleep Research, 14,8390.

Ator, N. A., & Griffiths, R. R. (2003). Drug abuse liabilityassessment in laboratory animals. Drug and AlcoholDependence, 70 (Suppl. 1), 5572.

Ayllon, T., & Haughton, E. (1963). Modification ofsymptomatic verbal behavior of mental patients.Behaviour Research and Therapy, 2, 8797.

Barnes-Holmes, D. (2005). For the radical behavioristbiological events are not biological and public eventsare not public. Behavior and Philosophy, 31, 145150.

Batterman, R. (2003). Intertheory relations in physics. Re-trieved December 17, 2006, from Stanford University,Encyclopedia of Philosophy Web site: http://plato.stanford.edu/entries/physics-interrelate/

Bauman, M. L., & Kemper, T. L. (1994). Neuroanatomicobservations of the brain in autism. In M. L. Bauman,& T. L. Kemper (Eds.), The neurobiology of autism(pp.6885). Baltimore: Johns Hopkins University Press.

Beck, B. (2006). Neuropeptide Y in normal eating and ingenetic and dietary-induced obesity. PhilosophicalTransactions of the Royal Society of London. Series B,Biological Sciences, 361, 11591185.

Bertalanffy, L. von (1968). General systems theory: Founda-tions, development, applications. New York: George

Braziller.Bijou, S. W., & Baer, D. M. (1978). Behavior analysis of child

development. Engelwood Cliffs, NJ: Prentice-Hall.Boulanger, L. M., & Shatz, C. J. (2004). Immune signaling

in neural development, synaptic plasticity and disease.Nature Reviews Neuroscience, 5, 521531.

Carnap, R. (1938). Logical foundations of the unity ofscience. In International encyclopaedia of unified science(Vol. 1, pp. 4262). Chicago: University of ChicagoPress.

Carr, E. G., & Smith, C. E. (1995). Biological setting eventsfor self-injury. Mental Retardation and DevelopmentalDisabilities Reviews, 1, 9498.

Cataldo, M. F., & Harris, J. (1982). The biological basis ofself-injury in the mentally retarded. Analysis andIntervention in Developmental Disabilities, 7, 2139.

Choe, Y. H., Songs, S. Y., Paik, K. H., Oh, Y. J., Chu, S. H.,Yeo, S. H., Kwon, E. K., Kim, E. M., Rha, M. Y., & Jin,D. K. (2005). Increased density of ghrelin-expressingcells in the gastric fundus and body in Prader-Willisyndrome. Journal of Clinical Endocrinology and Metab-olism, 90, 54415445.

Cohen-Cory, S. (2002, October 25). Developing synapse:Construction and modulation of synaptic structuresand circuits. Science, 298, 770776.

Cross, T. (2005). What is a disposition? Synthese, 144,21341.

DelParigi, A., Tschopo, M., Heiman, M. L., Salbe, A. D.,Vozarova, B., Sell, S. M., Bunt, J. C., & Tataranni, P. A.(2002). High circulating ghrelin: A potential cause forhyperphagia and obesity in Prader-Willi syndrome. Journal of Clinical Endocrinology and Metabolism, 87,54615464.

Drash, P. W., & Tudor, R. M. (2004). An analysis of autismas contingency-shaped disorder of verbal behavior.The Analysis of Verbal Behavior, 20, 523.

Dretske, F. (2004). Psychological vs. biological explana-tions of behavior. Behavior and Philosophy, 32, 167177.

Farooqui, S. M., Brock, J. W., & Zhou, J. (1996). Changesin monoamines and their metabolite concentrations

in REM sleep-deprived rat forebrain nuclei. Pharma-cology Biochemistry and Behavior, 54, 8591.

Faux, S. F. (2002). Cognitive neuroscience from a behav-ioral perspective: A critique of chasing ghosts withGeiger counters. The Behavior Analyst, 25, 161173.



16/18

Feyerabend, P. K. (1962). Explanation, reduction andempiricism. In H. Feigl, & G. Maxwell (Eds.), Scientificexplanation: Space and time: Minnesota studies in the philosophy of science: Vol. 3(pp. 2897). Minneapolis:University of Minnesota Press.

Gottesman, I. I., & Gould, T. D. (2003). The endopheno-

type concept in psychiatry: Etymology and strategicintentions. American Journal of Psychiatry, 160, 636645.Gottesman, I. I., & Shields, J. (1967). A polygenic theory of

schizophrenia. Proceedings of the National Academy of Sciences of the United States of America, 58, 99205.

Greenough, W. T., & Churchill, J. D. (2003). Morphologicalbasis of learning and memory: Vertebrates. In JohnH.Byrnes (Ed.), The encyclopedia of learning and memory(pp.404409). New York: MacMillan Publishing Co.

Harvey, M. T., Smith, R. L., May, M. E., Caruso, M.,Roberts, C., Patterson, T. G., Valdovinos, M., &Kennedy, C. H. (2004). Possible role for the 5-HT1Areceptor in the behavioral effects of REM sleepdeprivation on free-operant avoidance responding in

rat. Psychopharmacology (Berlin), 176, 123128.Hayes, S. C., & Barnes-Holmes, D. (Eds.) (2001). Re-

lational frame theory, A post-Skinnerian account ofhuman language and cognition. New York: Springer-

Verlag.Higgins, S. T., Heil, S. H., & Lussier, J. P. (2004). Clinical

implications of reinforcement as a determinant ofsubstance use disorders. Annual Review of Psychology,55, 431461.

Hitchcock, C. (1993). A generalized probabilistic theory ofcausal relevance. Synthese, 97, 335364.

Holm, V. A., Cassidy, S. B., Butler, M. G., Hanchett, J. M.,Greenswag, L. R., Whitman, B. Y., & Greenberg, F.

(1993). Prader-Willi syndrome: Consensus diagnosticcriteria. Pediatrics, 91, 398402.Holth, P. (2001). The persistence of category mistakes in

psychology. Behavior and Philosophy, 29, 203219.Jontes, J. D., & Smith, S. J. (2000). Filopodia, spines, and

generation of synaptic diversity. Neuron, 27, 1114.Kaas, J. H., Merzenich, M. M., & Killackey, H. P. (1983).

The reorganization of somatosensory cortex followingperipheral nerve damage in adult and developingmammals. Annual Review of Neuroscience, 6, 325356.

Kahng, S. W., Iwata, B. A., & Lewin, A. B. (2002).Behavioral treatment of self-injury, 1964 to 2000.American Journal on Mental Retardation, 107, 212221.

Kennedy, C. H. (2002). Effects of REM sleep deprivationon a multiple schedule of appetitive reinforcement.Behavior and Brain Research, 128, 205214.

Kennedy, C. H., & Meyer, K. A. (1996). Sleep deprivation,allergy symptoms, and negatively reinforced problembehavior. Journal of Applied Behavior Analysis, 29,133135.

Kennedy, C. H., Meyer, K. A., Werts, M. G., & Cushing, L.S. (2000). Effects of sleep deprivation on free-operantavoidance. Journal of the Experimental Analysis of Behavior, 73, 333345.

Kleim, J. A., Barbay, S., Cooper, N. R., Hogg, T. M., Reidel,C. N., Remple, M. S., & Nudo, R. J. (2002). Motorlearning-dependent synaptogenesis is localized to

functionally reorganized motor cortex. NeurobiologyLearning and Memory, 88, 6377.

Kranowitz, C. S. (2005). The out-of-sync child: Recognizing andcoping with sensory integration dysfunction. New York:Perigree Trade.

Leiberman, R. P., Teigen, J., Patterson, R., & Baker, V.(1973). Reducing delusional speech in chronic para-noid schizophrenics. Journal of Applied Behavior Anal-ysis, 6, 5764.

Leibman, M., & Salzinger, K. (1998). A theory-basedtreatment of psychotic symptoms in schizophrenia:

Treatment successes and obstacles to implementation. Journal of Genetic Psychology, 159, 404420.Lovaas, O. I. (1987). Behavioral treatment and normal

educational and intellectual functioning in youngautistic children. Journal of Consulting and ClinicalPsychology, 55, 39.

Lubinski, D., & Thompson, T. (1986). Functional units ofhuman behavior and their integration: A dispositionalanalysis. In T. Thompson, & M. Zeiler (Eds.), Analysisand integration of behavioral units (pp. 275314).Hillsdale, NJ: Erlbaum.

Mach, E. (1959). The analysis of sensations, and the relation ofthe physical to the psychical. (Translated from the 1st

German ed. by C. M. Williams; revised and supple-

mented from the 5th

German ed. by S. Waterlow).New York: Dover Publications (original work pub-lished 1914).

Marr, M. J. (1977). Behavioral pharmacology: Issues ofreductionism and causality. In J. E. Barrett, P. B. Dews,& T. Thompson (Eds.), Advances in behavioral pharma-cology: Vol. 7 (pp. 112). Hillsdale, NJ: Erlbaum.

Marr, M. J. (2003). The what, the how and the why: Theexplanation of Ernst Mach. Behavior and Philosophy, 31,18192.

Mayr, E. (1982). The growth of biological thought. Cambridge,MA: Harvard University Press.

McIlvane, W. J. (2003). A stimulus in need of a response: Areview of Relational frame theory: A post-Skinnerianaccount of human language and cognition. The Analysisof Verbal Behavior, 19, 2937.

McIlvane, W. J., & Dube, W. V. (2003). Stimulus controltopography coherence theory: Foundations and ex-tensions. The Behavior Analyst, 26, 195213.

Meehl, P. E. (1950). On the circularity of the Law of Effect.Psychological Bulletin, 47, 5275.

Meehl, P. E. (2006). Philosophy of science: Help orhindrance? In N. G. Waller, L. J. Yonce, W. M. Grove,D. Faust, & M. F. Lenzenweger (Eds.), A Paul Meehlreader: Essays on the practice of scientific psychology (pp.411431). Mahway, NJ: Erlbaum (reprinted fromPsychological Reports, 1993, 72, 707722).

Meerlo, P., Koehl, M., van der Borght, K., & Turek, F. W.(2002). Sleep restriction alters the hypothalamicpituitaryadrenal response to stress. Journal of Neuro-endocrinology, 14, 397402.

Meisch, R. A. (2001). Oral drug self-administration: Over-view of laboratory animal studies. Alcohol, 24, 17128.

Mellor, D. H. (1974). In defense of dispositions. ThePhilosophical Review, 83, 157181.

Michael, J. L. (1982). Distinguishing between discrimina-tive and motivational functions of stimuli. Journal of theExperimental Analysis of Behavior, 37, 149155.

Michael, J. L. (1993). Establishing operations. The BehaviorAnalyst, 16, 191206.

Michael, J. L. (2000). Implications and refinements of the

establishing operation concept. Journal of AppliedBehavior Analysis, 33, 401411.

Mill, J. S. (1986). A system of logic. Charlottesville, VA:Lincoln-Rembrandt Publisher (original work pub-lished in 1843).

438 TRAVIS THOMPSON


17/18

Moore, J. (2001). On psychological terms that appeal tothe mental. Behavior and Philosophy, 29, 167186.

Moore, J. (2002). Some thoughts on the SR issue and therelation between behavior analysis and behavioralneuroscience. The Psychological Record, 52, 261.

Moran, T. H. (2006). Gut peptide signaling in the controls

of food intake. Obesity, 14(Suppl. 5), 250253.Morris, E. K., Lazo, J. F., & Smith, N. G. (2004). Whether, when, and why Skinner published on biologicalparticipation in behavior. The Behavior Analyst, 27,153169.

Mumford, S. (1998). Laws of nature outlawed. Dialectica,52, 83101.

Nagel, E. (1961). The structure of science. London: Routledge& Kegan Paul.

Pickens, R., & Thompson, T. (1968). Cocaine-reinforcedbehavior in rats: Effects of reinforcement magnitudeand fixed-ratio size. Journal of Pharmacology andExperimental Therapeutics, 161, 122129.

Pons, T. P., Garraghty, P. E., Ommaya, A. K., Kaas, J. H.,Taub, E., & Mishkin, M. (1991, June 28). Massivecortical reorganization after sensory deafferentationin adult macaques. Science, 252, 18571860.

Prader, A., Labhart, A., & Willi, H. (1956). Ein Syndrom von Adipositas, Kleinwuchs, Kryptorchismus undOligophrenie nach Myatonieartigem Zustand imNeugeborenenalter. Schweizeriche Medizinische Wochen-schrift, 86, 12601261.

Prader-Willi syndrome, #176270. (2006). Retrieved De-cember 16, 2006, from Johns Hopkins University,Online Mendelian Inheritance in Man Web site:http://www.ncbi.nlm.nih.gov/entrez/dispomim.cgi?id5176270.

Prior, E., Pargetter, R., & Jackson, F. (1982). Three thesesabout dispositions. American Philosophical Quarterly, 19,251257.

Reichenbach, H. (1956). The direction of time. Berkeley:University of California Press.

Roeder, J. (1995). Can medication change behavior inautism? The Exceptional Parent, 25, 5054.

Ryle, G. (1949). The concept of mind. Chicago: TheUniversity of Chicago Press.

Sallows, G. O., & Graupner, T. D. (2005). Intensivebehavioral treatment for children with autism: Four-

year outcome and predictors. American Journal forMental Retardation, 110, 417438.

Salzinger, K. (1973). Schizophrenia: Behavioral aspects. New York: Wiley.

Sandman, C. A., Hetrick, W., Taylor, D. V., & Chicz-DeMet, A. (1997). Dissociation of POMC peptides after self-injury predicts responses to centrally acting opiateblockers. American Journal for Mental Retardation, 102,182199.

Schnaitter, R. (1984). Skinner on the mental and thephysical. Behaviorism, 12, 114.

Schrander-Stumpel, C. T., Curfs, L. M., Sastrowijoto, P.,Cassidy, S. B., Schrander, J. J., & Fryns, J. P. (2004).Prader-Willi syndrome: Causes of death in an in-ternational series of 27 cases. American Journal of Medical Genetics A, 124, 333338.

Schuster, C. R., & Brady, J. V. (1971). The discriminativecontrol of a food-reinforced operant by interoceptivestimulation. In T. Thompson, & R. Pickens (Eds.),Stimulus properties of drugs (pp. 133148). New York:

Appleton-Century-Crofts.

Schuster, C. R., & Thompson, T. (1969). Self-administra-tion of and behavioral dependence on drugs. AnnualReview of Pharmacology, 9, 483502.

Sedval, G., & Farde, L. (1995, September 16). Chemical brainanatomy in schizophrenia. The Lancet, 346, 743749.

Sidman, M. (1994). Equivalence relations and behavior: A

research story. Boston: Authors Cooperative.Skinner, B. F. (1938). Behavior of organisms. New York: Appleton and Company.

Skinner, B. F. (1945). The operational analysis ofpsychological terms. Psychological Review, 52, 270294.

Skinner, B. F. (1953). Science and human behavior. New York:Macmillan.

Skinner, B. F. (1972). Cumulative record. (3rd ed.) New York: Appleton-Century-Crofts.

Skinner, B. F. (1989). Recent issues in the analysis of behavior.Columbus, OH: Merril Publishing Co.

Sklar, L. (1967). Types of inter-theoretic reduction. TheBritish Journal for the Philosophy of Science, 18, 109124.

Society for stimulus properties of drugs, Drug discrimination

database. Retrieved on December 16, 2006, fromhttp://www.dd-database.org/

Spealman, R. D. (1979, June 15). Behavior maintained bytermination of a schedule of self-administered co-caine. Science, 204, 12311233.

Symons, F. J., Fox, N. D., & Thompson, T. (1998).Functional communication training and enaltrexonetreatment of self-injurious behavior: An experimentalcase report. Journal of Applied Research and IntellectualDisabilities, 11, 273292.

Taub, E. (1980). Somatosensory deafferentation research with monkeys: Implications for rehabilitation medi-cine. In L. P. Ince (Ed.), Behavioral psychology inrehabilitation medicine: Clinical applications (pp. 317401). New York: Williams & Wilkins.

Taub, E., Miller, N. E., Novack, T. A., Cook, E. W.,Fleming, W. C., Nepomuceno, C. S., Connel, J. S., &Crago, J. E. (1993). Technique to improve motordeficit after stroke. Archives of Physical Medicine andRehabilitation, 74, 347354.

Thompson, T. (2005). Paul E. Meehl & B. F. Skinner:Autitaxia, autitypy and autism. Behavior and Philosophy,33, 101131.

Thompson, T., Hackenberg, T., Cerutti, D., Baker, D., & Axtell, S. (1994). Opioid antagonist effects on self-injury in adults with mental retardation: response formand location as determinants of medication effects.

American Journal for Mental Retardation, 99, 85102.Thompson, T., & Pickens, R. (Eds.). (1971). Stimulusproperties of drugs. New York: Appleton-Century-Crofts.

Thompson, T., & Schuster, C. R. (1964). Morphine self-administration, food-reinforced and avoidance beha-

viors in Rhesus monkeys. Psychopharmacologia, 5, 5794.Thompson, T., & Zeiler, M. (Eds.). (1986). Analysis and

integration of behavioral units. Hillsdale, NJ: Erlbaum. Vanderbeeken, R., & Weber, E. (2002). Dispositional

explanations of behavior. Behavior and Philosophy, 30,4359.

Vanderschuren, L. J., & Everitt, B. J. (2005). Behavioraland neural mechanisms of compulsive drug seeking.European Journal of Pharmacology, 526, 7788.

Waddington, C. H. (1942). The epigenotype. Endeavour, 1,1820.

Weeks, J. R. (1962, October 12). Experimental morphineaddiction: Method for automatic intravenous injec-tions in unrestrained rats. Science, 138, 143144.



18/18

Wharton, R. H., Wang, T., Graeme-Cook, F., Briggs, S., &Cole, R. E. (1997). Acute idiopathic gastric dilation

with gastric necrosis in individuals with Prader-Willisyndrome. American Journal of Medical Genetics, 73,437441.

White, I. D., Harrison, S. J., & Mottershead, D. N. (1992).

Environmental systems: An introductory text. London: G.Allen and Unwin Publishers.

Yokel, R. A. (1987). Intravenous self-administration: Re-sponse rates, the effects of pharmacological chal-lenges, and drug preference. In M. A. Bozarth (Ed.),Methods of assessing the reinforcing properties of abuseddrugs (pp. 133). New York: Springer-Verlag.

Received: May 7, 2006Final acceptance: December 29, 2006

440 TRAVIS THOMPSON

Date post:	08-Apr-2018
Category:	Documents
Upload:	viviana-edith-arce
View:	227 times
Download:	0 times

Jeab, AC y Neurociencias

Documents