JOURNAL OF THE EXPERIMENTAL ANALYSIS OF BEHAVIOR 2005, 84, 305–311 NUMBER 3 (NOVEMBER)

RELATING BEHAVIOR AND NEUROSCIENCE: INTRODUCTION AND SYNOPSIS

WILLIAM TIMBERLAKE, DAVID W. SCHAAL, AND JOSEPH E. STEINMETZ
INDIANA UNIVERSITY AND STANFORD UNIVERSITY

_______________________________________________________________________________
Background crimination; finalizing distinctions among
stimulus functions; grouping reflexes by drives
B. F. Skinner, in a chapter on ‘‘Behavior and and emotions; and establishing the dynamics
the Nervous System’’ in his seminal work, The of reinforced behavior.
Behavior of Organisms (1938, pp. 418–432), Thirty-six years later in a chapter on ‘‘What
expressed both strong interest in and consid- is Inside the Skin?’’ in About Behaviorism (1974,
erable concern about relating behavior and pp. 207–218), Skinner reaffirmed the impor-
what he termed ‘‘neurology.’’ On the positive tance of a reductionist framework, and again
side, he subscribed to a unified reductionist rejected attributing the cause of a behavior to
science: ‘‘One of the objectives of science is a single neurobiological entity, whether it was
presumably the statement of all knowledge in a synapse, an anatomical structure, an emo-
a single language.’’ He treated this objective as tion, or a motivation. The possible exception
a fundamental reason to intensely cultivate he noted was appealing to neural events to fill
a behavioral approach because it would pro- inevitable gaps in an operant account. For
duce ‘‘. . . more rapid progress toward an example, because behavioral accounts of re-
inforcement are ‘‘necessarily historical,’’ they
ultimate synthesis [of the laws of behavior and
leave gaps between events that might be filled
the laws of the nervous system].’’
in by neural processes related to memory. It
At the same time Skinner spoke strongly
was clear, though, that any large-scale integra-
against ‘‘. . . proceeding from a behavioral fact
tion remained far in the future, following the
to its neural correlates instead of validating the
establishment of comprehensive and indepen-
fact as such, and then proceeding to deal with
dent behavioral and neural sciences.
other problems in behavior.’’ In short, his goal
From the publication of The Behavior of
was: first, to establish an independent science
Organisms to the present, the analysis and
of the control and dynamics of behavior,
control of behavior has proliferated in aca-
separate from neural, physiological, and cog- demic and applied settings around the world.
nitive references and speculations; and then, The Association for Behavior Analysis has
to bridge the gap between behavior and grown to around 4700 members spread among
neurobiology by a comprehensive integration. 43 countries. The study of neuroscience has
Skinner also placed strict preconditions on expanded even more rapidly. The Society for
both sciences. He argued that neurobiological Neuroscience is a thoroughly international
concepts had to be adequate to account for organization numbering over 37,000 mem-
the complexity of the dynamics of behavior, bers. Most relevant for our interest in relating
and that numerous behavioral issues needed behavior and neuroscience, Skinner’s meth-
to be resolved, among them: distinguishing ods have been widely adapted by neuroscien-
Type R (Operant) and S (Pavlovian) condi- tists. In fact, Skinner and his students played
tioning; clarifying the basis of temporal dis- crucial roles in founding psychopharmacolo-
gy, a field pursued in medical schools, drug
companies, and multiple academic depart-
The authors thank Colin Allen, Gary Lucas, and ments (Laties, 2003). Other scientists influ-
Armando Machado for their suggestions, and Len Green
for his perseverance and help. enced directly or indirectly by Skinner
William Timberlake and Joseph Steinmetz are at adopted operant procedures and apparatus
Indiana University; David Schaal is now at Accuray, Inc. to measure the effects of brain lesions or brain
Correspondence may be addressed to William Timber- stimulation on learning, discrimination, and
lake, Department of Psychology, Indiana University, 1101
E. 10th Street, Bloomington, IN 47405-7007, Telephone:
motivation (e.g., Grossman, 1979). Still others
812-855-4042 (e-mail: [email protected]). have used operant techniques to study drug
doi: 10.1901/jeab.2005.99-05 addiction, memory, the molecular basis of

305
306 WILLIAM TIMBERLAKE et al.

reinforcement, and effects of environmental tion, Choice, and Cortical Brain Activity; and
pollutants on learning and behavior. IV. Neural and Behavioral Analyses of Learn-
In short, there is considerable evidence for ing-Related Brain Circuitry and Addiction. We
the existence of independent sciences of begin this issue with a historical preface to the
behavior and of neurobiology, and there is problem of reductionism by the philosopher
research that combines aspects of both. What William Bechtel, and finish with a book review
is missing is the broad conceptual integration by David Schaal of Bennett and Hacker’s
that Skinner began pointing toward in 1938. It Philosophical Foundations of Neuroscience.
might be argued that the time is not yet right
because neuroscience has yet to account for
SYNOPSIS
the fine-grained dynamics of operant behavior
(or because operant conditioning has yet to Preface: Issues in Bridging Levels of Analysis
resolve the majority of the behavioral prob- William Bechtel (The challenge of characteriz-
lems Skinner laid out in 1938). But such ing operations in the mechanisms underlying
objections, however reasonable they may ap- behavior, pp. 313–325) notes two problematic
pear, ignore research that already has begun tendencies that need to be overcome in
to integrate neuroscience and behavior. relating phenomena across levels of analysis.
What seems clear at this point is that there The first tendency is to presume that causal
has not been (and will not be) one obvious workings (Bechtel uses the term ‘‘opera-
moment when scientists agree to pursue full tions’’) at a lower level of analysis are of the
integration of neuroscience and behavior (or same sort as those analyzed at higher levels.
among neuroscience, behavior, and cogni- Bechtel draws an example from physical
tion). Neither will there be a single set of chemistry and the phenomenon of fermenta-
experiments, or a single experimental ques- tion. As a parallel we might consider the
tion, that seamlessly joins these approaches. concept of reinforcement. By analogy to
Instead, there have been (and will continue to Bechtel’s example, attempts to reduce behav-
be) specific instances when enough is known ioral reinforcement to lower-level mechanisms
about neurobiological and behavioral phe- typically are based on concepts of behavioral
nomena to integrate levels successfully (e.g., reinforcement, which, instead of referring to
Schaal, 2003). Further, the potential for in- behavior-level operants, discriminative stimuli,
tegration will be greater as experimenters use and reinforcers, are used to designate parallel
causal manipulations and analyses that consid- elements at the level of circuits, individual
er both neuroscience and behavior. At the neurons, transmitters, ion channels, and/or
least, operant procedures can distinguish molecules.
among effects of motivation versus discrimina- The second tendency Bechtel calls attention
tion or memory disruption, while increased to is that of skipping levels of analysis when
knowledge of the involvement of specific brain invoking causal connections. Applied to re-
mechanisms can direct experimental attention inforcement, his warning calls attention to the
to specific manipulations, behaviors, and the multiple levels of mechanism that are omitted
timecourse of effects. when we attribute the reinforcement of lever
pressing to, say, GABA release. Such a correla-
Purpose and Organization tion, even if present, does not specify the
The purpose of this special issue of the mechanisms that connect the multiple levels
Journal of the Experimental Analysis of Behavior is of organization separating GABA release and
to present experimental and conceptual arti- lever pressing. As an engineering problem, we
cles that illustrate recent efforts to relate would be unable to duplicate the effect in
neuroscience and behavior in individual sub- a realistic model of an organism.
jects. In abstract terms, this issue is intended A simple example may help illustrate both
to illustrate mappings between behavior of Bechtel’s points. Consider accelerating a car.
and neuroscience that range from simple At the level of the driver, the cause involves
to complex. In concrete terms, this issue pressing down on the gas pedal. At the level of
focuses on four areas: I. Effects of Drugs and production of energy, the cause is the igniting
Genetics on Common Operant Tasks; II. of gasoline under pressure. In between are
Complex Stimulus Relations; III. Discrimina- multiple workings including: throttle linkage,
 INTRODUCTION AND SYNOPSIS 307

carburetor, gas tank, fuel pump, spark plugs, pressing for food, but they add additional
valves, timing rotor, pistons, cam shaft, drive measures, including: pressing an inoperative
train, clutch, transmission, axles, wheels, and lever, locomotor activity, and the time re-
tires. There are no ‘‘accelerators’’ to press at quired to move from lever to food hopper
lower levels, and no ‘‘explosions’’ at higher and back to lever.
levels. The mechanisms differ at each level, (3) Neill, Liu, Mikati, and Holmes–Pilocarpine
and they all are necessary for the car to seizures cause age-dependent impairment in auditory
accelerate. We cannot expect links between location discrimination. (pp. 357–370)
behavior and neurobiology to be any less Following pilocarpine injections at 20 or
complex. 45 days of age, rats were tested on auditory
location and sound/silence discriminations
Section I: Effects of Drugs and Genetics in Common beginning at 105 days of age. Pilocarpine
Operant Tasks injections at day 20 moderately impaired
The experiments reported in Section I use subsequent auditory location discrimination,
well-developed operant tasks to reveal multiple but injections at day 45 produced subsequent
effects of drugs and genes on behavior. failure to acquire either type of discrimination.
Researchers report on the role of reinforce- The results are related to clinical cases of
ment schedules in clarifying drug effects on humans with a history of seizures and impaired
extinction (Leslie, Shaw, Gregg, McCormick, auditory location discrimination.
Reynolds, & Dawson); mouse strain effects on (4) Bratcher, Farmer-Dougan, Dougan, Hei-
lever pressing for food (McKerchar, Zarcone, denreich, and Garris–The role of dopamine in
& Fowler); the developmental effects of drug- reinforcement: Changes in reinforcement sensitivity
induced seizures on hearing (Neill, Liu, induced by D1-type, D2-type, and nonselective
Mikati, & Holmes); the role of dopamine dopamine receptor agonists. (pp. 371–399)
agonists in reinforcement matching (Bratcher, These authors are concerned with the
Farmer-Dougan, Dougan, Heidenreich, & specific actions of D1 and D2 receptors in
Garris); and the effect of morphine on facilitating rewarded behavior in rats. They
stimulus control by color and time (Ward & studied the effects of injections of specific and
Odum). general dopamine agonists using the matching
(1) Leslie, Shaw, Gregg, McCormick, Rey- paradigm, because this well-developed pro-
nolds, and Dawson–Effects of reinforcement sched- cedure provides measures of sensitivity to
ule on facilitation of operant extinction by chlordi- reinforcement. They obtained dose-depen-
azepoxide. (pp. 327–338) dent changes in sensitivity produced by D1
These researchers test the effects of chlor- and general agonists, but not by a D2 agonist.
diazepoxide (a GABAergic drug) on the (5) Ward and Odum–Effects of morphine on
extinction of discrete-trial fixed-ratio or fixed- temporal discrimination and color matching: Gen-
interval responding in the inbred mouse strain eral disruption of stimulus control or selective effects
C57Bl/6 (a standard strain used in research). on timing ? (pp. 401–415)
Extinction under all schedules was subject to The authors used a three-component multi-
facilitation by drug injections of chlordiaz- ple schedule to compare the effects of
epoxide, but only if a number of extinction morphine on temporal discrimination and
sessions had already occurred, and more so on accuracy of color matching. Their intent was
FI than on FR schedules. The authors suggest to clarify the mechanisms by which morphine
that their effects are due to multiple processes affects timing. Psychophysical analyses of
in extinction. schedule behavior showed that morphine de-
(2) McKerchar, Zarcone, and Fowler–Differen- creased overall stimulus control by time, rather
tial acquisition of lever pressing in inbred and than selectively affecting timing abilities or
outbred mice: Comparison of one-lever and two-lever abililty to respond to the key color associated
procedures and correlation with differences in with the different schedules.
locomotor activity. (pp. 339–356)
The authors examine the potential useful- Section II: Complex Stimulus Relations
ness of testing for behavioral and procedural The studies in Section II focus on the
differences among multiple mouse strains relation of brain activity and complex stimulus
using the standard operant task of lever discriminations (equivalence classes, analo-
308 WILLIAM TIMBERLAKE et al.

gies, transitive inference, and biconditional cue designating either the same or different
discriminations involving context). The tech- lever. Sham-control rats picked the lever pre-
niques used to assess brain activity include: dicted by the context. In lesioned rats, impair-
evoked potentials in prefrontal cortex (Barnes- ment in using the context was proportional to
Holmes, Staunton, Whelan, Barnes-Holmes, the amount of damage to anterior cingulate and
Commins, Walsh, Stewart, Smeets, & Dymond; prelimbic areas. Presumed effects of anterior
and Barnes-Holmes, Regan, Barnes-Holmes, cingulate damage decreased across a conflict
Commins, Walsh, Stewart, Smeets, Whelan, & test trial, whereas effects tied to prelimbic
Dymond), f MRI in hippocampus (Dickins), damage continued through the trial.
and relative lesion damage to anterior cingu-
late and prelimbic areas of the cortex (Had- Section III: Discrimination, Choice, and Cortical
don & Killcross). Brain Activity
(1) Barnes-Holmes, Staunton, Whelan, Barnes- The researchers in Section III are con-
Holmes, Commins, Walsh, Stewart, Smeets, cerned with the relation of f MRI and single-
and Dymond–Derived stimulus relations, semantic cell recording from prefrontal and parietal
priming, and event-related potentials: Testing a be- cortex to standard choice tasks such as the
havioral theory of semantic networks. (pp. 417– training of discriminative stimuli in humans
433) (Schlund & Cataldo); delayed go/no-go dis-
(2) Barnes-Holmes, Regan, Barnes-Holmes, criminations in pigeons (Kalenscher, Güntür-
Commins, Walsh, Stewart, Smeets, Whelan, kün, Calabrese, Gehlen, Kalt, & Diekamp);
and Dymond–Relating derived relations as a model choosing between small, short-term rewards
of analogical reasoning: Reaction times and event- and larger, long-term rewards in humans
related potentials. (pp. 435–451) (Yarkoni, Braver, Gray, & Green), and special-
These authors use operant Matching-to- ized matching tasks using visual responses to
Sample discrimination training and subse- a simulated foraging environment in rhesus
quent testing procedures with humans to monkeys (Lau & Glimcher; and Corrado,
establish and test for the conceptual ‘‘frames’’ Sugrue, Seung, & Newsome).
relating multiple stimuli (forming an equiva- (1) Schlund and Cataldo–Integrating functional
lence class) or stimulus pairs (forming similar neuroimaging and human operant research: Brain
or different ‘‘analogies’’). Both measures of activation correlated with presentation of discrimi-
reaction time and bilateral event-related po- native stimuli. (pp. 505–519)
tentials in prefrontal cortex provide support These researchers train human subjects by
for the development of complex relations rewarding a button press in the presence of
based on simpler discrimination training. one set of discriminative stimuli, and reward-
(3) Dickins–On aims and methods in the neuroi- ing no responding by terminating the trial in
maging of derived relations. (pp. 453–483) the presence of another set of discriminative
This author reviews the use of neuroimaging stimuli. The subjects simply memorize a third
studies of the hippocampal area of the brain as set of (control) stimuli. Subsequent exposure
a means of analyzing learned phenomena to all three sets of discriminative stimuli
related to the emergence of transitive in- during fMRI scanning showed greater activa-
ference from the study of both stimulus tion in frontal and striatal brain regions for
equivalence and serial learning. He also briefly both sets of discriminative stimuli than to the
discusses the potential roles of evolution and control stimuli. These results support attribut-
ethology in relating brain and behavior. ing differences between brain imaging effects
(4) Haddon and Killcross–Medial prefrontal in human and nonhuman animals to the use
cortex lesions abolish contextual control of competing of a required operant response with non-
responses. (pp. 485–504) human animals.
These authors focus on the effects of medial (2) Kalenscher, Güntürkün, Calabrese, Geh-
prefrontal cortex damage on the behavior of len, Kalt, and Diekamp–Neural correlates of
rats trained in biconditional audio and visual a default response in a delayed go/no-go task. (pp.
discriminations. In the experiment, the lever 521–535)
associated with each audio (or visual) cue was These researchers study pigeons trained on
correct in one context, but not in another. a delayed go/no-go task in which an initial cue
The test consisted of one video and one audio signals the appropriate response. Retention
 INTRODUCTION AND SYNOPSIS 309

functions diverged across intervals indicating characteristics of choice behavior (hyperbolic

that the pigeons retained the specifics of the discounting, differential comparisons of op-
go stimulus, but not of the no-go stimulus. tion value, and empirical maximizing of
Single-cell recording from the avian nidopal- feeding). The strategies also produce local
lium caudolaterale (similar to the mammalian scalar value metrics highly correlated with the
prefrontal cortex) supported the view that the firing of individual neurons in the lateral
no-go response is a default response that is not intraparietal area of the cortex.
stored in the same way as the go response.
(3) Yarkoni, Braver, Gray, and Green–Prefron- Section IV: Neural and Behavioral Analyses of
tal brain activity predicts temporally extended Learning-Related Brain Circuitry and Addiction.
decision-making behavior. (pp. 537–554) The papers in the final section focus on
The authors use fMRI to explore neural integrating multiple neural cells and circuitry
activity underlying behavior during a task re- with learning in circumstances that include:
quiring the subjects to select small, short-term relating activation of a set of hippocampal
rewards to ultimately receive larger, long-term neurons to tasks tapping episodic memory
gains. Choice behavior is predicted by neural effects in rats (Eichenbaum & Fortin); relating
activity in a surprisingly specific area of the complex cerebellar circuits to eyeblink condi-
right lateral prefrontal cortex, a region pre- tioning phenomena in rabbits (Villarreal &
viously reported to be related to cognitive Steinmetz); and establishing the relation of
effort. In contrast, activity in the insular cortex single neurons in nucleus accumbens and
responded to fluctuations in amount of reward prefrontal cortex to drug-seeking relapse in
but did not predict choice behavior. rats (Rebec & Sun). The final paper in Section
(4) Lau and Glimcher–Dynamic response-by-re- IV (Winger, Woods, Galuska, & Wade-Galuska)
sponse models of matching behavior in rhesus raises the cautionary point that strictly neuro-
monkeys. (pp. 555–579) biological accounts of drug addiction may not
These researchers manipulate frequency be as useful for understanding and treating
and amount of reward in a discrete-trials addicted behavior as approaches based on an
matching procedure with rhesus monkeys. analysis of drugs as reinforcers.
Their goal is to identify brain areas relevant (1) Eichenbaum and Fortin–Bridging the gap
to integration of past reinforcers and choices. between brain and behavior: Cognitive and neural
They minimize neurophysiological recording mechanisms of episodic memory. (pp. 619–629)
problems by using eye rather than hand These researchers analyze behavioral and
movements for responses, and they minimize electrophysiological evidence supporting the
possible recording artifacts from jaw move- view that critical features of episodic memory
ments by using water rather than food re- (recollection, as opposed to recognition, of
inforcers. Using linear models of response-by- stimuli and places) are readily shown in studies
response behavior, they find it necessary to of rats using olfactory stimuli. These features
weight recent reinforcers more heavily and to include thresholds for retrieval effects (instead
consider patterning of choice responses to of continuous functions) and memory for
predict steady-state behavior, fluctuations in sequences of events. Their data show that
responding during transitions between densi- a particular set of neurons in the hippocampus
ties of reinforcement, and response-by-re- is selectively and reliably activated across
sponse patterns. a variety of tasks, thus providing the basis for
(5) Corrado, Sugrue, Seung, and Newsome– networks of neurons activated by retrieval
Linear-nonlinear-Poisson models of primate choice cues, including sequences of behavior.
dynamics. (pp. 581–617) (2) Villarreal and Steinmetz–Neuroscience and
These investigators also study the perfor- learning: Lessons from studying the involvement of
mance of rhesus monkeys in a matching a region of cerebellar cortex in eyeblink classical
paradigm using liquid rewards ( juice, in this conditioning. (pp. 631–652)
case) and eye movements in a dynamic video These investigators outline the methods of
simulation of a foraging environment. Their basic neuroscience used to explore the re-
aim also is to model the effects of past rewards lation between behavioral learning and the
on future choice. Their mathematically spec- brain, including brain lesions, stimulation,
ified strategies make contact with known pharmacology, anatomy, imaging, and record-
310 WILLIAM TIMBERLAKE et al.

ing. They review the results of these tech- humans show planning. Much of what the
niques in working out the complex neural authors of the book write will have a familiar
circuitry involved in classical conditioning of ring to behavior analysts who long have
the eyeblink, and show how the technique of criticized similar conceptual tendencies.
temporary brain inactivation combined with
single-cell recording can clarify the contribu-
tion of a specific region of the cerebellar POSTSCRIPT
cortex (Larsell’s lobule HVI). There appear to be two reasonable strategies
(3) Rebec and Sun–Neuronal substrates of relapse in bridging the gap between behavior and
to cocaine-seeking behavior: Role of prefrontal cortex. neuroscience. A pure strategy is to stay
(pp. 653–666) primarily on one side or the other until the
These researchers review animal models of science on both sides has developed more
drug addiction relapse (reinstatement) based fully. If you are a behaviorist, focus your efforts
on electrophysiological analysis relating the on amplifying the sophistication of your
nucleus accumbens and the prefrontal cortex measures, manipulations, and analyses. If you
as a final common pathway. They then use are a neuroscientist, focus on in vivo prepara-
single-unit recording in behaving animals to tions of brain cells, working out circuitry,
distinguish whether the activity of individual neuronal functioning, and cellular processes.
neurons in these brain structures reflects cues This strategy appears related to the initial
that signal the drug, the drug itself, or the ‘‘separate but equal’’ approach Skinner out-
motor sequence producing the drug. They lined in 1938. The gap between behavior and
provide preliminary findings that prefrontal neuroscience should be maintained until the
cortex processes information related to co- two sides produce parallel structures of con-
caine seeking but not information about the cepts and relations that can be easily integrat-
hedonic effects of the drug. ed. This solution simplifies the questions to be
(4) Winger, Woods, Galuska and Wade-Ga- asked on each side and the models developed.
luska – Behavioral perspectives on the neuroscience The difficulty is that it does not make clear
of drug addiction. (pp. 667–681) how to identify the proper time for integra-
These authors point out that neurophysio- tion, what constitutes parallel development,
logical accounts of drug addiction presume exactly how integration should occur, who will
that addiction is based on long-term brain do it, and how the science might be different.
changes caused by chronic administration of An alternative, mixed strategy, the one
drugs. They suggest the usefulness of the seemingly favored by most researchers in this
alternative perspective, namely that drug issue, is to develop specific, albeit limited,
addiction is a behavioral disorder in which bridges relating behavior and neuroscience.
drugs are preeminent reinforcers. The authors Given a focus on specific bridges, experimen-
argue that advances in treating drug addiction ters can focus more readily on areas where
and understanding mechanisms related to bridges appear interesting, possible, and most
excessive drug use will occur more rapidly by likely to be successful. As is evident from the
assessing the reinforcing effects of drugs. present articles, some bridges are based on
relating well-controlled tasks to manipulations
Book Review of drugs, transmitters, lesions, and measures of
Schaal–Naming our concerns about neuroscience: A brain activity. To increase the explanatory or
review of Bennett and Hacker’s Philosophical causal traffic the bridge can bear, it should be
Foundations of Neuroscience. (pp. 683–692) anchored to multiple locations both in behav-
The authors of the book under review ior and in the brain. For example, the classic
present a thorough critique of the tendency and continuing studies of how barn owls
of cognitive neuroscientists to commit the localize and react to sounds in space provide
mereological fallacy, the tendency to ‘‘explain’’ wonderful examples of how knowledge gained
psychological phenomena by ascribing to the about sensory neurobiology of the face feather
brain (or its parts) psychological powers structure and the brain clarifies how to
justifiably applied only to whole animals. For perform better experiments, which in turn
example, it is a fallacy to say that our forebrain reveal more about how the sensory-motor
‘‘plans’’ when attempting to explain why mechanisms work (e.g., Carew, 2000). Such
 INTRODUCTION AND SYNOPSIS 311

interactive complexity is gradually developing robiology with eyes wide open is an important
in areas with better-established bridges (e.g., direction for operant conditioning, and
the articles here by Eichenbaum & Fortin, learning in general, to proceed. It also is
Rebec & Sun, Villarreal & Steinmetz; see also important to recall Skinner’s goal of fully
the review by Thompson, 2005). understanding both sides of the behavior-
Certainly, rapidly constructed, poorly con- neurobiology gap, as well as Bechtel’s compat-
ceived bridges often have drawbacks (e.g., see ible point about the necessity of filling in the
the discussion by Winger et al.). Researchers working mechanisms that stitch together levels
caught up in the excitement of relating of analysis.
measures of neurobiology and behavior may
find that their anchors on one side or the
other do not bear the weight of their predic-
tions or assumptions. Skinner’s warning REFERENCES
against attributing causality to weakly an- Carew, T. J. (2000). Behavioral neurobiology: The cellular
chored, neurobiologically tinged concepts is organization of natural behavior. Sutherland, MA:
still good advice. However, his concern about Sinauer Associates, Inc.
causal anchoring applies also to the behavioral Grossman, S. P. (1979). The biology of motivation. Annual
Review of Psychology, 30, 209–242.
side. We should not forget that the causal Laties, V. G. (2003). Behavior analysis and the growth of
interpretation of an operant reinforcement behavioral pharmacology. The Behavior Analyst, 26,
contingency itself is firmly anchored only in 235–252.
the sense that we can manipulate its compo- Schaal, D. W. (2003). Explanatory reductionism in
nents and observe regular results (Skinner, behavior analysis. In K. A. Lattal & P. N. Chase
(Eds.), Behavior theory and philosophy (pp. 83–102). New
1938; Timberlake, 2004). It is not clear what is York: Kluwer Academic/Plenum Publishers.
to be done with phenomena such as autoshap- Skinner, B. F. (1938). The behavior of organisms: An
ing, misbehavior, and superstitious behavior, experimental analysis. New York: Appleton-Century-
which are not similarly anchored in typical Crofts.
operant manipulations. Such phenomena Skinner, B. F. (1974). About behaviorism. New York: Alfred
might profit from more secure causal connec- A. Knopf.
Thompson, R. F. (2005). In search of memory traces.
tions with behavior and with well-defined Annual Review of Psychology, 56, 1–23.
neurobiological processes. Timberlake, W. (2004). Is the operant contingency
The articles in this issue suggest that enough for a science of behavior? Behavior and
bridging the gap between behavior and neu- Philosophy, 32, 197–229.