See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/231181293

An introduction to principles of behavior

Chapter · January 2007

CITATIONS READS

10 2,391

3 authors, including:

Thomas Waltz Steven C Hayes

Eastern Michigan University University of Nevada, Reno
33 PUBLICATIONS   3,119 CITATIONS    476 PUBLICATIONS   47,467 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

LIFENGAGE randomised controlled trial of acceptance and commitment therapy for persisting psychosis View project

http://bostonanxiety.org/currentresearch.html View project

All content following this page was uploaded by Thomas Waltz on 31 May 2014.

The user has requested enhancement of the downloaded file.

Drossel. C., Waltz, T. J., & Hayes, S. C. (2007). An introduction to
principles of behavior.
Introduction to PrinciplesInofD. Woods & J. Kantor (Eds.)
Behavior 1
Understanding behavior disorders: a contemporary behavior
analytic perspective (pp. 21-46). Reno, NV: Context Press.

*Note that the pages breaks in this pre-publication

Chapter 2
proof differ slightly from those found in the book.

An Introduction to Principles of Behavior

Claudia Drossel, Thomas J. Waltz, & Steven C. Hayes

University of Nevada, Reno

Behavioral principles are not static. They develop systematically across the
years. Perhaps more than any other approach to psychology, behavioral psychology
is self-consciously committed to being a progressive science in which worthwhile
earlier work is sustained even while new frontiers are explored. We are not speaking
of conservatism for its own sake, but rather for the creation of a field that knows how
to keep its hard won gains. A field that cannot do so will appear to renew itself by
reinventing what was known earlier and forgotten. Scientific amnesia of that kind
gives only the appearance of progress, not its essence.
Built upon a century of research in behavioral psychology, in the last decades
a coherent and progressive approach to the topic of human language and cognition
has emerged from within behavior analysis. This volume is designed to attempt to
account for human psychopathology from the perspective of this comprehensive set
of modern behavioral principles. To do that, it is necessary to describe these
principles as well as the foundation upon which they rest.
The purpose of this chapter is to describe a workable core set of basic behavioral
principles that can be used to understand complex human behavior. We will outline
basic principles of learning and then extend those to the empirical analysis of human
language and cognition. The next chapter will then describe how these fundamental
principles are related to more middle-level constructs specifically focused on
complex issues encountered in applied work, such as issues of relationship or sense
of self.
Basic Behavioral Principles
Contingency Thinking
Everybody attempts to explain behavior: Parents ponder the reasons for their
children’s behavior; teachers attempt to influence their students’ behavior; star-
crossed lovers try to predict the behavior of their chosen one. This is both a blessing
and a curse. It is a blessing because it is the source of public interest in psychological
science. It is a curse because lay explanations of behavior strongly compete with
scientific ones (Hineline, 1980).
Behavioral psychology emerges from a pragmatic and contextualistic philoso-
phy of science that treats psychological events as the integrated action of a whole
 2 Chapter 2

organism interacting in and with a context considered both historically and

situationally (Hayes, Hayes, & Reese, 1988). These situated actions include all
forms of psychological activity: moving, speaking, thinking, feeling, reasoning,
problem-solving, and so on. It does not matter whether the action is publicly
observable or observable only to the person engaging in it. Even private events can
be studied scientifically, for reasons we will describe later (Skinner, 1945).
When behavior is approached in this pragmatic and contextualistic way, actions
and their contexts are whole units. For example, going to the store implies a place
to go from and to, a motivation to go, and a set of behavioral events that will
accomplish that purpose. Going to the store is not just the action of a leg, or the
twitches of muscle fibers. It is a whole event, but one with discernable aspects. Like
a person looking at different sides of a box, behavioral psychologists look at different
aspects of a whole behavioral episode to discern features that might help this entire
event to be predicted and influenced. The rules that are abstracted to do that need
to have precision (that is, only a certain number of things can be said about any
episode), scope (that is, ways of speaking about one episode should apply to others),
and depth (that is, successful ways of speaking at the level of psychological events
should comport with successful analyses at the level of others sciences, such as
biology, sociology, and so on).
The aspects – the “sides of the box” as it were – that are examined by behavioral
psychology include the states of affairs that were present when an action occurred,
the act itself, and the states of affairs that have been and are now produced by those
actions. Broadly speaking, the first set of events is termed “antecedents,” and the last
set is called “consequences.” These two considered together are all aspects of
“context.” The relationships among antecedents, actions, and consequences are
called “contingencies.”
These events can all be further divided for analytic purposes. Antecedents can
include environmental events that have a psychological function (e.g., behavior
occurs more or less frequently in their presence); conditions that alter the functions
of other antecedent and consequential events (e.g., motivational events; drugs;
augmentals); and the strength or occurrence of various other situated actions (e.g.,
repertoire; verbal rules). Consequences, on the other hand, include all events that
have reliably followed behaviors in the past.
Although behavior only occurs in the moment, a snapshot of behavior does not
reveal the stable antecedents or consequences nor the relationship among them. No
one situation is ever exactly like another. Thus, when we examine behavioral
patterns to predict or influence behavior in the future we focus on patterns that are
evident over time. When we want to understand a particular behavior pattern, we
try to determine those antecedents and consequences that help us predict and
influence from situation to situation, that is, those that reveal reliable patterns and
consistencies. For this reason, behavioral psychologists observe antecedents,
behavior, and consequences repeatedly and then identify patterns of events that
usually precede and follow an action.
Introduction to Principles of Behavior 3

All contingency streams are analyzed using the basic concepts in Figure 2.1. The
dot denotes an emitive probabilistic relation between terms while the arrow
indicates a contingent relation. Some analyses rely primarily on the terms in the top
panel, Antecedents < Behavior → Consequences. That integrated whole is called an
operant “three-term contingency,” simply because the behavior operates upon the
environment, and the integrated unit consists of three terms. It is “integrated”
because each can only be defined in terms of the other, and thus these are aspects
of a whole, not a mechanical assemblage. Other analyses might keep track explicitly
of the various events that alter these three events, any of which might be verbal. For
example, when most organisms are food-deprived, food-related antecedents and
consequences will come to the fore. The middle panel illustrates such a “four-term
contingency” analysis. This is not the end of the complexity, however. There is now
a good deal of evidence that the contingencies related to verbal behavior can vary
in the degree to which they coincide with the contingencies for other behavior; at
times, these contingencies may interact, compete with, or override one another.
Excessive influence by verbal relations and consequent insensitivity to other
contingencies is of particular interest in the clinical realm (see later chapters). For
this reason, it is often necessary to keep track of multiple sets of contingencies –
operating upon verbal and other behavior, and their interactions. When these
variables are included, a contingency analysis involves the more complex set of
considerations shown in the bottom panel of Figure 2.1.
Reinforcement and Punishment Contingencies
Although contingency analysis in a broad sense existed before the work of B.
F. Skinner, he formalized this approach in 1938 and added a focus on whether
specific contingencies resulted in an increase or decrease in the frequency of
behavior over time. Reinforcement and punishment contingencies are distinguished
by their effects on behavior.
Any behavior occurs in a given situation (the antecedent) and is followed by
particular consequences. If, over time, the behavior increases in frequency (due to
the contingency among antecedents, behavior, and consequences), the conse-
quences are characterized as reinforcers and the operation is called reinforcement.
If the behavior decreases in frequency (due to the contingencies), the consequences
are characterized as punishers. “Reinforcement” and “punishment” are thus contin-
gent relations (or dependencies) linking the frequency with which a behavior occurs,
its antecedents, and its consequences. These “three-term-contingencies” constitute
the simplest unit of analysis in behavioral psychology. It is important to remember
that this unit encompasses not only the events preceding and following the behavior,
but the dependent relations between them and resulting changes in the frequency
of that behavior over time. Any one of the three terms (antecedents, behavior, or
consequences) may be verbal.
4 Chapter 2

Analysis in terms of the three-term contingency

Antecedents < Behavior → Consequences

Analysis in terms of the four-term contingency

Antecedents < Behavior → Consequences

↔

Establishing operations and setting events

Analysis in terms of direct and derived four-term contingencies

Antecedents < Behavior → Consequences

↔

Establishing operations and setting events

↔

Verbal Antecedents < Verbal Behavior → Verbal Consequences

↔

Verbal establishing operations and setting events

Figure 2.1. Forms of Contingency Analysis.

Introduction to Principles of Behavior 5

For convenience and simplification, Table 2.1 describes possible behavior-

consequence outcomes in a binary fashion (that is, in terms of whether or not these
events occur or do not occur).

(1) A behavior (2) Following events are

occurs →

Presented Prevented/removed
Increases Positive Negative
reinforcement reinforcement
(3) Over time, (avoidance, escape)
the frequency of
behavior Decreases Positive Negative
punishment punishment
(timeout, penalty)

Table 2.1. Reinforcement and punishment contingencies.

Other behavior-consequence relations are not describable as binary events but

rather involve subtle changes in magnitude, duration, or quality. Thus, in a given
state of affairs behavior occurs and is followed by a changed state of affairs. In Table
1 the changed state of affairs can be characterized either by the presentation of events
or the prevention or removal of events. Skinner termed the presentation of events
“positive,” and their removal, postponement, prevention, or termination “negative.”
It helps to think of these terms arithmetically. Like a plus and minus sign, “positive
reinforcement” is “positive” in the sense of being additive to the current state of
affairs, while “negative” reinforcement is subtractive. The labels “positive” and
“negative” are not evaluative (i.e., denoting good and bad; pleasant or unpleasant;
or liked and disliked). This misunderstanding has caused a great deal of confusion.
The terms “positive” and “negative” only describe the contextual change that
follows the behavior from the point of view of the observer. To determine whether
the behavior was reinforced or punished, we have to look at the effects of the
consequences over time. If behavior increases in rate, we speak of reinforcement, and
if its frequency decreases, we speak of punishment.
Positive Reinforcement. Positive reinforcement is the strengthening of behavior
due to a contingency that involves the presentation of particular consequences.
Again, it is important to note that reinforcement is defined by the changing
frequency of behavior, not by the presumed “pleasantness” or “unpleasantness” of
events. This is part of why behavioral principles gain power over standard common-
sense interpretations of behavior, because sometimes behavior may be reinforced
by events that are usually thought to be aversive (e.g., access to restraints as a
 6 Chapter 2

reinforcer, Rooker & Roscoe, 2005; chiding by a teacher as a reinforcer, Madsen,

Becker, Thomas, Koser, & Plager, 1968).
When issues of what is “pleasant” or “desirable” become confused with positive
reinforcement, detection of the functions of events become difficult, especially in
applied situations. As this book emphasizes, applied behavioral approaches are
grounded in an analysis based upon the function of behavior (Eifert, 1996; Follette,
Naugle, & Linnerooth, 2000; O’Brien, McGrath, & Haynes, 2003). For example,
clients often identify the particular behavioral patterns that are the sources of
distress. The behavioral patterns that seemingly generate the problems have often
persisted, sometimes for years. When what is “pleasant” or “desirable” becomes
confused with what is reinforcing, this circumstance seems to contradict principles
of behavior. Why doesn’t distress—commonly conceptualized as an aversive event—
lead to a reduction in the frequency of such behavior patterns? For example, suppose
a client seeks therapy because of the occurrence of angry outbursts that continue
despite their threat to the client’s marriage. Just because the client characterizes
these outbursts as distressing does not mean that they are not maintained by
reinforcement (e.g., an intense interaction with the partner follows each angry
outburst).

Analysis in terms of the three-term contingency

No spousal interaction < Behavior (angry outbursts) → Intense Interaction

Figure 2.2. Contingency analysis.

Even if they look negative or distressing themselves, these intense interactions

might be reinforcing. They might be preferred to being ignored; they might prove
that the partner cares; they might lead to a sense of closeness with an otherwise
distant partner; they might feel alive; they might lead to sex; and so on. Thus, the
angry outbursts may be positively reinforced by the subsequent intensive emotional
engagement, whether or not it looks positive or reinforcing.
Negative Reinforcement (Avoidance, Escape). While positive reinforcement involves
the presentation of events following behavior, negative reinforcement describes an
increase in the rate of behavior if that behavior functions to escape or avoid
particular events. Suppose a parent escapes from a child’s tantrums at the checkout
counter of a grocery store by buying the child candy. If the candy stops the tantrum,
and this termination makes giving in to the child’s demands in public situations
more likely in the future, the parents’ giving-in has been “negatively reinforced.” If
the term “negative” is thought of in the sense of “subtraction” it makes sense: the
candy subtracts the tantrum from the parent’s immediate environment.
When both persons are examined, a disturbing pattern can be created by
interlocking contingencies of this kind. If tantruming increases in frequency over the
Introduction to Principles of Behavior 7

next weeks or months because the candy was presented, the child’s tantruming can
be characterized as having been positively reinforced. Now caregiver and child have
entered a “vicious cycle,” “sick social cycle,” or “coercive process” in which stopping
the disruptive behavior reinforces patterns that actually escalate that behavior and,
in turn, makes the caregiver’s escape or avoidance even more likely (e.g., Patterson,
1982).
Because the functional definition and the common-sense approach to rein-
forcement often do not overlap, the identification of reinforcers is an integral and
important part of behavior analysis. Reinforcer identification is exceptionally
difficult when behavior is being maintained by negative reinforcement. In the case
of avoidance, the event that maintains the behavior is postponed or prevented. It may
rarely occur at all when the avoidance behavior is strong. A person’s depressed
behavior, for example, may serve to prevent angry interactions in the household
(Hops, Biglan, Sherman, & Arthur, 1987). The long-standing success of the behavior
pattern (i.e., the absence of angry interactions) may have made it impossible for the
person to describe the relation between depressed behavior and familial disputes.
By the time the person seeks therapy, it may seem that depressed behavior occurs
“for no good reason.” Consequently, from a current lack of identifiable maintaining
factors it does not follow that problematic behavior patterns arose spontaneously
or that they will be insensitive to situational changes.
Behaviors maintained by negative reinforcement are often less flexible than
those maintained by positive reinforcement. Repertoires leading to successful
avoidance or escape are both narrow and efficient. For one thing, aversive control
tends to create a situation in which only a few behaviors are strong. For example,
if something stops you from breathing, within seconds the only behaviors of interest
are those that might lead to removal of the blockage. Furthermore, there is not much
creativity and variability built into such actions. Likewise, if you now discovered that
a certain behavior completely prevented the next episode, you would repeat
whatever worked and not deviate or experiment much. If the negative reinforcement
contingency were suddenly stopped under these conditions, your previously
functional avoidance repertoire would become obsolete and may weaken eventu-
ally. Yet, you might continue to engage in the now non-functional avoidance
behavior for quite some time. Concurrent verbal descriptions of these events (e.g.,
“reason-giving”) may also serve to maintain behaviors that do not work. For these
reasons, behavior that is negatively reinforced by avoidance or escape can be
extraordinarily narrow, persistent, and out of contact with a more flexible,
situationally appropriate repertoire.
Positive Punishment. As shown in Table 2.1 and parallel to the principle of
reinforcement, the principle of punishment also describes a relation among
antecedents, behavior, and consequences, such that behavior decreases in frequency
following the reliable occurrence of certain events over time. The epitome of a
positive punisher is the commonsense example of a hot stove: Touching the stove
reliably leads to the presentation of scorching heat, and given these consequences,
touching the stove will become less likely.
 8 Chapter 2

Punishment procedures, by definition, do not increase the rate of any specific

target behavior. If an increase in the rate of a particular behavior is desired,
reinforcement must be operative. Moreover, while punishment suppresses behavior,
it is not clear what will occur in its place. That is one of the weaknesses of procedures
based on punishment, and for this reason Skinner and other behavioral psychologists
(e.g., Sidman, 1989) consistently have opposed its application (for a review of the
ethical and effective use of punishment procedures, see Van Houten, 1983). The
attractiveness of positive punishment often lies in its immediate effects: the
annoying behavior stops. However, although the punished behavior has decreased
in rate, the reinforcers that maintain the behavior are still effective. In this situation
“symptom substitution” may occur (i.e., other problematic behavior patterns leading
to the reinforcer may emerge). Thus, both pragmatics and ethics suggest that each
punishment procedure be accompanied by the reinforcement of socially appropriate
behavioral patterns. For example, parents have many demands upon their time.
Especially if an infant is in the home, older siblings may receive the most attention
when they engage in disruptive behavior. Punishment of a particularly disruptive
behavior may quickly decrease its rate. However, if no appropriate means to get
attention are established, it is likely that another disruptive behavior will emerge.
Contrary to other approaches which may find fault with the child’s need for
attention, the behavioral perspective respects the child’s need for attention as a given
and focuses on providing access to attention by socially appropriate means.
Negative Punishment (Penalty, Timeout). While positive punishment procedures
involve the presentation of an event, negative punishment procedures characterize
a decrease in the rate of behavior because it is followed by the contingent removal,
withdrawal, postponement, or termination of an event. Thus, a teenager may
temporarily lose access to the family car as a consequence of breaking curfew. If, over
time, incidences of breaking the curfew go down in frequency, then taking away car
privileges functions as a negative punisher. Another example of negative punish-
ment is the temporary withdrawal of access to social activities following behavior.
If restless and agitated behavior with siblings at the dinner table, followed by
temporary removal from the table to eat alone, leads to a decrease in the frequency
of the problem behavior over time, then this procedure is correctly described as
“time-out from family dinner.”
Extinction
Importantly, for an event to be characterized as a reinforcer, not only does
behavior have to increase in rate when consequences are made dependent upon the
occurrence of the behavior, but withdrawal of the consequences has to eventually
result in a decrease in the rate of behavior. “Extinction” is the term used for procedures
that involve the breaking of a reinforcement contingency. To break a contingency,
one may either withhold the consequences completely, or one may present those
events regardless of the occurrence of particular behaviors. The same holds for
extinction of punishment contingencies, where breaking the punishment contin-
gency results in a subsequent increase in the rate of previously punished behavior.
Introduction to Principles of Behavior 9

Withdrawal of access to reinforcers often leads to a temporary escalation of

behavior or increased variability (Morgan & Lee, 1996). The escalation is termed an
“extinction burst,” a common phenomenon when extinction procedures are in
effect, which is then followed by a gradual decline in the behavior. In the case of
social distress, for example, a client’s social group may meet initial expressions with
much sympathy, compassion, and understanding. There is a reinforcement contin-
gency in effect. Accordingly, distress reports may increase in frequency, especially
if the client does not have other repertoires to maintain social involvement. As time
passes and behavior problems persist, the community may become less sympathetic
and less attentive to the client’s reports of distress. As a function of the community’s
withholding of attention, the client’s reports may start escalating. After the burst and
with extinction still in effect, the reports of distress are expected to decrease in rate
over time.
Another effect sometimes seen during extinction is that other, previously
extinguished behaviors or repertoires may “resurge” (e.g., Epstein, 1985; Wilson &
Hayes, 1996). For example, alcohol-seeking may “resurge” (i.e., relapse may occur)
when other, non-drug-related behaviors are extinguished (Podlesnik, Jimenez-
Gomez, & Shahan, 2006). Like an extinction burst, resurgence needs to be expected
as a normal indication that extinction is occurring.
The consistency with which behavior is reinforced or punished determines the
persistence of the particular behavior under extinction conditions. If behavior
always has been followed by reinforcement, discontinuing reinforcement will
decrease the rate of behavior very quickly. In other words, the person quickly discerns
that conditions have changed and that the behavior is not effective any more. A
parent who consistently escaped from a child’s tantruming by “giving in” and buying
toys or candy, has arranged a continuous schedule of reinforcement. “Giving in”
reinforces the child’s behavior, i.e., it increases its rate over time. When extinction
is implemented, and the parent remains firm in his or her decision to ignore the
tantrum and refrain from buying candy, the child’s behavior is expected to cease
completely after only a few times at the checkout line. However, if behavior has been
reinforced only once in a while, the parent has inadvertently established an
intermittent reinforcement pattern. Under these circumstances, the behavior is
expected to persist for quite some time. Intermittent reinforcement thus produces
higher and more persistent rates. For this reason, extinction procedures should only
be implemented if accidental reinforcement will not occur.
Motivation
The effectiveness of the reinforcer or punisher is measured in terms of the
frequency of the behavior it supports, and the subject of motivation (also termed
“establishing operations”) addresses the ways by which the effectiveness of
reinforcers or punishers may be changed. Food deprivation, for example, changes the
effectiveness with which food supports behavior: While at the beginning of the day,
the items available in the lounge snack machine do not seem appealing, they may
occasion a great deal of behavior (e.g., obtaining the necessary monetary change,
10 Chapter 2

operating the machine, devouring the items) in the middle of the afternoon, after
missing lunch.
Rather than referring to a physical or internal state, “motivation” as used here
refers to the practical procedures with which reinforcer or punisher effectiveness may
be changed. Many drugs of abuse establish the effectiveness of social reinforcers:
Ingestion of d-amphetamine increases the probability that a person will forego
monetary reinforcers for monologues and instead speak to a same-sex confederate
in a different room via a microphone (Higgins, Hughes, & Bickel, 1989). Other drugs
decrease the effectiveness of punishers: Contrary to the popular perception of
cannabis as an “amotivational” drug, experimental work has consistently shown that
people under cannabis’s influence are more likely to engage in non-preferred,
repetitious, or strenuous tasks (Snycerski, Laraway, & Poling, 2000). One may
speculate that cannabis may also make aversive social interactions more tolerable.
Motivational effects are not limited to drugs but may also emerge from the
interaction of contingencies. For example, a basic research program that investigated
the effects of activity on eating patterns showed that high-rate exercise schedules
decrease the reinforcing effectiveness of food. This amotivational effect, in turn,
leads to an increased engagement in exercising to the point at which the organism
is close to starvation. Activity anorexia, empirically supported by animal models,
seems to emerge from bidirectional and reciprocal changes in the effectiveness of
reinforcers (for a review, see Pierce & Epling, 1991).
Other motivational effects arise from the interactions of verbal processes with
other behavior and will be covered later in this chapter under the topics of
“augmentals.”
Practical Considerations
There is no fast-and-easy rule for the identification of components of the three
or n-term contingency. Once the theoretically functional background is in place, the
investigation is guided entirely by pragmatics. Skinner, alluding to the Jamesian
conceptualization of a stream of consciousness, acknowledged that one might also
speak of behavior as continuous (Skinner, 1953). The disruption of the stream and
the formation of units of analysis that lend themselves to intervention are activities
by the therapist or observer. Units are pragmatic, flexible, and dependent upon the
therapist’s conceptualization and purpose. Whether the therapist picked the “right”
unit is an experimental question to be answered by its utility: if the intervention
shows the desired results, an appropriate unit was chosen. Often there is a variety
of levels of analysis one can choose. In the case of smoking, for example, the positive
reinforcer may be conceptualized as the stimulation of cholinergic receptors and its
physiological consequences. On another scale of analysis, smoking may be
conceptualized as being maintained by the avoidance of withdrawal effects, where
withdrawal effects function as a negative reinforcer. Alternatively, one may think of
smoking as being maintained by the access to guaranteed social reinforcers, i.e.,
access to the smoking group that gathers on porches and balconies at parties and that
provides an immediate commonality, smoking. In each case, the intervention
Introduction to Principles of Behavior 11

differs. While nicotine replacement and slow titration of the dose may lead to
abstinence in the first two examples, the last example requires an intervention that
includes a social skills training component. Reinforcers and punishers thus can be
conceptualized from a molecular to a molar scale, and so can behavior. As we shall
see, however, adding an analysis of language to behavioral thinking focuses the
analysis on a smaller set of likely possibilities, which is one of its sources of applied
utility.
Behaviors that may look topographically different may belong to the same
functional class. The opening of a door may be accomplished with many different
topographies, such as turning a knob, pressing a bar, pulling a handle, nudging with
one’s hip, or pushing with a foot. In all cases the opening of the door, given a
particular door type, will reinforce a variety of topographically dissimilar behaviors.
The reader will note that this issue occurs repeatedly throughout this book.
Behavioral patterns associated with post-traumatic stress disorder, for example, may
be topographically dissimilar (say, drinking, unsafe sex, or social withdrawal) and
yet all be maintained by the same avoidance or escape contingency (see also Hayes,
Wilson, Gifford, Follette, & Strosahl, 1996).
Additionally, the topography of the class of behaviors that leads to particular
consequences may vary within a lifetime. In a clinical example, a client’s
agoraphobic reaction may be conceptualized as a discrete behavioral event or as part
of a larger pattern of behavior maintained by avoidance that permeates the client’s
life. While the first interpretation suggests exposure treatment in multiple settings,
the second would call for the identification of reinforcers in the client’s life that
could be activated to establish or maintain alternative repertoires competing with
avoidance. This pragmatic conceptual flexibility distinguishes behavior analytic
from traditional approaches.
Respondent conditioning
Respondent or “classical” conditioning refers to learning through stimulus-
stimulus relations. The conceptualization of operant and respondent processes as
distinct behavioral processes has not found empirical support (Jenkins, 1977). In
general it seems to be more useful to think of these differences in terms of
procedures.
In applied settings, respondent conditioning is most likely to be invoked post
hoc to describe the acquisition of the psychological functions of a previously neutral
event. For example, just as contiguous pairings of neutral with aversive events, such
as shock, heat, or loud noise, can be used to establish events that support avoidance
behavior in the laboratory, so too we can imagine how such histories might lead to
aversive emotional responses in real life.
The reason that this chapter has focused almost entirely on operant procedures
is that in practice, operant procedures are usually particularly effective in both
establishing and undermining psychological functions of events, regardless of
whether they are respondent or operant. If an event has acquired its discriminative
or motivational function through respondent conditioning, this function can be
12 Chapter 2

modified through operant procedures (Allan, 1998; Engel, 1986a, 1986b). Expo-
sure-based procedures, for example, may reinforce approach behaviors and extin-
guish escape or avoidance responding; the function of the event in question changes,
regardless of whether it was established through respondent pairing or operant
discriminative processes.
Sometimes researchers may refer to any procedure that does not explicitly
arrange consequences as a respondent procedure. While the procedure may
resemble that of respondent conditioning, the repertoires involved are often operant
and have a long history of reinforcement that maintains them even in the absence
of explicitly arranged consequences (e.g., Leader, Barnes, & Smeets, 1996).
Basic behavioral change procedures
Reinforcing one behavior pattern while extinguishing another one is termed
“differential reinforcement.” Differential reinforcement accomplishes two goals: (1)
it can shape novel behavior patterns, and (2) it can result in flexible and contextually
appropriate repertoires.
Shaping. Shaping is the differential reinforcement of successive approximations
to a target performance. A prototypical example for a shaping procedure is learning
how to hit a baseball (Miller, 1997). First, a child will swing at a ball resting upon
a tee. Once he or she reliably hits the ball in this stationary position, the coach will
start throwing the ball to the child from a short distance with minimum speed. The
criteria for distance and speed are systematically increased depending on the child’s
performance, i.e., after the child has met the preceding criterion reliably. Another
example of shaping is children’s “hot-and-cold” game, where orientation and
approach to a target location are described as “hotter,” while movement away from
the target is labeled “colder.”
Shaping also occurs in the therapeutic context. When five independent clinical
psychologist raters judged interactions between Carl Rogers and a successful long-
term therapy client, they found significant evidence of differential reinforcement in
Rogers’ client-centered therapy (Truax, 1966). Thus, therapist empathy, warmth, and
acceptance systematically varied with the nature of the client’s verbalizations, such
that clarity of expression and positive self-statements were shaped over the duration
of therapy. As with teaching a child to hit a baseball, the criterion for reinforcement
might have been adjusted over time: While at first even the slightest indication of
clarity or positive statement might have met with approval, the bar was raised ever
so slightly throughout the therapeutic process. “Unconditional positive regard”
indeed denotes the absence of punishment procedures and allows for gentle
behavior change through differential reinforcement.
Discrimination training. While shaping implies a changing criterion for rein-
forcement, in discrimination training the delivery of reinforcers depends upon the
context in which behavior occurs. When the behavior occurs more often in one
situation than another, that behavior is “discriminated” or bound by context. Most,
if not all, human behavior of interest to psychologists falls into this category. People
implicitly assess whether others are behaving according to the social demands
Introduction to Principles of Behavior 13

placed upon their behavior and evaluate behavior as appropriate to gender, age,
specific roles, or socio-economic status, for example. It is important to note that this
evaluation in turn has been differentially reinforced by the social or cultural context
of the individual evaluator. Behavior can be contextual at several scales of analysis;
and society, or on a smaller scale the therapist within session, is the arbiter of
whether behavior is appropriate to the situation.
Social skills are prototypical examples of discriminated operants. Most
individuals have learned to identify “a good time” to ask a favor, or “a bad time” to
ask a boss for a promotion. “A good time” or “a bad time” may be substantiated by
a “gut feeling,” indicating that the relevant contextual variables that correlate with
social success are often complex and subtle and not easily captured by instructions
or rules. Because of the subtlety of the contextual cues, discrimination training
usually requires the presentation of multiple examples of situations in which the
behavior, or particular magnitudes of behavior, will be reinforced. These positive
exemplars randomly alternate with situations in which reinforcement is not
available. Multiple exemplar discrimination training is an integral component of
most social skills training interventions. Interestingly, discrimination training is
possible even in situations in which the coach or therapist cannot explicitly specify
the criteria for reinforcement.
Azrin and Hayes (1984), for example, used multiple exemplar training to teach
college-age men to identify whether similar-aged women were interested in their
conversations with other men. Only the female speaker was videotaped during one-
minute conversational interactions. Audio was omitted. During the baseline
condition, the men gave estimates that did not closely match the degree of interest
reported by the women. After a training condition consisting of feedback on the
accuracy of each judgment, the accuracy of the men’s estimates increased by over
50 percent and their actual social skills increased as well – even though the raters
could not verbally say what they were picking up while learning to discriminate
interest. Thus, discrimination training is especially useful when it is difficult to
describe the detailed features upon which reinforcement is contingent, and explicit
rules cannot be provided.
In addition to subtle features of events, relations among them may also
discriminate the occurrence of behavior. Rapid approach of an object, for example,
is one of the relations in the presence of which escape and avoidance behavior will
be reinforced. Here, the relation between the position of observer and the gradually
increasing occlusion of the observer’s visual field governs stepping to the side,
ducking, or catching the approaching object. As another example, the Doppler Effect
reliably discriminates orientation to approaching or retreating vehicles. Music also
provides countless examples of a relational context: Musical pieces can be
identified regardless of their particular arrangement; or musical performances can
be discriminated as “on” or “off key.”
Most of our behavior occurs in a context that changes over time. Accordingly,
regularities involved in these changes also constitute discriminanda: Facial changes
due to aging, for example, occur with such specificity and regularity that we are able
14 Chapter 2

to replicate them with computer simulations (“morphing” faces from young to old).
These regularities within change allow us to call particular changes “aging.” Thus,
the context to which behavior is bound does not have to be a discrete event. It can
be a pattern of events dispersed in time or a relation among events.
Conditional discriminations. As the previous discussion of units of analysis
showed, n-term contingencies can be nested at various scales. In conditional
discriminations, the occurrence of a discriminated operant depends on an addi-
tional situational cue. To the chagrin of customers, for example, automated
switchboards often allow the selection of a specific extension (a discriminated
operant) only after a lengthy announcement has been followed by the instruction,
“Select now.” Selecting an extension before hearing this announcement is extin-
guished and thus often accompanied by some emotional behavior on the caller’s
part. Similarly, using touch screen monitors in airports, at automobile repair stores,
or in banks involves the selection of specific options (itself a discriminated operant)
only after the monitor’s display has changed and added the relevant instructions.
Conditional discriminations can also involve relations among events. If
instructed to choose the larger of three objects, a golf ball, a baseball, or a basketball,
a person would choose the basketball over the baseball and the baseball over the
golf ball. If asked to pick the smaller of these items, the correct choice would be
different. Thus, how a relational choice is made depends or is conditional upon the
presentation of the particular instruction. The training of a conditional discrimina-
tion of relations among events must present situations in such a way that the
criterion for reinforcement is not tied to particular objects but to the relations among
them. In other words, what remains stable and does not vary across multiple
exemplars is the relationship of objects to each other based on their formal
properties. Children’s educational TV programs, such as Sesame Street, give
countless examples of multiple exemplar training that illustrate relations, such as
far-near, over-under, before-behind, or same-opposite. An extension of these
conditional discriminations of events based on formal properties to abstract
relations is at the heart of the modern behavior analytic account of language and
cognition.
A Behavior Analysis of Language and Cognition
The Problem of Language
The modern behavior analytic account of language and cognition has emerged
within the last decades and is built upon the principles described in the preceding
pages. None of the principles are challenged or negated, but the unit of analysis is
expanded, such that histories of conditional discrimination training relate synergis-
tically to form complex repertoires, affecting the function of other events. While
earlier attempts to provide an account of language and cognition did not emphasize
the relational nature of verbal behavior (e.g., Skinner, 1957), the modern account
explicitly asserts that behavior “usually labeled as verbal include[s] a referential
quality […] or generalized relations among arbitrary stimuli” (Chase & Danforth,
Introduction to Principles of Behavior 15

1991, p. 206). Thus, Skinner’s (1957) analysis alluded to but failed to capture the
“arbitrary, social or culturally determined relations among events in the world,
symbols, pictures, gestures, and sounds” (Chase & Danforth, 1991, p. 206) that are
the warp and weft of being verbal (see also Barnes-Holmes, Barnes-Holmes, &
Cullinan, 2000; Sidman, 1994).
To illustrate this point, suppose a history of simple discrimination training (i.e.,
saying “x” in the presence of object “x” and not in the presence of object “y”). Skinner
(1957) termed this behavior “tacting” and explained that “tacting” denotes that the
speaker contacts the environment and provides the listener with access to events that
previously only affected the speaker’s behavior. The listener reinforces the speaker’s
tacting with social approval or by behaving with respect to the tacted events. Tacting
is bound by antecedent events, but the consequences for tacting may vary.
According to Skinner, in its simplest case, the function of tacting is much like
that of an instrument safety crosscheck that occurs between pilot and co-pilot in an
airplane cockpit. Individuals are ascertaining that they are behaving with respect to
similar events. Tacting increases in complexity as one thinks of complex social
relations that occasion “crosschecks” with other members of the social group,
especially if these relations only hold in certain contexts. However, problems with
Skinner’s definition arise when we are able to train a parrot to say “x” in the presence
of object “x” and not in the presence of object “y” (e.g., Pepperberg, 2005). The
parrot’s behavior looks topographically as if it were verbal, but it does not have the
common functions of human verbal behavior—to repeat, the “arbitrary, social or
culturally determined relations among events in the world, symbols, pictures,
gestures, and sounds” (Chase & Danforth, 1991, p. 206). This difference may be
irrelevant to the parrot example, but it bears heavily on interventions in human
language acquisition, when the establishment of rote responses in the presence of
particular events may not lead to the sophisticated repertoire required for appropri-
ate verbal functioning. Then, what other training might be necessary?
Relational Frame Theory (RFT)
RFT (Hayes, Barnes-Holmes, & Roche, 2001), supported by more than 70
studies to date (Hayes, Luoma, Bond, Masuda, & Lillis, 2006), builds upon a
person’s history of conditional discrimination training and adds four features
capturing the function of human verbal behavior. These are arbitrary applicability,
mutual entailment, combinatorial entailment, and transformation of stimulus
functions.
Arbitrary applicability. There is substantial evidence that relational conditional
discriminations, described above, may not only be based on the formal properties
of events, but once the relation has been learned it can be applied to any situation
based on contextual cues to do so. On a daily basis children ideally encounter
thousands of training exemplars with feedback for these and other linguistic
relations (see Moerk, 1996, for a review). For example, conditional discrimination
training may lead to the identification of the larger object when presented with the
word, “larger.” After a prolonged conditional discrimination training history, a child
16 Chapter 2

will be able to tact the relations “larger” and “smaller,” even when presented with
other abstracted relations (e.g., the value of a nickel or “5” versus the value of a dime
or “10”). Similarly, people who tact the relation “between” in a spatial context can
also apply “between” to describe their political stance by noting that it is “between
liberal and conservative.” In other words, the relational nature of discriminations
comes under the control of contextual cues other than the formal properties of the
related events. The same relation applies, whether the context is concrete or abstract,
but in one case it is control by formal properties and in the other by the relational
context. The relation is arbitrarily applicable as a result (e.g., a nickel “is smaller
than” a dime only by social convention).
Mutual entailment. A person will be able to behave with respect to relations (e.g.,
large-small, thick-thin, before-behind, far-near, over-under, same-opposite), even if
only one of these relations is identified. Thus, a dime being larger than a nickel
implies the nickel being smaller than the dime. This is called “mutual entailment”
and is based on a particular history of reinforcement for behaving with respect to
such entailed relations (e.g., Barnes-Holmes, Barnes-Holmes, Roche, & Smeets,
2001; Berens & Hayes, in press).
Combinatorial entailment. If multiple relations are presented, a person will be
able to infer relations among these events. If, for example, one is told that A was the
same as B, and B was the opposite of C, implied is that A is the opposite of C, and
C is the opposite of A. Behaving with respect to the relationships among mutually
entailed relations is called “combinatorial entailment.” Again, regardless of the
mutually entailed relation given (same-opposite, near-far, here-there, etc.), an
overarching combinatorially entailed relation among common members is also
implied. The emergence of explicitly untrained relations relies on an extensive
training history that occurs as a matter of course in the rearing of typically developing
children. This training history involves the explicit training of simple and
conditional discriminations, and mutually entailed and combinatorial relations.
Transformation of stimulus functions. While a person may infer certain relations
among events based upon the history just described, the events themselves may have
a history of participating in operant relations, for example as punishers, reinforcers,
motivators, or antecedent events. These psychological functions spread in accor-
dance with mutually and combinatorially entailed relations: this is termed the
transformation of stimulus functions. To revert to our previous example: If A
functioned as a punisher, and a person were instructed that A was the same as B, and
B was the opposite of C, then RFT would expect that A and B functioned as punishers
while C remained neutral or functioned as a reinforcer if these stimuli were
presented contingent upon behavior (Whelan & Barnes-Holmes, 2004). If, for
example, a person was told that “bad,” a word that presumably sets the context for
social punishment contingencies, was the same as “böse,” and “böse” was the
opposite of “lieb,” being told that one was “lieb” would not occasion the same
reaction as being told that one was “böse.” Note that no further training would be
necessary for these functions to emerge.
Introduction to Principles of Behavior 17

Figure 2.3 Behavior occasioned by emergent relations.

The solid arrows of Figure 2.3 depict directly trained relations; the dashed
arrows depict mutually entailed relations; and the dotted double-headed arrow
depicts combinatorially entailed relations. This emergence of consequential
functions in accordance with the mutually and combinatorially entailed relations
is called “transformation of stimulus function.”
Crel’s and Cfunc’s. Human language does not consist of isolated instances of
utterances involving arbitrary applicability, mutual and combinatorial entailment,
and transformation of stimulus function. Instead, each topographical unit (e.g.,
sentences, paragraphs, or chapters) contains multiple nested entailed relations and
multiple possible functions, all of which differentially affect behavior. As in any
discriminated operant, context regulates these actions. RFT distinguishes between
two kinds of context: the relational context (Crel) and the functional context (Cfunc)
[the “C” stands for “context” in both abbreviations]. In the above example, the word
“same” constitutes the relational context for a specific kind of relational response
in the presence of “bad” and “böse” (the “relata”), while the word “opposite” sets the
occasion for yet another type of relational response in the presence of “böse” and
“lieb.” Here, “same” and “opposite” function as Crel’s. On the other hand, Cfunc’s refer
to the context occasioning the transformation of behavioral functions in accordance
with the entailed relations. Each event may have a history of multiple behavioral
functions, and Cfunc’s are the contexts that govern the transformation. For example,
told that “lieb” is also the same as “dear,” being called “lieb” may reinforce behavior
in a familial or romantic context but punish it in a professional context; or hearing
the word may set the occasion for approach in the familial context and for escape
or avoidance in the professional one. Cfunc’s would be necessary to disambiguate the
behavioral functions.
Crel’s and Cfunc’s are originally established directly, through a history of
discrimination training; but as relational operants emerge they can be established
18 Chapter 2

indirectly, through entailed relations. Both Crel’s and Cfunc’s can be identified in the
context of larger functional verbal units, such as found in instructions (O’Hara,
Barnes-Holmes, Roche, & Smeets, 2004), metaphors (Stewart & Barnes-Holmes,
2001), analogies (for a review see Stewart & Barnes-Holmes, 2004), logic, and
problem-solving. Crel’s and Cfunc’s locally regulate the continuously ongoing stream
of behavior and become integrated with observed, sophisticated verbal repertoires.
They are not limited to particular words or phrases, however. Tone of voice, pauses,
gestures, whole conversations, or aspects of previous interactions with an individual
can all function as Crel’s and Cfunc’s for later interactions (“inside jokes” would be an
example for the latter). According to RFT, without such abstracted contextual
control, the automaticity of human language processes would be impossible.
The interaction of cognitions with other behavior. The processes of mutual and
combinatorial entailment and transformation of stimulus function could be termed
“automatic inferences,” but the RFT account greatly improves upon traditional
interpretations by providing an experimentally supported account of how cognitions
and behavior may interact. Through transformation of stimulus function, RFT
accounts for how events that have never been encountered obtain psychological
meaning.
Suppose a person experiences a panic attack in a small local store and
subsequently avoids leaving the home. During treatment, the client is asked to rank
(Crel) potential anxiety-provoking (Cfunc) locations, and paradoxically ranks the local
store in which the first panic attack was experienced as less anxiety-provoking than
Harrod’s department store in London, which in fact the person has never visited. To
the extent that the person has a history of describing Harrod’s as “larger than” or
“busier than” the local store, these multiple comparisons may come to bear upon
the ranking of anxiety-provoking locations. “Ranking” (Crel) provided the common
context for application of the comparative relations. As a result, although never
visited, Harrod’s may occasion more avoidance behavior than the local store.
To illustrate how humans acquire such a complex repertoire of behaving with
respect to entailed relations virtually “automatically,” let us use imitation as a
familiar example. Children’s imitative behavior is shaped by first explicitly
arranging reinforcement for “doing what the model does” in specific situations
(clapping hands, stomping feet, tapping fingers, making faces). As the number of
exemplars increases, the correspondence between the child’s and the model’s
behavior starts to guide the child’s behavior, for it is this relation that remains
constant across opportunities for reinforcement.
As with imitative behavior, the relations that have frequently provided the
criteria for reinforcement in conditional discrimination training situations come to
govern behavior without additional environmental support, i.e., no particular
prompts have to be given for the behavior to occur. However, the behavior continues
to contact social and other reinforcers. The difference between imitative repertoires
and the ones just described in the “Harrod’s” example is that the criterion for
reinforcement (“match the model”) can be described in fewer words than the history
Introduction to Principles of Behavior 19

of conditional discrimination training that leads to combinatorial entailment and

transformation of stimulus function. Note that the ease of description by the
psychologist is not an indicator for the ease with which these repertoires are acquired
and performed. Describing the contingencies that operate upon “walking,” for
example, is a quite daunting task, yet walking is acquired by almost everybody.
RFT does not question or alter the history of conditional discrimination training
necessary for the acquisition of complex repertoires; instead, it emphasizes that the
criteria for reinforcement continue to govern behavior in the absence of explicitly
arranged contingencies as the contexts influencing such responding (Crel’s and
Cfunc’s) become increasingly abstract. It posits that people’s extensive verbal histories
of reinforcement lead to an almost automatic application of mutual and combina-
torial entailment to any encountered situation. The extensive history of trained
conditional discriminations among relations bears on all verbal utterances, every-
thing a person says or thinks. RFT is a logical extension of explicitly arranged
conditional discrimination training and accounts for many verbal phenomena, such
as generativity, figurative language in general, and vocabulary spurts in childhood
in particular (Barnes-Holmes, Barnes-Holmes, & Cullinan, 2001; Stewart, Barnes-
Holmes, Hayes, & Lipkens, 2001).
Rule-governed behavior
Because multiple contingencies operate upon behavior, the influence of
emergent stimulus functions partly depends on the availability of alternative
reinforcers that support behaviors competing with those functions. This interplay
between multiple contingencies (such as that shown in the bottom panel of Figure
1) is studied under the topic of rule-governed behavior. Note that any instances of
verbal behavior analyzed under this and the following rubrics involve arbitrary
applicability, mutual entailment, combinatorial entailment, and transformation of
stimulus function. These characteristics may not be independently verified or
discussed but are implied by the scope of the analysis. As outlined above,
transformation of stimulus function in accordance with entailed relations describes
how events that have not been experienced come to affect behavior.
Behavior that is bound or discriminated by a particular verbal context is called
“rule-governed.” A person’s typing repertoire quickly illustrates the difference
between rule-governed and contingency-shaped behavior: When typing is first
learned, a person may recite each letter and its respective keyboard position and
tentatively move the respective fingers on the keyboard to produce the desired result.
“Same” and “before/after” relations (O’Hara, Barnes-Holmes, Roche, & Smeets,
2004), as to what letter corresponds to the movement of which finger and to what
effect, may guide the initial acquisition. As a person’s repertoire becomes more
fluent, it becomes less dependent on particular rules. Consequences in the forms
of typos on the screen or annoying beeps by the computer shape proficiency, until
thinking about which finger to move actually interrupts the now fluent performance.
Behavior that was first entirely rule-governed gradually becomes more and more
20 Chapter 2

affected by the direct contingencies. Most skills are acquired in this fashion: An
expert chef, for example, may not be able to state the rule by which he or she decides
upon the final touches of a meal, although cooking was initially verbally instructed.
Ideally, the degree of verbal control wanes.
Augmenting. An augmental is a particular type of instruction that changes the
reinforcing or punishing effectiveness of an event through the transformation of
stimulus functions according to entailed relations. Formative augmentals are those
that establish events as reinforcers or punishers (e.g., being told that one is able to
exchange a paper ticket for a promotional item; see also Whelan & Barnes-Holmes,
2004), while motivative augmentals increase or decrease the effectiveness of already
established events. Again, whether the event actually comes to function as a
reinforcer or a punisher is only evident by the effect on the behavior over time. To
the degree that the augmental facilitates contact with the actual reinforcing or
punishing contingencies, it might facilitate tracking which we discuss below.
Pliance. Rule-following can be conceptualized as two additional categories,
pliance and tracking. Similarly to the earlier example of imitation, where “doing
what the model does” is the occasion for reinforcement, in pliance the correspon-
dence between a person’s behavior and what that person has been told to do provides
the context in which pliance is reinforced (Zettle & Hayes, 1982). Because only a
social/verbal community could discriminate and differentially reinforce the rule-
behavior correspondence, the reinforcers for pliance are always socially mediated,
but it is the unit that is reinforced, not the type of reinforcer, that defines pliance.
Pliance occurs across a variety of situations and is maintained by real or perceived
social demands and not by the demands of the actual situation. Milgram’s (1963)
classic study of obedience, where nearly two-thirds of the subjects complied with
instructions to administer increasing levels of shock to a confederate despite signs
of the presumed victim’s distress, illustrates how pliance can overwhelm appropriate
social behavior. Although pliance is established by a rule-giver who reinforces
correspondence with a particular rule or request, the verbal processes described
above may extend pliance to self-generated demands that are maintained by verbally
constructed social communities (e.g., “my mother is watching me from heaven”). For
these reasons pliance can be greatly over-extended, and people may rigidly fulfill
these demands even when they run counter to the actual contingencies. Hence,
pliance can be a major source of clinical distress.
Tracking. Tracking is rule-governed behavior maintained by contingencies
(Zettle & Hayes, 1982), that are contacted based upon the formal properties of
behavior regulated by a rule. The consequences may be of any kind, including social
consequences – what defines it as tracking is that the behavior itself (not the fact
that the behavior was occasioned by a rule) produces the consequence. Closely
following written driving directions, for example, may be necessary to find one’s way
for a first visit to a friend’s home. Random driving may have led to the same result
–the arrival at the friend’s home is determined solely by the formal properties of the
driving, however it was produced. In that sense, the reinforcers for tracking are not
arbitrary.
Introduction to Principles of Behavior 21

It is impossible to determine whether a particular instance of following a rule

constitutes pliance or tracking, because pliance and tracking are antecedent-
behavior-consequent relations that occur over time. Generally speaking, pliance will
persist regardless of the ultimate utility of the behavior because it involves an
extensive history of reinforcement for rule-following per se that may compete with
other available consequences. Conversely, tracking repertoires will adjust and be
further refined by the situation at hand. Establishing repertoires may first require
pliance; as the individual comes in contact with natural contingencies, tracking
repertoires emerge. Such is the case when a couple engages in communication skill
training during marriage counseling. The first few instances of following the rules
may be maintained by the therapist’s approval or praise. With practice, the couple
may be able to use the skills dependent upon their impact in actual situations; the
instructions from therapy have functioned to track the social utility of these
communication skills. Similarly a young child may learn not to go into the street
to avoid parental punishment; later, not going into the street may be a matter of
avoiding cars.
In the absence of a skill deficit, excessive pliance may be implicated when it
appears that an individual’s verbal behavior is the only obstacle to more effective
action in a particular situation. “Same” and “before/after” relations governing
behavior maintained by avoidance are especially problematic, for—as long as the
aversive event does not occur, regardless of its likelihood—the rule-following is
reinforced. Additionally, if a person’s verbal community has slowly shaped a
person’s description of his or her own performance without particular regard to the
relevant behavior-environment contingencies, the performance itself is more likely
to be consistent with the verbal description and less likely to track changes in the
actually prevailing environmental contingencies (Catania, Matthews, & Shimoff,
1982). Thus, even when behavior looks as if it were tracking contingencies, only a
change in those contingencies reveals whether the person’s behavior is indeed
flexible and adjusts to current demands (Hayes, Brownstein, Haas, & Greenway,
1986). Because of the risk of producing rigid rule-following given the social demand
characteristics of the therapeutic setting, Hayes et al. (1986) advise considerable
care when using instructions. As a general guideline, general, experiential, or
metaphorical rules support more flexibility.
Other behavior of clinical interest
Manding. While being vague may be advisable when providing instructions,
manding benefits from conciseness and specificity. If a particular verbal utterance
is always followed by the same listener-initiated consequences (Skinner, 1957) and,
to be consistent with the modern account, participates in multiple “arbitrary, social
or culturally determined relations” (Chase & Danforth, 1991, p. 206), the speaker
is said to engage in “manding.” Manding differs from “demanding” or “requesting”
because it is defined by its function rather than syntactic properties of the utterance.
Thus, if saying, “the food is bland,” is always followed by an array of salts and spices
on the table, the person is effectively manding these consequences. Like tacting,
22 Chapter 2

manding for novel events (as transformation of stimulus function) also occurs in
accordance with a history of conditional discrimination training involving arbitrary
and entailed relations (Murphy, Barnes-Holmes, & Barnes-Holmes, 2005). While
Skinner posited a direct history of reinforcement for manding (but see “magical
mands,” Skinner, 1957, p. 48), the modern account allows for emergence through
indirect and relational processes.
Because many individuals have difficulty expressing their needs, deficient
mand repertoires are often therapeutic targets. Especially in couple therapy, one may
observe one partner who consistently mands with statements that are not identifi-
able as demands or requests, and the other partner who displays excessive pliance.
Under these conditions, the latter partner may complain of too heavy demands,
while the former partner reports never asking for anything. Generally, non-specific
and imprecise mands can be a source of much interpersonal tension.
Like all verbal operants, the frequency of a particular mand depends upon its
effectiveness in producing consequences. If concise manding is occurring at a low
rate, it may have been extinguished or punished. Assertiveness training is partially
targeted at helping individuals to mand more effectively.
Tacting. As described earlier, Skinner’s (1957) conceptualization of “tacting” as
verbal behavior governed by a non-verbal event and reinforced by the social
community does not explicitly reflect the relational aspects of human language
studied by RFT. However, when “tacting” is considered more broadly (i.e., when a
behavioral history of arbitrary applicability, mutual and combinatorial entailment,
and transformation of stimulus function is assumed and behavior with respect to
emergent—rather than directly trained—stimulus functions becomes possible), then
the category of “tacting” is also one of clinical utility.
Tacting is verbal behavior bound by a specific non-verbal context that remains
invariable (e.g., saying “table” in the presence of any table). The function of tacting
may be to direct the listener to particular aspects of the context or of the speaker’s
behavior, to identify particular features of the situation, or to describe a situation
that is inaccessible to the listener. The reinforcement contingencies for tacting
involve nonspecific approval (e.g., praise, confirmation, or gratitude) for the accurate
identification of particular events.
The correspondence between the described and the actual state of affairs is of
special interest when events are inaccessible to the listener, or private. Given that
inaccessibility, how do speakers learn to reliably report? Skinner (1945) suggested
that correspondence is likely to be reinforced when the listener can rely on collateral
events, collateral behavior, a history with similar events under accessible conditions,
or metaphorical speech to determine the accuracy of the report (see also Stewart,
Barnes-Holmes, Hayes, & Lipkens, 2001).
Private events are often accompanied by public collateral events, as when
reports of nausea come with pallor. Collateral behavior may be holding one’s
stomach or vomiting. Both types of collaterals may make it more likely that a
caregiver reinforces a child’s report of nausea by keeping him or her home from
Introduction to Principles of Behavior 23

school, thereby teaching the child to tact “nausea.” Later, the history of speaking of
nausea with public collaterals may lead to reliable tacting when these collaterals are
not present. Alternatively, for more sophisticated speakers, relations applicable to
different contexts may be used to describe one’s ailment, e.g., “my stomach is
churning.” Importantly, private events participate in the relational processes
described earlier (see also DeGrandpre, Bickel, & Higgins, 1992), suggesting that
private events may be able to take on novel and emergent functions.
The verbal community shapes the tacting of private events in the same fashion
as the tacting of public events. Note that the private-public distinction is defined
by the observer’s accessibility to the relations that occasion the tact; public and
private events are not different in kind or differentially amenable to scientific study.
In the therapeutic situation, the client’s current out-of-session context and history
constitute private events to which the therapist does not have access.
The interesting implication of tacting, as conceptualized above, is that an
individual’s skill of reliably tacting private events depends largely upon the practices
of the social community. Social milieus that fail to establish correspondence
between the reporting of private experiences and the social interpretation of the
actual events have been characterized as invalidating: “An invalidating environment
is one in which communication of private experience is met by erratic, inappropri-
ate, and extreme responses” (Linehan, 1993, p. 49). In the absence of an effective
tact repertoire for private experiences, individuals may present with flat or
inappropriate affect, or they may escalate as responding shifts from tacting an event
to manding for the audience to affirm the observation. Thus, an individual’s social
audience can unduly influence the reliability of tacting by imposing a threat of
punishment, an opportunity to gain favors, or by reinforcing a particular report over
another, regardless of correspondence (see Loftus, 2003, for a review of audience
effects on memory). From this view, exaggerating, lying, or escalating are behaviors
warranted by a client’s history and current circumstances.
Conclusion
Behavior analysis (for a review see Baum, 1994; for philosophy, see Chiesa,
1994) is an evolving approach that involves verbally tacting antecedent-behavior-
consequence relations. This job has become more complex with an appreciation of
the important role of relational operants and rule-governance, but it is no less a form
of behavior analysis.
All interventions in this book build upon the basic principles reviewed in this
chapter. One can use them directly to facilitate contact with the larger antecedent-
behavior-consequence patterns affecting clients’ lives in order to relieve suffering
and empower a more effective style of living. But applying these principles often is
helped by middle level constructs that economically describe and integrate a wide
range of behavioral concepts and their extension into the behavior analysis of
language and cognition. These facilitative middle level constructs will be covered
in the next chapter.
24 Chapter 2

References
Allan, R. W. (1998). Operant-respondent interactions. In W. O’Donohue (Ed.),
Learning and behavior therapy (pp. 146-168). Needham Heights, MA: Allyn &
Bacon.
Azrin, R. D., & Hayes, S. C. (1984). The discrimination of interest within a
heterosexual interaction: Training, generalization, and effects on social skills.
Behavior Therapy, 15, 173-184.
Barnes-Holmes, D., Barnes-Holmes, Y., & Cullinan, V. (2000). Relational frame
theory and Skinner’s Verbal Behavior: A possible synthesis. The Behavior Analyst,
23, 69-84.
Barnes-Holmes, Y., Barnes-Holmes, D., & Cullinan, V. (2001). Education. In S. C.
Hayes, D. Barnes-Holmes, & B. Roche (Eds.), Relational frame theory: A post-
Skinnerian account of human language (pp. 181-195). New York: Kluwer Aca-
demic/Plenum Publishers.
Barnes-Holmes, Y., Barnes-Holmes, D., Roche, B., & Smeets, P. (2001). Exemplar
training and a derived transformation of function in accordance with symmetry.
The Psychological Record, 51, 287-308.
Baum, W. M. (1994). Understanding behaviorism: Science, behavior, and culture. New
York: Harper Collins.
Berens, N. M., & Hayes, S. C. (in press). Arbitrarily applicable comparative relations:
Experimental evidence for a relational operant. Journal of Applied Behavior
Analysis.
Catania, A. C., Matthews, B. A., & Shimoff, E. (1982). Instructed versus shaped
human verbal behavior: Interactions with nonverbal responding. Journal of the
Experimental Analysis of Behavior, 38, 233-248.
Chase, P. N., & Danforth, J. S. (1991). The role of rules in concept learning. In L.
J. Hayes & P. N. Chase (Eds.), Dialogues on verbal behavior (pp. 205-225). Reno,
NV: Context Press.
Chiesa, M. (1994). Radical behaviorism: The philosophy and the science. Boston: Authors
Cooperative, Inc.
DeGrandpre, R. J., Bickel, W. K., & Higgins, S. T. (1992). Emergent equivalence
relations between interoceptive (drug) and exteroceptive (visual) stimuli.
Journal of the Experimental Analysis of Behavior, 58, 9-18.
Eifert, G. H. (1996). More theory-driven and less diagnosis-based behavior therapy.
Journal of Behavior Therapy and Experimental Psychiatry, 27, 75-86.
Engel, B. T. (1986a). An essay on the circulation as behavior. Behavioral and Brain
Sciences, 9, 285-295.
Engel, B. T. (1986b). If it looks like a duck, walks like a duck, and quacks like a duck,
it’s a duck: Neurally mediated responses of the circulation are behavior.
Behavioral and Brain Sciences, 9, 307-318.
Epstein, R. (1985). Extinction-induced resurgence: Preliminary investigations and
possible implications. Psychological Record, 35, 143-153.
Introduction to Principles of Behavior 25

Follette, W. C., Naugle, A. E., & Linnerooth, P. J. N. (2000). Functional alternatives

to traditional assessment and diagnosis. In M. J. Dougher (Ed.), Clinical behavior
analysis (pp. 99-125). Reno, NV: Context Press.
Hayes, S. C., Barnes-Holmes, D., & Roche, B. T. (Eds.). (2001). Relational frame theory:
A post-Skinnerian account of human language and cognition. New York: Plenum.
Hayes, S. C., Brownstein, A. J., Haas, J. R., & Greenway, D. E. (1986). Instructions,
multiple schedules, and extinction: Distinguishing rule-governed from sched-
ule-controlled behavior. Journal of the Experimental Analysis of Behavior, 46, 137-
147.
Hayes, S. C., Hayes, L. J., & Reese, H. W. (1988). Finding the philosophical core:
A review of Stephen C. Pepper’s World Hypotheses. Journal of the Experimental
Analysis of Behavior, 50, 97-111.
Hayes, S. C., Luoma, J. B., Bond, F. W., Masuda, A., & Lillis, J. (2006). Acceptance
and commitment therapy: Model, processes and outcomes. Behaviour Research
and Therapy, 44, 1-25.
Hayes, S. C., Wilson, K. G., Gifford, E. V., Follette, V. M., & Strosahl, K. D. (1996).
Emotional avoidance and behavioral disorders: A functional dimensional
approach to diagnosis and treatment. Journal of Consulting and Clinical Psychology,
64, 1152-1168.
Higgins, S. T., Hughes, J. R., & Bickel, W. K. (1989). Effects of d-amphetamine on
choice of social versus monetary reinforcement: A discrete-trial test. Pharmacol-
ogy, Biochemistry and Behavior, 34, 297-301.
Hineline, P. N. (1980). The language of behavior analysis: Its community, its
functions, and its limitations. Behaviorism, 8, 67-86.
Hops, H., Biglan, A., Sherman, L., & Arthur, J. (1987). Home observations of family
interactions of depressed women. Journal of Consulting & Community Psychology,
55, 341-346.
Jenkins, H. M. (1977). Sensitivity of different response systems to stimulus-
reinforcer and response-reinforcer relations. In H. Davis & H. M. B. Hurwitz
(Eds.), Operant-Pavlovian interactions (pp. 47-66). Mahwah, NJ: Lawrence Erlbaum
Associates, Inc.
Leader, G., Barnes, D., & Smeets, P. M. (1996). Establishing equivalence relations
using a respondent-type training procedure. The Psychological Record, 46, 685-706.
Linehan, M. M. (1993). Cognitive-behavioral treatment of borderline personality disorder.
New York: Guilford.
Loftus, E. F. (2003). Make-believe memories. American Psychologist, 58, 867-873.
Madsen, C. H., Becker, W. C., Thomas, D. R., Koser, L., & Plager, E. (1968). An
analysis of the reinforcing function of “sit-down” commands. In R. K. Parker
(Ed.), Readings in educational psychology (pp. 265-278). Boston: Allyn & Bacon.
Milgram, S. (1963). Behavioral study of obedience. Journal of Abnormal and Social
Psychology, 67, 371-378.
Miller, L. K. (1997). Principles of everyday behavior analysis (3rd ed.). New York: Brooks/
Cole Publishing Company.
26 Chapter 2

Moerk, E. L. (1996). Input and language learning processes in first language

acquisition. In H. W. Reese (Vol. Ed.), Advances in child development and behavior
(Vol. 26, pp. 181-228). New York: Academic Press.
Morgan, D. L., & Lee, K. (1996). Extinction-induced response variability in humans.
The Psychological Record, 46, 146-160.
Murphy, C., Barnes-Holmes, D., & Barnes-Holmes, Y. (2005). Derived manding in
children with autism: Synthesizing Skinner’s Verbal Behavior with relational
frame theory. Journal of Applied Behavior Analysis, 38, 445-462.
O’Brien, W. H., McGrath, J. J., & Haynes, S. N. (2003). Assessment of psychopathol-
ogy with behavioral approaches. In J. R. Graham & J. A. Nagleri (Eds.), Handbook
of psychology: Vol. 10. Assessment psychology (pp. 509-529). New York: John Wiley
& Sons, Inc.
O’Hara, D., Barnes-Holmes, D., Roche, B., & Smeets, P. (2004). Derived relational
networks and control by novel instructions: A possible model of generative
verbal responding. The Psychological Record, 54, 437-460.
Patterson, G. R. (1982). Coercive family process. Eugene, OR: Castalia Publishing Co.
Pepperberg, I. M. (2005). An avian perspective on language evolution: Implications
of simultaneous development of vocal and physical object combinations by a
grey parrot (psittacus erithacus). In M. Tallerman (Ed.), Language origins:
Perspectives on evolution (pp. 239-261). Oxford: Oxford University Press.
Pierce, W. D., & Epling, W. F. (1991). Activity anorexia: An animal model and
theory of human self-starvation. In A. Boulton, G. Baker, & M. Martin-Iverson
(Eds.), Neuromethods: Vol. 18. Animal models in psychiatry I (pp. 267-312). Clifton,
NJ: Humana Press.
Podlesnik, C. A., Jimenez-Gomez, C., & Shahan, T. A. (2006). Resurgence of alcohol
seeking produced by discontinuing non-drug reinforcement as an animal model
of drug relapse. Behavioural Pharmacology, 17, 369-374.
Rooker, G. W., & Roscoe, E. M. (2005). Functional analysis of self-injurious behavior
and its relation to self-restraint. Journal of Applied Behavior Analysis, 38, 537-542.
Sidman, M. (1989). Coercion and its fallout. Boston: Authors Cooperative, Inc.
Sidman, M. (1994). Equivalence relations and behavior: A research story. Boston: Authors
Cooperative, Inc.
Skinner, B. F. (1938). The behavior of organisms. New York: Appleton-Century-Crofts.
Skinner, B. F. (1945). The operational analysis of psychological terms. Psychological
Review, 52, 270-277.
Skinner, B. F. (1953). Science and human behavior. New York: Macmillan.
Skinner, B. F. (1957). Verbal behavior. Englewood Cliffs, NJ: Prentice Hall, Inc.
Snycerski, S., Laraway, S., & Poling, A. (2000). Basic research with humans. In A.
Poling & T. Byrne (Eds.), Introduction to behavioral pharmacology (pp. 111-139).
Reno, NV: Context Press.
Stewart, I., & Barnes-Holmes, D. (2001). Understanding metaphor: A relational
frame perspective. Behavior Analyst, 24, 191-199.
 Introduction to Principles of Behavior 27

Stewart, I., & Barnes-Holmes, D. (2004). Relational frame theory and analogical
reasoning: Empirical investigation. International Journal of Psychology & Psycho-
logical Therapy, 4, 241-262.
Stewart, I., Barnes-Holmes, D., Hayes, S. C., & Lipkens, R. (2001). Relations among
relations: Analogies, metaphors, and stories. In S. C. Hayes, D. Barnes-Holmes,
& B. Roche (Eds.), Relational frame theory: A post-Skinnerian account of human
language and cognition (pp. 73-86). New York: Kluwer Academic/Plenum
Publishers.
Truax, C. B. (1966). Reinforcement and non-reinforcement in Rogerian psycho-
therapy. Journal of Abnormal Psychology, 71, 1-9.
Van Houten, R. (1983). Punishment: From the animal laboratory to the applied
setting. In S. Axelrod & J. Apsche (Eds.), The effects of punishment on human behavior
(pp. 13-42). New York: Academic Press.
Whelan, R., & Barnes-Holmes, D. (2004). The transformation of consequential
functions in accordance with the relational frames of same and opposite. Journal
of the Experimental Analysis of Behavior, 82, 177-196.
Wilson, K. G., & Hayes, S. C. (1996). Resurgence of derived stimulus relations.
Journal of the Experimental Analysis of Behavior, 66, 267-281.
Zettle, R., & Hayes, S. C. (1982). Rule-governed behavior: A potential theoretical
framework for cognitive-behavioral therapy. Advances in Cognitive-Behavioral
Research and Therapy, 1, 73-118.

View publication stats