Vous êtes sur la page 1sur 16

The Behavior Analyst 1993, 16, 191-206 No.

2 (Fall)

Establishing Operations
Jack Michael
Western Michigan University
The first two books on behavior analysis (Skinner, 1938; Keller & Schoenfeld, 1950) had chapter-length
coverage of motivation. The next generation of texts also had chapters on the topic, but by the late 1960s
it was no longer being given much treatment in the behavior-analytic literature. The present failure to
deal with the topic leaves a gap in our understanding of operant functional relations. A partial solution
is to reintroduce the concept of the establishing operation, defined as an environmental event, operation,
or stimulus condition that affects an organism by momentarily altering (a) the reinforcing effectiveness
of other events and (b) the frequency of occurrence of that part of the organism's repertoire relevant to
those events as consequences. Discriminative and motivative variables can be distinguished as follows:
The former are related to the differential availability of an effective form ofreinforcement given a particular
type of behavior; the latter are related to the differential reinforcing effectiveness ofenvironmental events.
An important distinction can also be made between unconditioned establishing operations (UEOs), such
as food deprivation and painful stimulation, and conditioned establishing operations (CEOs) that depend
on the learning history of the organism. One type of CEO is a stimulus that has simply been paired with
a UEO and as a result may take on some of the motivative properties of that UEO. The warning stimulus
in avoidance procedures is another important type of CEO referred to as reflexive because it establishes
its own termination as a form of reinforcement and evokes the behavior that has accomplished such
termination. Another CEO is closely related to the concept of conditional conditioned reinforcement and
is referred to as a transitive CEO, because it establishes some other stimulus as a form of effective
reinforcement and evokes the behavior that has produced that other stimulus. The multiple control of
human behavior is very common, and is often quite complex. An understanding ofunlearned and learned
establishing operations can contribute to our ability to identify and control the various components of
such multiple determination.
Key words: establishing operations, motivation, multiple control

In commonsense psychology, what a sidered a part of the topic of motivation.

person does is generally thought to be a To some extent this replacement is rea-
function of two broad factors, knowledge sonable. With the discovery of the role
and motivation. For any particular be- of reinforcement in the maintenance of
havior to occur (except for "involuntary" behavior-schedules of intermittent re-
acts such as reflexes), the behaver must inforcement-many examples of insuf-
know how and must also want to do it. ficient motivation could be better inter-
A good deal of traditional psychological preted as examples of insufficient ongoing
theory concerns the different kinds of reinforcement. The replacement was also
wants and the way they interact with oth- attractive because the more common
er mental functions; much ofapplied psy- motivational terms-wants, needs,
chology is concerned with getting people drives, motives-usually referred to in-
to do things that they know how to do ner entities whose existence and essential
but don't want to do. Motivation seems features were inferred from the very be-
to be an important topic, yet the basic havior that they were supposed to ex-
notion plays only a small role in the ap- plain.
proach currently referred to as behavior Reinforcement history is not, howev-
analysis. er, a complete replacement for motiva-
In applied behavior analysis or behav- tive functional relations. Skinner (1938,
ior modification, the concept of rein- chap. 9 and 10, 1953, chap. 9) clearly
forcement seems to have taken over much distinguishes deprivation and satiation
of the subject matter that was once con- from other kinds of environmental vari-
ables and relates these operations to the
Address correspondence concerning this article traditional concept of drive, as did Keller
to the author, Psychology Department, Western and Schoenfeld (1950, chap. 9). Skinner's
Michigan University, Kalamazoo, MI 49008-5052. treatment of aversive stimulation (e.g.,
1953, chap. 11) is very similar to his with behavioral effects similar to those
treatment of deprivation, and Keller and of deprivation and aversive stimulation,
Schoenfeld classify aversive stimulation but that cannot be easily classified as ei-
as one of the drives (1950, chap. 9). Later, ther, are likely to be ignored or misclas-
in his treatment of verbal behavior (1957, sified (usually as discriminative stimuli).
pp. 28-33,212-214), Skinner again iden- What follows is an attempt to provide a
tifies deprivation and aversive stimula- more thorough and systematic approach
tion as independent variables that are than usually appears, and one that cor-
quite different in function from rein- rects this latter difficulty.
forcement and stimulus control.'
Subsequent behavioral texts at first
continued to provide a separate chapter Establishing Operation Defined in
on deprivation (e.g., Holland & Skinner, Terms of Two Features
1961; Lundin, 1961, 1969; Millenson, An establishing operation-the term
1967; Millenson & Leslie, 1979), but was first used by Keller and Schoenfeld
more recent texts have almost dropped (1950) and later by Millenson (1967)-
the topic (e.g., Catania, 1979, 1984; Fan- is an environmental event, operation, or
tino & Logan, 1979; Mazur, 1986, 1990; stimulus condition that affects an organ-
Powers & Osborne, 1976). The hand- ism by momentarily altering (a) the re-
book by Honig (1966) and the later one inforcing effectiveness of other events and
by Honig and Staddon (1977) each con- (b) the frequency3 of occurrence of that
tain a chapter on motivation by Teitel- part of the organism's repertoire relevant
baum (1966, 1977); these differ from the to those events as consequences.
earlier treatments in being concerned The first effect can be called reinforcer
largely with relations between physiolog- establishing and the second evocative.
ical variables and behavior. The Honig Thus, food deprivation is an establishing
and Staddon handbook also contains a operation (EO) that momentarily in-
chapter by Collier, Hirsch, and Kanarek creases the effectiveness offood as a form
(1977), in which feeding behavior is an- of reinforcement. But food deprivation
alyzed in the context ofits ecological sig- not only establishes food as an effective
nificance. Like that of Teitelbaum, this form of reinforcement if the organism
approach is very different from the earlier should encounter food; it also momen-
ones, and is one that is to some extent tarily increases the frequency of the types
critical of some of the assumptions about of behavior that have been previously
motivation in Skinner's earlier treat- reinforced with food. In other words, it
ments. Neither the physiological nor the evokes any behavior that has been fol-
ecological type of analysis seems to have lowed by food reinforcement. This evoc-
been incorporated in the more recent ative effect is probably best thought of as
"nontreatments" of the topic of moti- (a) the result of a direct effect of the EO
vation. on such behavior, (b) an increase in the
The present state of affairs, with mo- evocative effectiveness of all SDS for be-
tivative variables being dealt with as re- havior that has been followed by food
inforcement history, deprivation and sa- reinforcement, and (c) an increase in the
tiation, or aversive stimulation, is not frequency of behavior that has been fol-
entirely satisfactory, however.2 Variables lowed by conditioned reinforcers whose
' The topic of emotion is closely related to mo-
tivation in these treatments, either as an adjacent munity (but see Morris, Higgins, & Bickel, 1982,
chapter (chap. 11 in Skinner, 1938; chap. 10 in especially pp. 161 and 167 for a contradictory view).
Keller & Schoenfeld, 1950; chap. 10 in Skinner, I In this context, frequency should be taken to
1953) or as part of a group of closely related in- mean both number of responses per unit time and
dependent variables (as in Skinner, 1957). relative frequency, the proportion of response op-
Kantor's settingfactor (1959, p. 14) includes portunities in which a response occurred. This us-
motivative variables, but until recently this concept age makes it possible to avoid such controversial
has not been much used within the behavioral com- terms as response strength and response probability.

effectiveness depends on food depriva- veloping an SD, as when a rat's lever-

tion. Food satiation-consuming food- pressing behavior is brought under the
is an EO working in the opposite direc- control of an auditory stimulus in a lab-
tion; it is actually more accurate to think oratory setting, consists of allowing re-
ofmotivative variables as establishing or inforcement for lever pressing in the
abolishing operations and to think of their presence of the stimulus, but extinguish-
evocative effect as an increase or a de- ing the response in the absence of the
crease in the momentary or current stimulus -letting the response occur, but
frequency4 ofthe relevant kind of behav- not following it with reinforcement. Al-
ior. though this point is seldom made, the
An effort has sometimes been made to extinction condition always occurs at a
interpret the evocative effect of an EO as time when the unavailable consequence
the result of stimuli produced by the rel- would be effective as reinforcement if it
evant deprivation (hunger pangs, dry were obtained. The unavailability of an
mouth and throat) functioning as dis- ineffective consequence is a behaviorally
criminative stimuli (SDs) for the relevant neutral event. It doesn't constitute an ex-
behavior (e.g., Staats, 1963, pp. 11 1-1 12). tinction condition, nor does it contribute
At this point it will be useful to provide to a correlation with differential avail-
a somewhat more restrictive definition of ability. Although it is true that food re-
the discriminative relation than usually inforcement is in a sense unavailable in
appears in basic texts and to contrast this the absence of the stimuli that we call
relation with that of the establishing op- hunger pangs, this unavailability is not
eration. the kind that contributes to a correlation.
An SD is a stimulus condition that has This issue is also critical for an under-
a history of correlation with the differ- standing of the role of painful stimulation
ential availability of an effective form of as an EO, as discussed later.
reinforcement given a particular type of To summarize, a useful contrast can
behavior. Differential availability im- usually be made as follows: Discrimi-
plies that the relevant consequence has native variables are related to the differ-
been in some way more available in the ential availability of an effective form of
presence than in the absence of the stim- reinforcement given a particular type of
ulus condition or SD. Lowered availabil- behavior; motivative variables are relat-
ity or unavailability in the absence of the ed to the differential reinforcing effective-
stimulus condition further implies that ness of environmental events.
the unavailable event would have been It seems that most events, operations,
effective as reinforcement if it had been or stimuli that alter the reinforcing effec-
obtained. It is with respect to this latter tiveness of other events also alter the mo-
requirement that most motivative vari- mentary frequency of occurrence of any
ables fail to qualify as discriminative behavior that has been followed by those
variables. The ordinary procedure for de- other events. Why this should be the case
is probably not a question that can be
answered by staying within the science of
4A change in the momentary or current frequen- behavior, but can eventually be under-
cy of all behavior that has been followed by a par- stood in neurophysiological terms or in
ticular type of reinforcement is to be contrasted
with a change in the future frequency of the partic- terms of evolutionary biology. Although
ular type of behavior that preceded a particular the reinforcer-establishing effect is im-
instance of reinforcement. Changes in future fre- portant for identifying the various EOs
quency define function-altering relations (reinforce- and distinguishing them from other types
ment, punishment, extinction), whereas changes in
current frequency define evocative relations (the of variables, it is the evocative effect of
effects of a discriminative stimulus and the various the EO that is most consistent with com-
kinds of establishing operations). For a detailed monsense concepts of motivation and
treatment of this distinction, both for operant and with motivational concepts appealed to
respondent relations, see Michael (1983). In that in other psychological theories. To
paper, function-altering effects are referred to as
repertoire altering. "want" something can be most easily de-

fined in terms of the momentary fre- perature below normal values increases
quency of the behavior that has typically the reinforcing effectiveness of a change
obtained whatever is wanted-the evoc- in the opposite direction, and more re-
ative EO effect. That the thing wanted motely of warm objects, places, and
will function as more effective reinforce- clothing. These changes also increase the
ment when and if it is obtained -the re- momentary frequency of the behavior
inforcer-establishing EO effect-is not a that has accomplished them. An increase
convenient behavioral interpretation of above normal values is an EO that has
"want," because it refers to an event that reinforcer-establishing effects in the op-
is in the future with respect to the time posite direction and evokes the behavior
the "want" is observed. that has resulted in a body-cooling effect.
These EOs could also be dealt with under
the category of aversive stimulation,
Unconditioned Establishing which includes painful stimulation, but
Operations (UEOs) it may be clearer to consider them sep-
For all organisms there are events, op- arately, especially considering the prob-
erations, and stimulus conditions whose lems of aversive stimulus as an omnibus
reinforcer-establishing effects are un- term, as discussed later.
learned. They depend upon the evolu- Variables relevant to sexual reinforce-
tionary history of the species, and vary ment. For many nonhuman mammals,
from species to species. Note that it is hormonal changes in the female are trig-
the unlearned aspect of the reinforcer- gered by time passage, ambient light con-
establishing effect that results in an EO ditions, daily average temperature, or
being classified as "unconditioned." The other features of the environment related
behavior evoked by the EO is usually phylogenically to successful reproduc-
learned. Said another way, we are prob- tion. These environmental features, or the
ably born with the capacity for our be- hormonal changes themselves, can be
havior to be more reinforceable by food considered UEOs in making sexual con-
as a result of food deprivation and more tact with the male an effective form of
reinforceable by pain cessation as a result reinforcement for the female, and si-
of pain onset, but we have to learn most multaneously causing the production of
of the behavior that produces food and chemical (olfactory) attractions that
terminates pain and is thus typically function as EOs for the male. The various
evoked by these EOs. hormonal changes may also elicit (or sen-
Deprivation and satiation. Some forms sitize for easier elicitation by other stim-
of deprivation serve as unconditioned es- uli) certain behaviors (e.g., the assump-
tablishing operations (UEOs) for humans tion of sexually receptive postures) that
and related mammals. Food, water, ac- then function as UEOs or as respondent
tivity, and sleep are things that these an- unconditioned elicitors (UEs)5 for sexual
imals can be deprived of with resultant behavior by members of the opposite sex.
reinforcer-establishing and evocative ef- Superimposed on this collection of UEOs
fects. Food and water consumption, en-
gaging in activity, and sleeping are op-
erations that work in the opposite I
The term unconditioned stimulus actually refers
direction. Salt ingestion, perspiration, and to two quite different functions, for which it is con-
blood loss are operations that have much venient to have separate terms. In the Pavlovian
the same effect on behavior as water de- arrangement, for example, food elicits salivation
when it is placed in the mouth, and it also alters
privation does -water becomes more ef- the function of other stimuli present at the same
fective as a form of reinforcement, and time (e.g., a tone) so that in the future, when those
behavior that has been reinforced with stimuli are presented by themselves, they too elicit
water becomes momentarily more fre- salivation. The food can be said to function as an
unconditioned elicitor (UE) and as an uncondi-
quent. tioned conditioner (UC). For a more detailed treat-
Temperature changes. A decrease in ment of this distinction and related terms see Mi-
skin temperature or in general body tem- chael (1983).

and UEs is a deprivation effect that may ment if it had been obtained) in the ab-
also function as a UEO. sence of the stimulus. The correlation
In the human, the situation is quite between painful stimulation and conse-
complex and not well understood. The quence availability fails in the second
behavioral role of hormonal changes in component. In the absence of shock, there
the female is not clear, nor is the role of is no effective consequence that could
chemical attractants in changing the have failed to follow the response in an
male's behavior. There do seem to be analogue to the extinction responding that
deprivation effects for both sexes. In ad- occurs in the absence of an SD. In the
dition, tactile stimulation of erogenous absence of shock, the relevant establish-
regions of the body seems to function as ing operation is absent. The fact that the
a UEO in making further similar stim- lever press does not turn off the "non-
ulation even more effective as reinforce- present" shock is in no sense extinction
ment and in evoking the behavior that responding, but rather is behaviorally
has in the past achieved such further neutral, like the unavailability of food
stimulation. But the possible role of reinforcement when one is satiated. Con-
learning in these various relations cannot trast this situation with extinction when
be ruled out. an SD is absent-the food-deprived rat
Painful stimulation: Escape. The onset presses the lever but fails to receive food,
of painful stimulation establishes the re- which would at that moment be effective
duction or offset ofthis stimulation as an as reinforcement if it were obtained. The
effective form of reinforcement and absence of the shock is more like the ab-
evokes the behavior that has achieved sence of food deprivation than like the
such reduction or offset. That painful absence of the SD. With no shock present,
stimulation is a UEO rather than an SD shock termination is not an effective form
is not well appreciated. The issue can be of reinforcement.
most easily analyzed by reference to a Note that in addition to its EO evoc-
typical laboratory shock-escape proce- ative effect, the onset of a painful stim-
dure. The response that turns offthe shock ulus also functions as a respondent UE
(often a lever press) is clearly evoked by in eliciting a number of smooth muscle
the shock onset, and it is also clearly a and gland responses such as increased
part of an operant rather than a respon- heart rate, pupillary dilation, adrenal se-
dent relation because it was developed cretion, and so forth; as a respondent UC
through the use of shock offset as rein- (unconditioned conditioner) in condi-
forcement. If the only known operant tioning these responses to other stimuli
evocative relation were the SD, then shock present at the time; as an SP (it is con-
onset would seem to be an SD; but, like venient to use the symbol SR for uncon-
the stimuli called hunger pangs, it fails ditioned reinforcement and Sr for con-
to qualify as a discriminative variable be- ditioned reinforcement; similarly, SP is
cause its absence has not been a condition used for unconditioned punishment and
in which an effective form of reinforce- SP for conditioned punishment) in de-
ment was unavailable for a particular type creasing the future frequency of any type
of behavior. of behavior that precedes it; possibly as
To repeat the argument, an SD is a an SD for any response that has been cor-
stimulus condition that has been corre- related with the availability of some form
lated with the availability of a type of of reinforcement other than pain reduc-
consequence given a type of behavior. A tion (e.g., lever pressing for food); and
correlation with availability has two possibly as a respondent CE (conditioned
components: An effective consequence elicitor), if pain has been paired with some
(one whose EO was in effect) must have other UC (unconditioned conditioner)
followed the response in the presence of such as food, in which case pain might
the stimulus, and the response must have elicit salivation (this would probably work
occurred without the consequence (which only with mild pain because intense pain
would have been effective as reinforce- as a UE would probably elicit autonomic

activity that is incompatible with sali- favor of more specific terms. In any case,
vation). proper use of aversive stimulus is often
It is possible that, in general, there is problematic, and appetitive doesn't seem
a close correlation among several of these to be catching on, but this is not the place
functions. Maybe any stimulus, the onset to do other than identify the difficulty.
of which can function as a UEO in evok- Unconditioned reinforcement. Many
ing its own termination, will also func- introductory behavioral treatments iden-
tion as a UE and a UC with respect to tify environmental changes that function
certain smooth muscle and gland re- as reinforcers in the absence of any par-
sponses (the activation syndrome, for ex- ticular learning history as unconditioned
ample), and will also function as an SP in reinforcers (SRs). They are contrasted with
weakening any response that precedes its conditioned reinforcers (Srs), whose re-
onset. This seems to be the implication inforcing effectiveness depends upon
of much current use of the term aversive some association with unconditioned re-
stimulus, where the specific behavioral inforcers. Unconditioned establishing
function is not identified. It is not clear operations are the environmental events,
at present just how close the correlation operations, or stimulus conditions that
among these functions is, nor is it clear regulate the momentary effectiveness of
that the advantages of an omnibus term unconditioned reinforcers (and also the
of this sort outweigh the disadvantage of momentary effectiveness of the condi-
its lack of specificity. It is clear that some tioned reinforcers based on those uncon-
use of this term is simply a behavioral ditioned reinforcers). The UEO and the
translation of commonsense expressions SR are obviously closely related, but it
for "feelings," "states of mind," and so would be a mistake to consider the terms
on, an undesirable usage that is fostered synonymous. Food deprivation is the
by the term's lack of specificity. The same UEO that regulates the effectiveness of
problem arises when "reinforcement" is food as an SR. The above list of UEOs
used without the implication of a implies a corresponding list of uncondi-
strengthening effect on preceding behav- tioned reinforcers.
ior, but rather as a synonym for "pleas- It is possible that there are forms of
ant" or "desired," as in the too often heard unconditioned reinforcement that are not
"That's very reinforcing!" related to any particular UEO. For ex-
To keep the term reinforcement spe- ample, infants' behavior seems to be
cific to its strengthening function and still reinforceable by a variety of mild stim-
have an omnibus term for positive events, ulus changes6 (Finkelstein & Ramey,
one might use appetitive as the opposite 1977; Kalnins & Bruner, 1973; McKirdy
of aversive. In this usage, an appetitive & Rovee, 1978; Rovee & Fagen, 1976;
stimulus or condition would be one that Watson, 1967; Watson & Ramey, 1972)
elicits certain smooth muscle and gland not easily related to any obvious EO. The
responses, conditions neutral stimuli so concept of intrinsic motivation, as pro-
that they elicit similar responses, increas- posed by Deci and Ryan (1985, chap. 1
es the future frequency of the type of be- and 2), seems to imply that "signs of
havior preceding its onset (SR), suppress- competence" and "signs of self-deter-
es the behavior that removes it (UEO), mination" always function as effective
and decreases the future frequency of the forms of unconditioned reinforcement,
behavior that precedes its termination although they can be temporarily weak-
(SP). This use of appetitive is often pre- ened by the presence of other UEOs and
sented in basic textbooks, but seems not other momentarily strong forms of re-
to be used much in the area of behavior inforcement. What such signs might con-
analysis, possibly because it still seems
to refer primarily to eating.
Perhaps the omnibus usage is essential 6 I am indebted to Henry Schlinger for reminding
for effective behavioral discourse, or per- me of this point and for directing my attention to
haps it can be and should be avoided in the references shown.

sist of for the untrained organism is not cies. It thus appears that one ofthe effects
clear, and current evidence for such re- of painful stimulation is simply to elicit
inforcement is equally well interpreted in aggressive behavior as an unconditioned
terms of Skinner's concept of generalized response. But the issue is still not entirely
conditioned reinforcement (1953, pp. 77- clear, because some or all of the effect
81). However, the issue is essentially an may be the result ofpain as a UEO (rather
empirical one, and either interpretation than as a UE), where pain increases the
is compatible with a behavioral approach effectiveness of some form of reinforce-
to motivation.7 ment specific to that EO, such as signs of
Painful stimulation: Aggression. In ad- damage to another organism or the feel
dition to establishing its own reduction of one's teeth being pressed against some-
as a form of reinforcement and evoking thing. The aggressive behavior would thus
the behavior that has produced such re- be an example of what Skinner later re-
duction, painful stimulation in the pres- ferred to as an intermingling of phylo-
ence of another organism evokes aggres- genic and ontogenic contingencies (1974,
sive behavior toward that organism. pp. 45-50, where the EO has its rein-
Skinner, in Science and Human Behavior forcer-establishing effect, which then
(1 9 5 3), discussed this effect in attempting leads to a form of rapid operant shaping
to explain the responses that vary to- of the appropriate behavior. These var-
gether in emotion. He identified two bas- ious mechanisms need not be mutually
es for such covariation. exclusive, and there is considerable evi-
Responses which vary together in an emotion do
dence for multiple provenances for these
so in part because of a common consequence. The kinds of behavior.
responses which grow strong in anger inflict damage Other emotional EOs. The above anal-
upon persons or objects. This process is often bi- ysis of painful stimulation as an EO with
ologically useful when an organism competes with respect to aggressive behavior can prob-
other organisms or struggles with the inanimate ably be extended to the other operations
world. The grouping of responses which define an-
ger thus in part depends upon conditioning. Be- that produce so-called "emotional be-
havior which inflicts damage is reinforced in anger havior" or "emotions." In Science and
and is subsequently controlled by the conditions Human Behavior (1953), Skinner de-
which control anger. (p. 163) scribes the operant aspect of emotion as
This covariation is based on ontogenic a predisposition, as follows:
factors or on a history of reinforcement. The "angry" man shows an increased probability
Phylogenic factors may also be involved. of striking, insulting, or otherwise inflicting injury
Some of the behavior involved in an emotion is and a lowered probability of aiding, favoring, com-
apparently unconditioned, however, and in that case forting, or making love. The man 'in love" shows
the grouping must be explained in terms of evo- an increased tendency to aid, favor, be with, and
lutionary consequences. For example, in some spe- caress, and a lowered tendency to injure in any way.
cies biting, striking, and clawing appear to be "In fear" a man tends to reduce or avoid contact
strengthened during anger before conditioning can with specific stimuli-as by running away, hiding,
have taken place. (pp. 163-164) covering his eyes and ears; at the same time he is
less likely to advance toward such stimuli or into
About 10 years after Science and Hu- unfamiliar territory. These are useful facts, and
man Behavior was published, Ulrich and something like the layman's mode of classification
has a place in a scientific analysis. (p. 162)
Azrin (1962) first reported their discov-
ery of the phenomenon that came to be Skinner doesn't specifically identify the
known as elicited aggression: the un- reinforcer-establishing effect of the EO
learned occurrence ofbiting, striking, and here, but in the previous passage this im-
so on, as a result of painful stimulation, plication is quite clear. Of course, as he
found to occur in a wide variety of spe- points out in discussing "the total emo-
tion" (p. 166), one must add the respon-
dent UE and UC effects of emotional op-
7 For a detailed behavioral analysis of the topic erations, and we should also add the
of intrinsic motivation, see Bernstein (1990) and possible effects of such operations as un-
Dickinson (1989). conditioned reinforcement or punish-
ment for the behavior that precedes the exist for many of the unconditioned re-
occurrence of the operation, for a com- flexes, in which case respondent evoca-
plete picture of the "total emotion" tion, like operant evocation, should be
(Skinner, 1953, p. 166). conceptualized as jointly controlled by
EOs and punishment. If EOs must be the EO and UE (or CE).8
in effect for events to function as rein-
forcement, it is reasonable to consider Conditioned Establishing Operations
their function with respect to punish- (CEOs)9
ment. Painful stimulation as punishment
functions as its own EO, but other forms Ordinary forms of conditioned rein-
of punishment, particularly those involv- forcement do not require a special EO for
ing withdrawal or removal of reinforcing their effectiveness; the UEO appropriate
events, would be expected to function as to the relevant unconditioned reinforce-
punishment (i.e., would decrease the fu- ment is sufficient. In other words, many
ture frequency of the kind of behavior learned forms of reinforcement do not
that they followed) only if those events require learned EOs. Nevertheless, there
were effective as reinforcement at the time are variables that alter the reinforcing ef-
they were withdrawn. It is not punish- fectiveness of other events, but only as a
ment to take food away from a food-sa- result of the individual organism's his-
tiated organism. tory. These are learned or conditioned
The evocative effect is more complex. establishing operations (CEOs). As with
Suppose removal of food from a food- the UEOs, they also alter the momentary
deprived animal was used as a form of frequency of the type of behavior that has
punishment for some particular type of been reinforced (or punished) by those
behavior (which was being reinforced in other events. There are at least three kinds
some other way -for example, by access of CEOs. They are all stimuli that were
to sexual stimulation). We must assume motivationally neutral prior to their re-
that the current weakening effect of this lation to another EO or to a form of re-
history ofpunishment would be seen only inforcement or punishment. They differ
when the organism was food deprived. in terms of the nature of their relation to
This relation has not received much the- the behaviorally significant event or con-
oretical or experimental attention, but dition. The simplest relation is a corre-
seems to follow naturally from existing lation in time; the neutral event is paired
knowledge of reinforcement, punish- with or systematically precedes a UEO
ment, and establishing operations. (Note (or another CEO). As a result ofthis pair-
that this relation makes the detection of ing, the neutral event may acquire the
a punishment effect doubly complex: The motivational characteristics of the UEO
effect of punishment will not be seen un- that it is paired with. I refer to this as a
less the EOs relevant to the reinforce- surrogate'0 CEO.
ment maintaining the behavior and to
the punishment weakening it are both in
effect -in the example above, unless sex- 8 I am grateful to Michael Commons for pointing
ual stimulation and food are both effec- out the possible relevance of the EO concept to
tive as reinforcement.) respondent relations.
A respondent analogy. As described I
In Michael (1982), I suggested the term estab-
above, the EO is strictly an operant evoc- lishing stimulus and the symbol SE for a learned
ative relation, but there may well be a motivative relation, with establishing operation (EO)
referring to the unlearned relation. It now seems
respondent analogy. Food deprivation that conditioned establishing operation (CEO) works
probably increases the evocative and better because of the easier contrast with uncon-
function-altering effectiveness of food as ditioned establishing operation (UEO). This ap-
a stimulus (that is, as UE and UC) as well proach also leaves establishing operation (EO) as a
useful term for the general motivative relation,
as of stimuli that have been correlated without specifying provenance.
with food (as CEs and conditioned con- 1" The term surrogate was suggested by Michael
ditioners [CCs]). Similar relations may Urbach.

A more complex relation is one in forcer refers to just such a relation. The
which a stimulus systematically precedes stimulus upon which the effectiveness of
some form of worsening,1' and if the the conditioned reinforcer depends is a
stimulus is terminated prior to the oc- CEO, in that it establishes the effective-
currence ofthis worsening, the worsening ness of another event as reinforcement
does not occur. This relation is exempli- and evokes any behavior that has pro-
fied by the warning stimulus in an avoid- duced this other event. This type of CEO
ance procedure, and this type of stimulus can be called transitive, in contrast to the
acquires the capacity to establish its own reflexive CEO. (Again, this is the gram-
termination as an effective form of con- matical usage, as with a transitive verb
ditioned reinforcement and to evoke any that takes a direct object.) As with the
behavior that has accomplished this ter- reflexive CEO, one must consider both
mination. In the opposite direction, a the positive and the negative case. With
stimulus that systematically precedes a conditional conditioned punisher, the
some form of improvement, and whose stimulus upon which the effectiveness of
termination prevents the occurrence of the conditioned punisher depends is a
the improvement, will acquire the ca- CEO, in that it establishes the effective-
pacity to establish its own termination as ness of another event as punishment and
a form of conditioned punishment and suppresses any behavior that has pro-
to suppress any behavior that has accom- duced this other event. In an earlier paper
plished this termination. In an earlier pa- (Michael, 1988) I suggested the term
per (Michael, 1988) I referred to these as blocked-response CEO for this relation,
a threat CEO and a promise CEO. It now because many human examples were
seems more reasonable to refer to a CEO characterized by a stimulus change func-
that establishes its own termination as a tioning as an SD for a response that could
form of reinforcement or punishment as not take place until some object was
a reflexive CEO, a term that is more in- available, and thus functioning as a CEO
dicative of the effect of this CEO in al- in establishing the object as a condi-
tering its own function. (Reflexive here is tioned reinforcer and evoking the behav-
meant in the grammatical sense, not as ior that had obtained such an object. The
referring to a reflex. This usage is thus slotted screw example (described below)
somewhat similar but not identical to the has this pattern, but some CEOs of this
mathematical and logical use that occurs type are simply a stimulus upon which
in the context of equivalence relations.) the reinforcing effectiveness of another
An even more complex relation exists stimulus depends, but with no response
in the correlation of a stimulus with the blocked (like the nonhuman example of
correlation between another stimulus and lever pressing). The three CEO types will
a form of unconditioned reinforcement. now be considered in detail.
The term conditional conditioned rein- Surrogate CEO: Correlating a stimulus
with a UEO. The development of the CE
(conditioned elicitor), Sr (conditioned re-
I use the term worsening to refer to any stim-
inforcer), and SP (conditioned punisher)
ulus change that would function as punishment for each involves pairing or correlating a
the type of behavior that preceded it. I avoid the neutral event with a behaviorally effec-
term punishment, because in the context ofdescrib- tive one as a way of giving the neutral
ing this CEO I am not referring to the decrease in event some of the behavioral properties
future frequency of any behavior. Similarly, I use of the effective one. It is not unreasonable
improvement for a change that would function as
reinforcement for the type of behavior that preced- to suppose that properties of the EO could
ed it, but I am not referring to an increase in the be developed in the same way. The ques-
future frequency of any behavior in this context. tion is, would a stimulus that had been
The term aversive stimulus would be appropriate, correlated with a UEO become capable
except for my general uneasiness about omnibus of the same reinforcer-establishing and
terms, as expressed earlier. Worsening should be
considered a term from everyday usage, not a tech- evocative effects as that UEO? The terms
nical term. learned drive or acquired drive appeared

quite often in the early learning literature. deprivation ate significantly more than
A chapter by Miller, entitled "Learnable the group with the history of 1 hr of de-
Drives and Rewards," was included in privation. There were several attempts to
the 1951 Stevens Handbook of Experi- replicate these results during the next
mental Psychology. Much speculation re- several years; some were successful, but
garding human behavior has taken the most failed to produce similar results. In
form of postulating various learned mo- a comprehensive review of this line of
tives, but as Miller pointed out, "the ex- research, Cravens and Renner (1970)
perimental work on learned drives and identified several major methodological
rewards is limited almost exclusively to problems with most of the research, and
(1) fear as a learnable drive and fear re- concluded that the results were essen-
duction as a reward or (2) learned re- tially uninterpretable.
wards and drives based on hunger and Mineka (1975) suggested that gusta-
food" (1951, pp. 435-436). Although tory and olfactory stimuli are more ap-
couched in the language of hypothesized propriate as conditioned elicitors for a
internal drive states, the work on fear as hunger drive than are, the visual stimuli
a learnable drive is primarily concerned that had been used in most of the pre-
with the motivative characteristics of the vious studies. She conducted a series of
warning stimulus in an avoidance situ- experiments comparing visual and gus-
ation (the reflexive CEO described in de- tatory stimuli, with favorable results when
tail below). Miller's section on learned the latter were used, but then failed to
rewards and drives based on hunger and replicate those results in a subsequent ex-
food (1951, pp. 454-462) dealt exten- periment, and ultimately concluded that
sively with the development and use of the phenomenon may not exist. Mineka
learned rewards (conditioned reinforc- also made an interesting point about the
ers), but considered learned drives only possible biological uselessness of such
briefly. Subsequent to that publication, learned appetitive drives, in that eating
there was some research on the possibil- more than is appropriate for a given de-
ity of developing a learned appetitive privation level simply because one has
drive in a laboratory situation. The ques- been hungry in that particular stimulus
tion was whether stimuli correlated with condition before would not be to the or-
high levels of food deprivation would ganism's long-term advantage or surviv-
produce, by themselves, a momentary in- al.
crease in the frequency of the behavior Not much research of this type has ap-
that had been reinforced by food. Also, peared since Mineka's 1975 report, but
would they increase, by themselves, the it would be premature to exclude the pos-
effectiveness of food, water, and so on, sibility of this CEO on the basis of the
as forms of reinforcement? (Of course unclear empirical evidence or hypothe-
these questions could be asked just as sized negative survival value. Depriva-
reasonably about other UEOs, such as tion-satiation UEOs typically build up
water, sleep, activity, or sex deprivation, slowly, and it is not easy for a stimulus
but most of the research involved food to become correlated with the extreme
deprivation.) values of such a build-up. UEOs with
In the first experiment of this type, Cal- more rapid onset, however, are often
vin, Bicknell, and Sperling (1953) placed paired with relatively unique stimuli that
rats in a distinctively striped box for 30 might be expected to develop CEO prop-
min a day for 24 days. During this train- erties. Would stimuli that were correlat-
ing one group was placed in the box while ed with decreases in temperature, for ex-
food deprived for 22 hr, and the other ample, have CEO effects similar to the
group was placed in the box while de- effects of those temperature decreases
prived for only 1 hr. After training, both themselves? In the presence of such stim-
groups were allowed to eat in the striped uli, would warmth be more reinforcing
box following 11.5 hr of food depriva- than would be appropriate for the actual
tion; the rats with the history of 22-hr temperature, and would behavior that has
produced such warmth be more frequent CE. Thus, with painful stimulation and
than it ought to be for the actual tem- aggressive behavior, if the potential CEO
perature? evokes some arbitrary response (such as
Note that this question is not about the lever pressing) that has been developed
possibility of conditioned elicitation or by reinforcement with access to another
conditioned reinforcement or punish- organism to attack, then it is functioning
ment. It is well known that a neutral stim- as a CEO rather than a CE, because there
ulus (e.g., a tone) paired with cold (hand is no UE for such behavior. The issue
dipped in ice water) will come to elicit would not be clear if the behavior studied
appropriate smooth muscle responses was striking, biting, and so forth, because
(peripheral vasoconstriction) when it is these may be elicited by painful stimu-
presented alone. It is also quite clear that lation as a UE. Similarly, with sexual mo-
if the onset of the cold stimulus functions tivation, if the previously neutral stim-
as punishment (SP) for any response that ulus evoked an arbitrary response (such
precedes it, then so too will any stimulus as lever pressing) that had been rein-
that is correlated with such onset (SP). forced with access to sexual stimulation,
Neither of these functions (CE, SP) is syn- it is functioning as a CEO rather than a
onymous, however, with the operant CE, but the issue would be unclear if the
evocative effect of an EO, although their behavior was pelvic thrusting, which
occurrence might always be a good basis might have been elicited by a UE.
for predicting the CEO effect, which might The possibility of developing a surro-
well be based on the same physiological gate CEO based on painful stimulation
processes. I know of no research bearing as a UEO for escape behavior must be
directly on the existence of a CEO evoc- carefully distinguished from the next type
ative effect based on pairing with tem- of CEO (to be discussed below). The issue
perature changes, but the possibility is whether correlating a neutral stimulus
seems worth considering. with painful stimulation will increase the
With sexual motivation, EOs for ag- effectiveness of pain reduction as a form
gressive behavior, and the other emo- of reinforcement and evoke the behavior
tional EOs, the issue has not been ad- that has been reinforced with pain re-
dressed in terms specific to the CEO, duction. It is not clear what it means to
because its distinction from CE, Sr, and increase the effectiveness of pain reduc-
SP has not been previously emphasized. tion when no pain is present, but such a
There is evidence that a stimulus corre- stimulus in the presence of mild pain
lated with painful stimulation will in- might cause the mild pain reduction
crease the frequency of aggressive behav- (along with the reduction of the CEO) to
ior when presented alone (Farns, Gideon, be more like the reduction of more severe
& Ulrich, 1970), but it is not clear wheth- pain. Much less difficult to measure would
er it is functioning as CEO, CE, or both. be the extent to which such a stimulus
The basic experimental design is simple evoked the pain-escape response in the
enough: Correlate a neutral stimulus con- absence of pain. In the typical shock-es-
dition with a UEO, and then see if by cape experiment, all that is necessary is
itself it increases the reinforcing effec- to precede the onset of shock with a warn-
tiveness of the consequence relevant to ing stimulus and see if the shock-escape
the UEO and increases the momentary response is increased in frequency by the
frequency of the behavior that has been onset of the warning stimulus. Note that
developed through reinforcement by that this is not the typical escape-avoidance
consequence. Reinforcing effectiveness is procedure. As described below, there is
not easy to quantify, but the evocative no question that a stimulus that system-
effect should be easy to measure, and its atically precedes the delivery of a second,
presence should be evidence enough for painful stimulus will evoke the behavior
the CEO effect. One must, of course, use that terminates the first stimulus and thus
behavior that is clearly of learned operant avoids the onset of the pain. Here the
origin to prevent confusion of CEO with question is whether the warning stimulus
will evoke the response that terminates stimulus and starts the intertrial interval,
the pain, even though the pain is not yet thus avoiding the shock. As a result of
present, and even though such a response exposure to this procedure, many organ-
has not prevented the onset of pain. isms acquire a repertoire that consists of
The situation can be clarified by ref- making the relevant response during most
erence to an unusual type of avoidance of the occurrences of the warning stim-
experiment, one with escape and avoid- ulus.
ance responses of quite different topog- Recall the analysis of the role of shock
raphies. Imagine a rat in a procedure in as an EO for the escape response, the
which a lever press terminates the shock reinforcement for which is shock termi-
but a wheel turn terminates the warning nation. The warning stimulus has a sim-
stimulus and avoids the shock. The pres- ilar function, except that its capacity to
ent CEO would be demonstrated if the establish its own termination as an effec-
warning stimulus evoked the lever press; tive form of reinforcement is of ontogen-
the CEO discussed below would be dem- ic provenance-due to the individual's
onstrated ifthe warning stimulus evoked own history involving the correlation of
the wheel turn. The occurrence of such the presence of the warning stimulus with
behavior could have other interpreta- the onset of the painful stimulation. In
tions, but it will probably be possible to other words, the warning stimulus as a
exclude these with appropriate experi- CEO evokes the so-called avoidance re-
mental designs. This process, like the ones sponse, just as the painful stimulation as
described above, seems intuitively quite a UEO evokes the escape response. In
plausible, but research directed precisely neither case is the relevant stimulus cor-
at the CEO issue has not yet been con- related with the availability of the re-
ducted. sponse consequence, but rather with its
Reflexive CEO: Correlating a stimulus reinforcing effectiveness.
with worsening or improvement. In the In more general terms, any stimulus
traditional discriminated'2 avoidance that is positively correlated with the on-
procedure, an intertrial interval is fol- set of painful stimulation becomes a CEO,
lowed by the onset of an initially neutral in that its own offset will function as re-
warning stimulus, which is in turn fol- inforcement and it will evoke any be-
lowed by the onset of painful stimula- havior that has been followed by this re-
tion-usually electric shock. Some arbi- inforcement. But this set of functional
trary response (i.e., one that is not part relations is not limited to painful stim-
of the animal's phylogenic pain-escape ulation as a form of worsening (or even
repertoire) terminates the painful stim- to worsening, as will be seen later). It is
ulation and starts the intertrial interval. well known that organisms can learn to
The same response, if it occurs during avoid forms ofstimulus change other than
the warning stimulus, terminates that the onset of pain. Stimuli that warn of a
lowered frequency of food presentation,
12 The term discriminated arose so that this type
increased effort, a higher response ratio
ofprocedure could be distinguished from an avoid-
requirement, longer delays to food, and
ance procedure with no programmed exteroceptive so forth will all evoke the behavior that
stimulus except for the shock itself. It also implies terminates such stimuli. These events
that the warning stimulus is a discriminative stim- have in common a form of worsening,
ulus for the avoidance response, but the main point and stimuli positively correlated with
of the present section contradicts this practice; thus, such events are often called conditioned
we should develop a new name for this type of
procedure. Sometimes this procedure is called aversive stimuli, without specifying any
avoidance without a warning stimulus and is then particular behavioral function. It is pos-
contrasted with avoidance with a warning stimulus. sible that these stimuli will generally
This may be preferable to discriminated, but it im- function as CEOs in evoking the behav-
plies the effect of the stimulus on the organism-it ior that terminates themselves, as con-
warns the organism -and it would be preferable if
the terms for procedures did not presuppose their ditioned punishment (SP) for any behav-
behavioral functions. ior that precedes their onset, and as

conditioned elicitors (CEs) for smooth occurrence, but that is not the functional
muscle and gland responses of the same relation under consideration. Its CEO ef-
type that are produced by painful stimuli. fect consists of its establishing its own
And, of course, the offset of such stimuli offset as effective punishment and sup-
will function as reinforcement (Sr) for any pressing (as an opposite to "evoking")
behavior that precedes that offset. As any behavior that has been so punished.
mentioned above in connection with the The relation is quite plausible, although
term aversive stimulus, there are advan- I know of no research that is directly rel-
tages and disadvantages to such omnibus evant. Stimuli that are negatively corre-
terms, and in any case their availability lated with some form of worsening (safe-
does not obviate the necessity for more ty signals) have the same status as those
specific terms. With respect to the CEO, that are positively correlated with im-
the case can be most clearly stated as provement. Their onset would be ex-
follows: Any stimulus condition whose pected to establish their removal as a form
presence or absence has been positively of punishment and to suppress any be-
correlated with the presence or absence havior that had been so punished. Sim-
of any form of worsening will function ilarly, but in the opposite direction, a
as a CEO in establishing its own termi- stimulus that has been negatively corre-
nation as effective reinforcement and in lated with improvement would be ex-
evoking any behavior that has been so pected to establish its removal as rein-
reinforced. forcement and evoke behavior that has
It may be useful to repeat the argument been followed by such reinforcement.
against such stimuli being considered as None of these last relations have been
discriminative stimuli. A discriminative verified directly by research but seem to
relation involves a correlation with the follow naturally from existing knowl-
availability of a type of consequence giv- edge.
en a type of behavior. A correlation with There is an important additional re-
availability has two components: An ef- quirement involved in the correlation
fective consequence (one whose EO was histories discussed above. It is essential
in effect) must have followed the re- for the stimulus not only to have preced-
sponse in the presence of the stimulus, ed the worsening or improvement but also
and the response must have occurred for removal of the stimulus to have sys-
without the consequence (which would tematically prevented the worsening or
have been effective as reinforcement if it improvement. If removal of the stimulus
had been obtained) in the absence of the does not prevent the worsening or im-
stimulus. The correlation between the provement, there is no correlation (D'A-
warning stimulus and consequence avail- mato, Fazzaro, & Etkin, 1968; see also
ability fails in the second component. In Fantino & Logan, 1979, pp. 273-275).
the absence of the warning stimulus, there Transitive CEO: Conditional condi-
is no effective consequence that could tioned reinforcement and punishment.
have failed to follow the response in an When a stimulus condition (S1) is cor-
analogue to the extinction responding that related with the correlation between an-
occurs in the absence of an SD. The fact other stimulus (S2) and some form of
that the avoidance response does not turn improvement (or worsening), the pres-
off the nonpresent warning stimulus is in ence of the S1 establishes the reinforcing
no sense extinction responding, but rath- (or punishing) effectiveness of S2 and
er is behaviorally neutral, like the un- evokes (or suppresses) the behavior that
availability of food reinforcement for a has been followed by that reinforcement
food-satiated organism. or punishment. Again, this relation has
Now consider a stimulus that is posi- not been directly researched, but it fol-
tively correlated with some form of im- lows easily from existing principles and
provement. Such a stimulus would clear- concepts. Many (probably most) forms
ly function as conditioned reinforcement of conditioned reinforcement or condi-
(Sr) for any response that preceded its tioned punishment are themselves con-
ditional upon other stimulus conditions. In other words, SI onset should function
This notion is sometimes referred to by as conditioned reinforcement, but is con-
saying that conditioned reinforcing effec- ditional upon the presence of S2; it is thus
tiveness is dependent upon a "context." a form of conditional conditioned rein-
Imagine a food-deprived animal in an forcement. The stimulus upon which its
environment in which it can always pro- reinforcing effectiveness is conditional is
duce a 10-s buzzer sound by pressing a the supposed CEO. Pigeons learn to stop
lever. Distinctive visual stimuli are re- producing SI except in the presence of
lated to the relation of this auditory stim- S2, but in the experiments cited above it
ulus to food. In the presence of a red has not been possible to exclude the pos-
overhead light, the 10-s buzzer sound sibility that S2 is simply functioning as
ends with the delivery of food. In the the first discriminative stimulus in a two-
absence of the red light, the buzzer sound response chain.
lasts for 10 s and then ends without any The first element in a chain evoked by
food delivery. This is a situation in which an SD is often a CEO of this type. Con-
the auditory stimulus functions as a con- sider a rat in a chamber in which an au-
ditioned reinforcer, but is conditional ditory stimulus is related as an SD to the
upon the color of the overhead light. Thus availability of food for a lever press. But
the buzzer onset is not effective as rein- the lever cannot be pressed until it is lo-
forcement until the red overhead light cated, so the auditory stimulus evokes
comes on. When it does, with a well- visual search behavior, which is rein-
trained animal, the lever press will be forced by seeing the lever. The auditory
evoked. What is the reinforcement for stimulus is not related to the availability
the lever press? Obviously the buzzer on- of this reinforcement, however, but rath-
set. How does the red overhead light er to its value. (Once the lever is seen,
evoke the lever press? Not as an SD, be- the other elements of the chain-ap-
cause it is not correlated with availability proaching, touching, pressing-are con-
of the buzzer-the buzzer is actually trolled by a succession of SDs, but the first
available irrespective of the light condi- element is not.) Similarly, in an avoid-
tion, but it is not an effective form of ance situation, the warning stimulus
reinforcement in the absence of the red evokes the avoidance response as a re-
light. It evokes the lever press as a CEO, flexive CEO, but if this requires locating
a stimulus change that alters the rein- an operandum, the visual search behav-
forcing effectiveness-the value-of the ior is evoked by the warning stimulus as
buzzer sound and evokes the behavior a transitive CEO, which is correlated with
that produces it. Only in the red light has the value of seeing the lever, not the avail-
the buzzer been paired or correlated with ability of its sight.
food, so only in the red light is it an ef- This type of CEO is exemplified by
fective form of conditioned reinforce- many human examples. A workman is
ment. The basic relation is still that of disassembling a piece of equipment. His
correlation, but of a more complex type. assistant hands him tools as he requests
The buzzer's correlation with food is it- them. In the process of disassembling, he
self correlated with the light color. encounters a slotted screw that must be
There have been several attempts to removed, and requests a screwdriver. The
demonstrate this type of CEO with non- sight of the slotted screw "evoked" the
humans (Alling, 1991; McPherson & Os- request, the reinforcement for which is
borne, 1986, 1988). The experiments receiving the screwdriver. To refer to the
have in common the following arrange- slotted screw as an SD for the request,
ment. One stimulus condition, S1, can however, raises the same difficulty as be-
be produced in both the presence and fore. This stimulus has not been differ-
absence of another stimulus condition, entially correlated with successful re-
S2. Onset of the first stimulus system- quests-screwdrivers are not more
atically precedes food reinforcement in available when slotted screws are around
the presence of S2, but not in its absence. than in their absence, but rather more

valuable. The slotted screw should be extensive histories. However, they would
considered a CEO for the request, not an be expected in the repertoires of long-
SD. Here the slotted screw is like the red lived species in their natural environ-
light. In its presence, screwdrivers have ments, and most certainly in those of hu-
been correlated with successful disassem- mans.
bly and are therefore valuable.
Another common human example is a General Implications
stimulus related to some form of danger, Motivation as a topic within behavior
in its evocation of protective behavior. analysis can be reintroduced as consid-
A night watchman patrolling an area hears eration of those variables-establishing
a suspicious sound and pushes a button operations-that momentarily alter the
on his radio phone that causes the other effectiveness of other events as reinforce-
night watchman to answer the phone and ment (and punishment) and simulta-
ask if help is needed. The suspicious neously alter the frequency of those types
sound is not an SD in the presence of of behavior that have been followed by
which such help is more available, but that reinforcement (or punishment). A
rather more valuable. Note that this ef- clear distinction is possible between mo-
fect of the danger signal is not to produce tivative and discriminative variables in
its own termination, but rather to in- terms of whether they are related to the
crease the value of some other event. reinforcing effectiveness of an event or to
This type of analysis seems to be re- its availability. The application of this
quired irrespective of the direction of the distinction is especially critical for the
first or second correlations, and irrespec- proper interpretation of some of the ef-
tive of whether the final event is im- fects of painful stimulation. When ap-
provement or worsening. To consider one plied to learned functional relations, this
more example, let the buzzer be a stim- distinction permits classification of a
ulus that is negatively correlated with a number of seemingly discriminative re-
worsening of some sort, in other words, lations as conditioned establishing op-
let the buzzer be a safety signal; but let erations, further enlarging the topic of
this correlation be in effect only when the motivation within behavior analysis and
overhead light is red. Under other stim- facilitating useful identification of the
ulus conditions, the buzzer is uncorre- various factors involved in the multiple
lated with any form of worsening. Now control of human behavior.
let the lever press be a response, main-
tained by an unrelated form ofreinforce-
ment, that has in the past also terminated REFERENCES
the buzzer, and thus in the red light ter-
minated the "safety," a form of condi- Alling, K. (1991). The effects of a conditioned
establishing operation on performance of a two-
tioned punishment for the lever press. component chain. Unpublished master's thesis.
The lever press, of course, also termi- Western Michigan University, Kalamazoo.
nates the buzzer when the red light is not Bernstein, D. J. (1990). Of carrots and sticks: A
on, but in this case it is not punishment, review of Deci and Ryan's Intrinsic Motivation
because in the absence of the red light the and Self-Determination in Human Behavior.
Journal ofthe ExperimentalAnalysis ofBehavior,
buzzer is uncorrelated with worsening. 54, 323-332.
We would expect that when the red light Calvin, J. S., Bicknell, E. A., & Sperling, D. S.
came on and the buzzer was on, any ten- (1953). Establishment of a conditioned drive
dency to press the lever would be "sup- based on the hunger drive. Journal of Compar-
ative and Physiological Psychology, 46, 173-175.
pressed." The red light functions as a CEO Catania, A. C. (1979). Learning. Englewood Cliffs,
to cause buzzer offset to function as ef- NJ: Prentice-Hall.
fective punishment and to suppress any Catania, A. C. (1984). Learning (2nd ed.). Engle-
behavior that has been so punished. These wood Cliffs, NJ: Prentice-Hall.
higher order relations may not play a ma- Collier, G., Hirsch, E., & Kanarek, R. (1977). The
operant revisited. In W. K. Honig & J. E. R.
jor role in the typical experiment with Staddon (Eds.), Handbook of operant behavior
nonhumans because they require such (pp. 28-52). Englewood Cliffs, NJ: Prentice-Hall.
Cravens, R., & Renner, K. E. (1970). Conditioned uli. Journal of the Experimental Analysis of Be-
appetitive drive states: Empirical evidence and havior, 37, 149-155.
Michael, J. (1983). Evocative and repertoire-al-
theoretical status. Psychological Bulletin, 73, 212-
220. tering effects of an environmental event. The
D'Amato, M. R., Fazzaro, J., & Etkin, M. (1968). Analysis of Verbal Behavior, 2, 19-21.
Anticipatory responding and avoidance discrim- Michael, J. (1988). Establishing operations and
ination as factors in avoidance conditioning. the mand. The Analysis of Verbal Behavior, 6,
Journal of Experimental Psychology, 77, 41-47. 3-9.
Deci, E. L., & Ryan, R. M. (1985). Intrinsic mo- Millenson, J. R. (1967). Principles of behavioral
tivation and self-determination in human behav- analysis. New York: Macmillan.
ior. New York: Plenum. Millenson, J. R., & Leslie, J. C. (1979). Principles
Dickinson, A. M. (1989). The detrimental effects ofbehavioral analysis (2nd ed.). New York: Mac-
of extrinsic reinforcement on intrinsic motiva- millan.
tion. The Behavior Analyst, 12, 1-15. Miller, N. E. (1951). Learnable drives and re-
Fantino, E., & Logan, C. A. (1979). The experi- wards. In S. S. Stevens (Ed.), Handbook of ex-
mental analysis of behavior. San Francisco: W. perimental psychology (pp. 435-472). New York:
H. Freeman and Company. John Wiley & Sons.
Farris, H. E., Gideon, B. E., &Ulrich, R. E. (1970). Mineka, S. (1975). Some new perspectives on con-
Classical conditioning of aggression: A devel- ditioned hunger. Journal of Experimental Psy-
opmental study. The Psychological Record, 20, chology: Animal Behavior Processes, 104, 143-
63-67. 148.
Finkelstein, N. W., & Ramey, C. T. (1977). Learn- Morris, E. K., Higgins, S. T., & Bickel, W. K. (1982).
ing to control the environment in infancy. Child The influence of Kantor's interbehavioral psy-
Development, 48, 806-819. chology on behavior analysis. The Behavior An-
Holland, J. G., & Skinner, B. F. (1961). The anal- alyst, 5, 159-173.
ysis of behavior. New York: McGraw-Hill. Powers, R. B., & Osborne, J. G. (1976). Funda-
Honig, W. K. (1966). Operant behavior: Areas of mentals of behavior. St. Paul, MN: West.
research and application. New York: Appleton- Rovee, C. K., & Fagen, J. W. (1976). Extended
Century-Crofts. conditioning and 24-hour retention in infants.
Honig, W. K., & Staddon, J. E. R. (Eds.). (1977). Journal of Experimental Child Psychology, 21,
Handbook of operant behavior. Englewood Cliffs, 1-11.
NJ: Prentice-Hall. Skinner, B. F. (1938). Thebehavioroforganisms.
Kalnins, I. V., & Bruner, J. S. (1973). The coor- New York: Appleton-Century-Crofts.
dination of visual observation and instrumental Skinner, B. F. (1953). Science and human behav-
behavior in early infancy. Perception, 2, 307-314. ior. New York: MacMillan.
Kantor, J. R. (1959). Interbehavioral psychology. Skinner,B.F. (1957). Verbalbehavior. NewYork:
Granville, OH: Principia Press. Appleton-Century-Crofts.
Keller, F. S., & Schoenfeld, W. N. (1950). Prin- Skinner, B. F. (1974). About behaviorism. New
ciples ofpsychology. New York: Appleton-Cen- York: Vintage Books.
tury-Crofts. Staats, A. W. (1963). Complex human behavior
Lundin, R. W. (1961). Personality, an experi- (with contributions by C. K. Staats). New York:
mental analysis. London: MacMillan. Holt, Rinehart and Winston.
Lundin, R. W. (1969). Personality, a behavioral Teitelbaum, P. (1966). The use of operant meth-
analysis. London: MacMillan. ods in the assessment and control ofmotivational
Mazur, J. E. (1986). Learning and behavior. En- states. In W. K. Honig (Ed.), Operant behavior:
glewood Cliffs, NJ: Prentice-Hall. Areas of research and application (pp. 565-608).
Mazur, J. E. (1990). Learning and behavior (2nd New York: Appleton-Century-Crofts.
ed.). Englewood Cliffs, NJ: Prentice-Hall. Teitelbaum,P. (1977). Levelsofintegrationofthe
McKirdy, L. S., & Rovee, C. K. (1978). The re- operant. In W. K. Honig & J. E. R. Staddon
inforcing efficacy of visual and auditory com- (Eds.), Handbook of operant behavior (pp. 7-27).
ponents in infant conjugate conditioning. Journal Englewood Cliffs, NJ: Prentice-Hall.
of Experimental Child Psychology, 25, 80-89. Ulrich, R. E., & Azrin, N. H. (1962). Reflexive
McPherson, A., & Osborne, J. G. (1986). The fighting in response to aversive stimulation. Jour-
emergence of establishing stimulus control. The nal of the Experimental Analysis of Behavior, 5,
Psychological Record, 36, 375-386. 511-520.
McPherson, A., & Osborne, J. G. (1988). Control Watson, J. S. (1967). Memory and contingency
of behavior by an establishing stimulus. Journal analysis in infant learning. Merril-Palmer Quar-
ofthe Experimental Analysis ofBehavior, 49, 213- terly, 13, 55-76.
227. Watson, J. S., & Ramey, C. T. (1972). Reactions
Michael, J. (1982). Distinguishing between dis- to response-contingent stimulation in early in-
criminative and motivational functions of stim- fancy. Merril-Palmer Quarterly, 18, 219-227.