by Alfred A. Barrios, PhD.
Self-Programmed Control Center, Los Angeles, CA


A comprehensive theory of hypnosis is presented which attempts to explain the three main aspects of hypnosis in terms of principles of condi­tioning and inhibition.

  1. Hypnotic induction is explained as a conditioning procedure for producing an inhibitory set.
  2. Hypnotic phenomena (increased responsiveness to suggestion) occur because this set can inhibit stimuli (both sensory and cognitive) which would ordinarily contradict the suggested response.
  3. Post-hypnotic behavior changes are explained as occurring through a process of higher-order conditioning; this conditioning being facilitated by the inhibitory set which inhibits stimuli that would be incompatible with the new association.

The theory is felt to be broad enough to cover not only hypnosis and suggestion, but also such related areas as persuasion, the placebo effect, and faith, as well as to throw added light on the area of conditioning.

Throughout the years many extraordinary phenomena have been attributed to the effects of hypnosis and great claims made as to its efficacy in therapy. Yet, in spite of such claims, it seems that there continues to be relatively little interest shown in it by the psychological and psychiatric community. Why?

It is felt that the reason for the continued apathy toward hypnosis is not that the claims made for it are untrue or exaggerated, but that it is still virtually an unknown. This unknown quality has led to the arousal of fears (an innate response to an unknown), many misconceptions and various unjust criticisms, and, consequently, rejection or avoidance of the area. What we are in need of, then, is a rational theory or explanation of hypnosis, one that will tie it down to known laws and facts. and, thus, help us to make the most of this vast, unexplored area. The following theory is presented as an attempt at achieving this goal.

The theory will be divided into three major sections, one each for what is felt are the three major aspects of hypnosis. Each section will start off with definitions of terms, then the major hypotheses and their corollaries will be presented together with available evidence in support of them, followed by suggestion for further tests.   There are a total or seven hypothesis making up the theoretical system. Hypotheses I - III deal with the first aspect, the hypnotic induction   Hypotheses IV and V deal with the second aspect, hypnotic phenomena.  Hypotheses VI and VII deal with the third aspect, post hypnotic suggestion. The reason for dividing the theory into three sections is to emphasize the fact that when one attempts to explain hypnosis, he has to do more than just explain hypnotic phenomena. He has also to explain how the hypnotic state was produced and how hypnosis can produce post-hypnotic behavior changes. Most previous theories deal only with hypnotic phenomena, per se.

The overall explanation presented will he based mainly on principles of conditioning and inhibition delineated in the postulates. Briefly, hypnotic induction will be explained as the conditioning of an inhibitory set, a set which increases responsiveness to suggestion by inhibiting stimuli and thoughts incompatible with a suggested response. The various hypnotic phenomena, including the phenomenon of post-hypnotic suggestion, will then be explained in terms of this set.

The theory presented herein was part of a doctoral dissertation completed in 1969 at the University of California, Los Angeles.  The work was supported in part by a Public Health Ser­vice fellowship (MPM-l3, 264-cl) from the National Institute of Mental Health, Public Health Service.


One of the major purposes of any theory or explanation should be to tie the phenomena to be explained down to known laws. This can be done by first stating the known laws and then showing how the theory (the system of hypotheses explaining the phenomena) can be deduced from these laws or can be shown to be compatible with them. In what follows, we will present the known ''laws'' (the postulates) that the theory of hyp­nosis will be tied down to.

As the reader goes through the postulates, he should keep in mind that at the present stage the science of psychology has not yet advanced to the point where we can really speak in terms of "laws" in the sense of firmly established and accepted laws.  Thus, he is not expected to accept absolutely the validity of the postulates. However, it is felt that sufficient evidence will be presented to show that these postulates are reasonable approximations to established laws. How close an approxima­tion they are will, to a large extent, be mirrored by how valid are the hypotheses and corollaries deduced from these laws. Thus, if the hypo­theses of the present theory are tested and validated, then the validity of the postulates themselves will be further strengthened.

Postulate I.  Reciprocal Inhibition – When an organism is attending or responding to one Stimulus, there will be a reciprocal inhibition of incompatible stimuli and responses.

Sherrington (1906, 1940) was one of the first to discover the phenomenon of reciprocal inhibition. He found that "incompatible movements such as turning the eyes to the right and left are so controlled in their nerve centers that with increased activity of one muscle goes decreased activity of its antagonist. The same type of inhibition is observed in human attention and distraction, since in attending to one object, you cease attending to another." (Woodworth and Schlosberg, 1954, p. 669).

The latter contention is supported by the work of Hernández-Peón (1959) who has shown that when an organism is attentive to one stimulus other stimuli impinging upon it tend to be inhibited.  This centrifugal inhibition of afferent sources has been demonstrated for all sense modalities  (Lindsley, 1961). Which stimulus will be most likely to be attended or responded to (and, therefore, which stimuli will be inhibited) in a given situation will depend on a number of different factors, such as: stimulus intensity, novelty of the stimulus, acquired significance of the stimulus, sense modality, etc. (Berlyne, 1960). Some types of stimuli, then will have preference or dominance over others, and they, in turn, will have dominance over others, and so on, thus forming a "stimulus dominance hierarchy" (SDH).

Corollary 1 – If a dominant stimulus is itself inhibited or eliminated, those stimuli below it in the hierarchy which it was reciprocally inhibiting will now be responded to more strongly.
Postulate II.  Cognitive Stimuli – Behavior is determined by cognitive stimuli as well as sensory stimuli.

We know that an organism's behavior in a given situation can be determined by certain innate behavior patterns. A pin prick will evoke a pain response (withdrawal of the injured part, crying out, heart racing, palms sweating, etc.); salt on the tongue will elicit salivation; stimulation of the erogenous zones will evoke certain patterns of physiological res­ponses; etc. Such stimulation seems to trigger "built-in" or  innate patterns of behavior.  But we also know that organisms do not always make the same response to the same stimulus. Learning or conditioning can and does play a very big part, especially with humans, in modifying behavior. For example, the response resulting from stimulation of the erogenous zones will vary from individual to individual as a result or the individual's previous experience; i.e., his previous conditioning.  If a person has been taught that sex is something dirty and bad, he could easily respond with feelings of disgust or guilt rather than with the normal ("built-in") sexual response. Thus, we can say that stimulation can also trigger "acquired" or learned patterns of behavior.

One way of conceptualizing this modification of behavior by learning is to think of the organism as reacting not only to sensory stimuli but reacting as well to what may be called memory, recorded. or "cognitive" stimuli. A sensory stimulus can be defined as coming to the organism via the sensory pathways. A "cognitive stimulus" will herein be defined as a stimulus emanating from engrams (permanent traces or recordings of past experiences in the brain). It is postulated that this stimulus is as potentially capable of initiating and directing behavior as any sensory stimulus. This means, for instance, that a stimulus dominance hierarchy can be made up of both sensory and cognitive stimuli.

These engrams are felt to he formed through a process of conditioning (see Postulate III, below) and are triggered by the conditioned stimulus. This conditioned stimulus can be either a sensory stimulus or another cognitive stimulus. For example, the thought of a steak (a cognitive stimulus) can be triggered by the smell of a steak cooking (a sensory stimulus) or the thought of a particular restaurant specializing in steaks (a cognitive stimulus).

Under the heading of cognitive stimuli we would find such things as thoughts, images, beliefs, sets, values, attitudes, ideas, etc. A cognitive stimulus can also be looked upon as the equivalent of  Hull's [1933] "pure stimulus act", Tolman's [1932] "expectancy", Osgood's [1948] "representational mechanism",  etc.

The reason for using the term "cognitive stimulus" rather than such terms as "expectancy", "thought", or "cognition" is that inclusion of the term "stimulus" more strongly implies action.  In the past, cognitive theorists have been usually criticized for leaving their subjects "lost in thought".

Support for the contention that permanent records of previous ex­perience (engrams) are stored in the brain comes from at least two sources. First, there is the work done by Penfield [1954] where he has reported that electrical stimulation of the temporal cortex of humans causes the subject to experience images so vivid that they are difficult to tell from reality. These hallucinations are reenactments of actual experiences from the recent or distant past. ("Both old and recent memories are evoked with equal ease.")

"In general, the recollections produced by stimulation seem to be as clear as they would be seconds after the experience. In fact, they are apparently as clear as they were during the ex­perience... It is an episode in which action goes forward and the patient is an actor. He may seem to see and hear and react as well."
[p. 99]

The work of Penfield ties in with clinical reports that brain tumors in the temporal cortex can also lead to complex and elaborate hallucina­tions [Weinberger and Grant, 1940]. It is proposed that the irritation due to the tumor and the electrical stimulation both serve to trigger the engram which, in turn, leads to the hallucinations.

A second source of evidence in support of the existence of engrams which might be used is the recent work which implicates RNA in the process of memory storage. These studies, summarized in a number of recent articles and books [Brazier, 1964; Landauer, 1964; Gaito and Zavala, 1964;  Jacobson, 1966], suggest that previous experiences are recorded  in the brain by restructuring of the RNA molecule. According to Landauer [1964], for example, when two stimuli, the CS and US, are paired, RNA representing the CS enters the neurons activated by the US. The result of incorporating the new RNA, which represents the CS, alters the recipient. or US cells, so as to make them more likely to fire in the presence of the spreading electrical activity generated by the CS. Thus, the engrams we are talking about could be thought of as the restructured RNA molecules that have entered the neurons normally activated by the US. A cognitive stimulus would be the stimulation propagated by the altered US cells upon stimulation by the CS.

One very important implication from the above engram concept is that all recorded experiences are subject to "replay" if the appropriate engram is triggered. Extinction or forgetting would be explained in terms or an interference hypothesis; that is, "replay" would fail to occur if there were more dominant stimuli present which led to responses incompatible with the response evoked by the CS.  If these competing responses could be eliminated, then the appropriate engram could be triggered (i.e. the appropriate cognitive stimulus could be evoked).

Postulate III.  Conditioning – If an organism attends to two stimuli occurring in close contiguity, these two stimuli will become associated so that upon later occurrence of the first stimulus the reaction to the second will occur.

This postulate is essentially the "S-S Contiguity" interpretation of conditioning with the added stipulation that the organism must be aware of or attentive to the two stimuli. This awareness or attention addendum has recently been shown to be necessary by a number of investigators: Guthrie [1959]; Speilberger [1962] Dulany [1962]; Maltzman [1966]; and Trabasso and Bower [1968]. Thus, according to this postulate, (1) association occurs between stimuli and not a stimulus and response as called for by the S-R approach; and (2) contiguity of the attended stimuli is the necessary and sufficient condition for conditioning to take place and not drive or need reduction as called for by the "Law of Effect" approach. It is the authors opinion that the evidence indicates that this is the more general and parsimonious of the three major systematic points of view that have dominated the psychology of learning (namely, the S-S Contiguity, the S-R Contiguity and the S-R Effect approaches).

As pointed out, the S-S Contiguity approach says first of all that association occurs between stimuli and not between a stimulus and a response. This, of course, does not mean that a stimulus cannot become associated with a response. The S-S position would explain an association between a stimulus, S1 and a response, R2, by positing that the  association takes place between S1 and S2 where S2 is a stimulus which normally evokes R2 . It is felt the S-S position is more general than the strict S-R approach because as well as explaining association between stimuli and responses, it can also explain the formation of associations between stimuli where no visible response is involved. (One of the major shortcomings or the S-R position, we feel, has been that it is more difficult for S-R theorists to conceive of conditioning taking place when no visible response is known to occur.) Evidence in support of the conten­tion that associations can take place between stimuli without necessitating a response comes from a number of areas of study. Among them are: (1) sensory preconditioning, (2) perceptual learning, and (3) learning without overt response. An extensive review of these areas can be found in Kimble [1961].

In addition to saying that associations take place between stimuli, Postulate III states that contiguity of the stimuli in the focus of attention is the necessary and sufficient condition for the association to take place. This is opposed to the "Effect" position which proposes that, in addition to contiguity, some form of drive or need reduction is necessary for the association to take place. Although there is no denying that reward or drive reduction can facilitate conditioning, there is considerable evidence to show that conditioning can, however, still take place without the necessity of drive reduction.

The evidence against a strict "Effect" position comes from several areas of study (also reviewed in Kimble, 1961). These are: (1) the latent learning studies, (2) the saccharine studies, (3) the exploration studies, and (4) the brain stimulation studies, in addition to the sensory preconditioning and perceptual learning studies already mentioned.

Corollary 2 – Whatever would raise the stimuli to be paired in the stimulus dominance hierarchy should facilitate the conditioning.

This follows from the postulate since the latter states that the CS and US must be in the focus of attention to be paired. If there are other, more dominant, stimuli present, this condition will not be met. Thus, anything that would inhibit competing stimuli should facilitate conditioning.

Corollary 3 – Words can act as conditioned stimuli which can evoke cognitive stimuli mediating responses similar to those evoked by the original unconditioned stimuli.

Pavlov was one of the first to recognize that words could act as conditioned stimuli.

"Obviously for man speech provides conditioned stimuli which are just as real as any other stimuli. . . . Speech on account of the whole preceding life of the adult, is connected up with all the internal and external stimuli which can reach the cortex, signaling all of them and replacing all of them, and therefore it can call forth all those reactions of the organism which are normally determined by the actual stimuli themselves."
[Pavlov, 1960, p. 407]

That words can act as conditioned stimuli is supported by a number of experiments. As pointed out by Platinov [l959], Vasileyva found that he could condition a stable defensive motor response to the word "bell".  Hudgins [193]) was able to condition the pupils of his subjects eyes to contract upon thinking the word "contract".  Menzies [1941], by associating the word "crosses", with immersion of the hand in cold water, was able to condition his subjects so that when they said the word "crosses" a drop in the temperature of the hand resulted. This contention is also concurred with by Hull:

"In the suggestion experiments the words of the experimenter presumably are merely performing the function served by the ar­bitrary sounds, temperatures, etc. (conditioned stimuli) in the conditioned reflex experiments."
[Hull, 1933, p.280]

An interesting point to ponder is that the reinforcing effects of the drive reducers (such as food and sex) might themselves be subsumed under a stimulation explanation of reinforcement. This is the case if we consider the possibility that it is the drive reducer’s resulting stimulation of arousal which plays the major role in reinforcement rather than reduction of a drive, per se. This seems to fit in with the position taken by Sheffield's [1966] "Drive Induction" and Miller's [1963] "Go-Mechanism" explanation of drive reduction in conditioning. The reason that most drive reducers can be such effective reinforcers could be that they are stimuli which, due to their high arousal value, would be placed high in a stimulus dominance hierarchy, as well as place any stimulus they become associated with high in the hierarchy.

That words can evoke responses similar to those evoked by the un­conditioned stimuli they are a substitute for is also supported by the available evidence.  For instance, Schultz [1950], Vandell, Davis and Uugston [1943], Max [1937], and Jacobson [1938], among others, "have shown quite satisfactorily that thought can give rise to specific patterns of muscular tension and activity, particularly in those muscles that are symbolically represented in the thought in question" [Weitzenhoffer, 1953, p. 246].  There are also a number of experiments where it has been shown that various physiological and perceptual responses can be evoked by means of waking suggestion. These are best summarized in Barber’s two review articles on the physiological effects of suggestion [1961, 1965]. Among the responses he reports evoked by waking suggestion, we find such things as heart acceleration and deceleration, color blindness, deafness, autonomic changes. salivation, analgesia, and allergic dermatitis. (Heart acceleration, for example, could be produced by words associated with fear-producing stimuli; i.e., by suggesting something fearful.)

Corollary 4 – A reciprocal inhibitory response can be conditioned like any other response if it occurs contiguously with the conditioned stimulus.

First of all, we know that an inhibitory response can be conditioned just like any other response.  For example, Pavlov [1960], referring to experiments in his laboratory by Volborth, concluded that "if an inhibitory stimulus is applied simultaneously and repeatedly for short periods of time together with some neutral stimulus, the latter also develops an inhibitory function of its own" [p.106; see also p.404].

Under Postulate I we saw that when an organism is responding to one stimulus, there occurs a reciprocal inhibition of any stimuli that would lead to incompatible responses.  The case in favor of the contention that this type of inhibitory response can be conditioned is very nicely presented by Wolpe [1958] in his book, Psychotherapy by Reciprocal Inhibition.  For example, among other things, he refers to Pavlov’s experiment where a strong electric current was made the conditioned stimulus for a feeding response in a dog:

"The current was in time gradually increased (with feeding) until it was extremely strong, but even then no defensive reaction was manifested.  In other words, the pathways normally connecting the electrical stimulus with the defense reaction had become inhibited. It would appear that at every stage of the experiment the performance of the feeding response involved a reciprocal inhibition of the mild defense reaction aroused by the electrical stimulus.After many repetitions of the procedure, in the course of which the current was gradually stepped up, so great a degree of conditioned inhibition of the defense reaction to the current was established that even very strong electrical stimuli were unable to evoke that reaction, but evoked only the feeding response."
[Wolpe, 1958, p.30]

The important thing to note here is that in conditioning the feeding response, the inhibitory response – inhibition of the defense reaction – was simultaneously being conditioned.

Wolpe also cites as evidence various experiments done on cats whereby neurotic anxiety reactions were overcome by opposing them with feeding reactions.  To this evidence can be added Watson’s "Peter and the Rabbit" experiments wherein a phobia of rabbits was gradually extinguished by having the child eat his meals in the presence of the feared rabbit [Watson, 1957, pp. 172-175].   Wolpe’s position is apparently supported by at least one learning theorist.  Osgood discussing what he refers to as "an hypothesis of reciprocal inhibition of antagonistic reactions" states:

"Simultaneous with every increment in excitatory habit tendency in the association of a given stimulus with a given reaction, there is also generated an equal increment of inhibitory habit tendency in the association of the same stimulus with the directly antagonistic reaction."
[Osgood, 1948, p. 150]
  In other words, simultaneous with learning any response, the S is also learning not to make the directly antagonistic response.

Corollary 5 – If a set to inhibit certain stimuli is conditioned to a given CS, the presence of this CS will facilitate the occurrence of any response that would ordinarily be interfered with by these stimuli.

This corollary is derived from Postulates I and III.  It is given substantial support by the work of Harlow on learning sets and error factor theory [Koch, 1959].  In a number of experiments he has shown that when monkeys are given a series of different discrimination problems to learn, a "learning set" is gradually established which facilitates the making of new, different, discrimination responses.  (The CS referred to in the postulate in this case would be any stimulus or stimuli which are always present from problem to problem, such as the presence of the experimenter himself.)

Harlow explains this facilitation in terms of learned (conditioned) inhibition.  He proposes a hypothesis similar to Wolpe’s – that in learning to make a particular response the organism learns to inhibit all interfering or incompatible stimuli, or what he calls error factors (EF’s).  In fact, he goes so far as to say that "learning is nothing but suppressions or inhibitions of EF’s" [Koch, 1959, p. 526].  Harlow feels that when the monkey is asked to make a new discrimination response, this learned inhibition of EF’s facilitates the making of the new response.  This is because many of the EF’s inhibited in learning the previous problems are potential interferers of the new response as well.

Also in support of Corollary 5 is the fact (as pointed out by Harlow) that in most learning experiments the investigator often finds it quite advantageous to "adapt" his animals to the experimental situation prior to the start of the learning.  "Psychologists have been doing this for decades, e.g., ‘adapting’ rats on a straight-away before training them on a multiple unit maze, thereby doubtless reducing error-producing factors in advance of the ‘learning’ situation" [Koch, 1959, p. 526]. This adaptation procedure can be looked upon as the establishment of a conditioned inhibition of irrelevant responses.  This conditioned inhibition is evoked in the learning situation by the stimuli that are common to both the adaptation trials and the learning trials.


In the following an attempt will be made to explain hypnotic induction in terms of the principles of conditioning and inhibition outlined in the above postulates.  It will be shown how the hypnotic induction can be explained as the conditioning of an inhibitory set – a set which increases responsiveness to suggestion by inhibiting stimuli (sensory and cognitive) incompatible with the suggestion.  This explanation will then be condensed into three major hypotheses and evidence presented in their support.  Finally, some of the major individual factors that can influence the hypnotic induction, such as prestige, expectation, fears, and age, are discussed and their roles explained in terms of the theory.

The first step in the explanation will be to define the terms to be used, then we shall attempt to fit the HI into a conditioning paradigm.



The definition of suggestion given in Warren’s [1934] dictionary is: "A suggestion is a stimulus, usually verbal in nature, by which an individual seeks to arouse activity in another by circumventing the critical, integrative functions" [p.267].  The following is the definition of suggestion given by McDougall [1908, p.100]:  "Suggestion is a process of communication resulting in the acceptance with conviction of the communicated proposition in the absence of logically adequate grounds for its acceptance."  Hull defines suggestion as:  "A true suggestion response is one in which the subject’s own symbolic process, instead of becoming active either in facilitating or resisting the tendency to action naturally arising from the experimenter’s words, remains passive so far as the particular act suggested is concerned" [Hull, 1933, p.307].

Lindzey [1954, p.27], summarizing a number of definitions of suggestion, states:

"In these and in similar definitions, attention is called to some arbitrary restriction in the determinants of behavior.  The individual is not employing all relevant ideas, nor his full intelligence.  Granted that suggestion proceeds according to the laws of association (conditioning), still we must also allow for the blocking of normal association, so that the end result in behavior is due to a selected field on determinants."

The definition of suggestion which will be used in the present paper is:

"A suggestion is a stimulus or set of stimuli, usually verbal in nature, by which an individual (1) evokes a cognitive stimulus in another, and (2) at the same time evokes an inhibitory set which tends to inhibit stimuli (sensory or cognitive) incompatible with the cognitive stimulus evoked."

The only major difference between this definition and the previous ones mentioned is the addition in the parentheses – that sensory stimuli, as well as cognitive stimuli tend to be inhibited by the inhibitory set.  All of the above definitions seem to stress the inhibition of cognitive stimuli and do not mention inhibition of sensory stimuli.

It should be stressed that both hypnotic and waking suggestion have an inhibitory set component.  The only difference between hypnotic and waking suggestion is that for a given individual, the former should have a larger inhibitory set component as a result of the hypnotic induction.  The size of the inhibitory set for waking suggestion will vary from individual to individual depending on certain factors, such as prestige for example, (these are discussed in a later section).  This means that for a particular suggestion, the response could be greater for one individual in the waking state than for another individual in the hypnotic state.


Hypersuggestibility is defined as "a state where the cognitive stimulus evoked by a suggestion is responded to more readily or strongly than usual because the usually competing stimuli have been reduced or inhibited. (The usual responsiveness to suggestion could be predetermined for each individual.)

There are, of course, numerous ways other than hypnotic induction for bringing about a state of hypersuggestibility.  For example, sensory deprivation is known to lead to such a state [Jackson and Pollard, 1962; Jackson and Kelly, 1962; Pollard, Uhr and Jackson, 1963].  The hallucinogenic drugs (e.g., LSD and mescaline), which act as inhibitors, are also known to produce states of hypersuggestibility [Barrios, 1965; Sjobert, 1965; Solursh and Rae, 1966].


Hypnosis is defined as "a state of hypersuggestibility arrived at by means of a hypnotic induction."

It should be mentioned that the evocation of the cognitive stimulus alone will cause a certain amount of inhibition of competing stimuli just as the evocation of any stimulus would.  However, in a suggestion we find the additional inhibitory "aid" of the inhibitory set.

It is a hypersuggestible state (i.e., more suggestible than normal) because when a suggestion is given, the inhibitory set part of suggestion for a given individual is greater in scope that it is in the normal state.

Hypnotic Induction (HI)

Hypnotic induction is defined as "the giving of two or more suggestions in succession so that a positive response to one increases the probability of responding to the next one."

That the author is not alone in his feelings that a positive response to a series of suggestions or assertions leads to a state of hypnosis is illustrated by the following statements made by Skinner [1957]:

"With respect to a particular speaker, the behavior of the listener is also a function of what is called belief (a term very similar to suggestibility). ...Our belief in what some one tells us is similarly a function of, or identical with, our tendency to act upon the verbal stimuli which he provides. If we have always been successful when responding with respect to his verbal behavior, our belief will be strong [pp.159-160]."

"The listener reacts to the behavior of a given speaker to an extent determined by the consequences of past reactions. The speaker can build confidence or belief by saying many things which are obviously true or quickly confirmed or by resorting to rhetorical devices [p.365]."

"Various devices used professionally to increase the belief of a listener (for example by salesmen or therapists) can be analyzed in these terms. The therapist may begin with a number of statements which are so obviously true that the listener's behavior is strongly reinforced. Later a strong reaction is obtained to statements which would otherwise have led to little or no response. Hypnosis is not at the moment very well understood, but it seems to exemplify a heightened 'belief' in the present sense [p.160]."

From the definition of HI used in the present paper, then, the reader can begin to see the fairly broad scope of the theory of hypnosis presented in this paper. It can not only be used to explain the phenomenal effects of hypnosis, in the accepted sense of the term, but also the hypnotic effects (persuasiveness) of salesmen, lawyers, politicians, etc.; the hypnotic effects (placebo effect) of psychotherapists and doctors of medicine; and even the hypnotic effects (faith) of ministers and faith healers.

This definition of HI does not differentiate between waking suggestions and trance or sleep suggestions. That is, we can conceive of the "formal hypnotic induction" suggestions of eyelid closure, drowsiness, sleep, etc., as just so many more waking suggestions. "Sleep suggestions", however, may additionally further aid the hypnotic induction since the sleep-like state thus produced may provide for even greater inhibition of stimuli competing with the suggestions. As Hull puts it,

"It is a very general custom of hypnotists to give suggestions of relaxation while inducing the trance... The present hypothesis assumes that this relaxation has the effect more or less completely of suppressing the spontaneous activity of the symbolic thought processes. With this suppression should disappear the control normally exercised by symbolism over the lower levels of activity. This should leave the latter more completely exposed to the influence of suggestive stimuli from outside sources."
[Hull, 1933, p.310)

It should be stressed that in the present theory sleep suggestions are not a necessary condition for hypnotic induction. Thus, the use of the term "hypnotic", which means "tending to produce sleep", is perhaps misleading and it might be appropriate to eventually change it.

Hypnotic Induction in a Classical Conditioning Paradigm

In this section, we will attempt to show how the hypnotic induction is actually a conditioning process.

Understanding the Conditioning Paradigm

Before we show how the hypnotic induction fits into the conditioning paradigm we must first be sure we understand the conditioning paradigm. First of all, as pointed out in the conditioning postulate, for a process to be called conditioning it must involve two stimuli presented together contiguously and in the focus of attention. In classical conditioning the two stimuli are usually referred to as the CS and the UCS. The CS is usually some neutral stimulus (i.e., no observable response is evoked or at least not the response to be conditioned) and the UCS is a stimulus which evokes some innate response (e.g., food-salvation; shock-withdrawal). However, and this is an important point to keep in mind, there is nothing that says that the UCS has to evoke an innate response. The UCS, or second stimulus in the pair, can be one that evokes a learned or previously conditioned response. In classical conditioning, this is referred to as higher-order conditioning and as Hebb [1949] has pointed out, most conditioning in the mature organism is of this higher-order variety.

Another thing to keep in mind is the nature of the response conditioned (the CR). We know that in a conditioning situation the experimenter (E) is not always interested in the entire response to the UCS. He usually focuses on one component of the UCR which he is interested in associating with the CS. Usually, this component is some positive response (e.g., salvation, eye-blink, withdrawal, etc.). However from our reciprocal inhibition postulate, we know that occurring with each positive response there is also a reciprocal inhibitory response. Now, in applying the conditioning paradigm to hypnotic induction, we will be focusing on the inhibitory component rather than the positive component.

Finally, a third thing to keep in mind is that the CS need not be something as obvious as a bell ringing, but can also be the very presence of the experimenter and  any action on his part which is repeated prior to each time the UCS is presented.

The Classical Conditioning Components in the HI
    Conditioned Stimulus (CS) – The general stimulus situation in which the hypnotist makes a suggestion during a hypnotic induction.

    Unconditioned Stimulus (UCS) – The particular words of the hypnotist evoking the cognitive stimulus which leads to the particular suggested response. This UCS, of course, is a "second-order" UCS in that it evokes its response as a result of previous conditioning and not innately.

    As an illustration of the distinction between this CS and UCS (since at first glance they do appear to be indistinguishable), let us take the suggestion, "You will now feel your arm pulled up." The CS would be the first part of the suggestion; i.e. the hypnotist saying the words: "You will now... " which essentially would precede each suggestion in the hypnotic induction series. The UCS would be the specific suggestion "...feel your arm pulled up." The CS is always the same. The UCS varies with each suggestion.

    Unconditioned Response (UCR) – The reciprocal inhibition of stimuli incompatible with the particular suggested response. This inhibitory response occurs automatically when the suggested response occurs. Three of the major classes of stimuli inhibited would be the following:

    1. specific stimuli in direct contradiction to the suggested response. For example, if it is suggested that the subject see a watermelon on a table that is actually empty, the sensory stimulus "empty table" would be in direct contradiction to the image "watermelon on table;
    2. general irrelevant but attention-catching stimuli like a door slamming or an itch; and
    3. certain general negative attitudes (e.g., skepticism) and fears (e.g., fear of the unknown) which interfere with the conditioning during hypnotic induction in a number of ways. (This will be discussed further in a later section.)

    Conditioned Response (CR) – The inhibition of stimuli incompatible with the previous suggested responses. This CR is evoked upon the presentation of the CS, i.e., when the hypnotist makes a suggestion. It can be measured directly in terms of inhibition of competing stimuli [e.g. Hernández-Peón, et al, 1960] or it can be measured indirectly by measuring the increase in strength of the suggested response, since according to the stimulus dominance hierarchy postulate, by inhibiting competing stimuli the inhibitory set would increase response to the suggestion.

    It could very well be that this conditioned inhibitory response will not appear to be a discrete one, that is, evoked only when the hypnotist makes a suggestion, but may instead appear to be a continous one, present even between suggestions. This would best be explained in terms of generalization. The inhibitory response, rather than become associated specifically with the CS we have designated (i.e., the hypnotist making a suggestion), may also become associated with other components of the whole hypnotic setting as well.

    Trial – That period during which a suggestion is made and the suggested response occurs. Each trial need not necessarily be thought of as involving a different type of response. When a suggestion is repeated (and each suggestion is usually repeated a number of times in most hypnotic inductions), a trial can be thought of as taking place with each repetition. Each repetition would be expected to lead to a stronger response and thus a greater inhibitory set.

    Reinforcement – A contiguous point of view. As pointed out in Postulate III (Conditioning), all that is necessary for conditioning to take place is that the two stimuli (the CS and the UCS) occur contiguously and in the focus of attention.
Analysis of the HI in Terms of the Classical Conditioning Components

Let us now look at the hypnotic induction in the above terms.

First suggestion (first trial). When the initial suggestion is responded to (i.e., when the first presentation of the UCS leads to its response), a certain proportion of stimuli incompatible with the suggested response are reciprocally inhibited (the UCR). This inhibition is conditioned to the hypnotist making a suggestion (the CS) so that when the hypnotist makes his next suggestion (when the CS is presented again) a conditioned inhibitory response (the CR) is evoked; i.e., those stimuli inhibited during the response to the first suggestion are now automatically inhibited prior to the response to the second suggestion.

Second suggestion (second trial). Since a certain proportion of competing stimuli will now automatically be inhibited when the hypnotist makes a suggestion the probability of evoking the next suggested response is increased. When this second suggestion is positively responded to, another proportion of incompatible stimuli will be inhibited. As was the case in the first trial, this inhibitory response becomes conditioned to the CS so that on the succeeding trial an even greater proportion of competing stimuli arc now inhibited, thus further facilitating the response to suggestion.

Succeeding suggestions (succeeding trials). Each succeeding suggestion positively responded to will now add to the proportion of incompatible stimuli inhibited until, conceivably, a point is reached where all incompatible stimuli are automatically inhibited when the hypnotist makes a suggestion. (It is as if S's focus of attention were gradually being narrowed more and more until the only stimuli reacted to were the cognitive stimuli evoked by the hypnotist.) This would be called a "deep" state of hypnosis.

The hypnotic induction is ended by suggesting that S will "awaken" or come back to normal, whereupon the inhibitory set which has been developed during the induction becomes inhibited itself. That the hypnotic conditioning can be so quickly "extinguished" by such a suggestion should not be so surprising to those who are familiar with the part that cognitive factors can play in extinction [Spence, 1963].

It should be mentioned that very often in a hypnotic induction, after getting a positive response to a suggestion, the hypnotist will then remove the suggestion. For example, after getting a positive response to the suggestion, "Your arm will become as stiff and rigid as a steel bar", the hypnotist usually follows with, "Your arm will now loosen up and is now no longer rigid." Knowing this, one might be inclined to think that such removal or reversal of the suggestion would undo the conditioning of the inhibitory set accomplished by that suggestion. Actually, this apparent paradox can be explained if we realize that most of the stimuli that would have to be suppressed in order for the first suggestion ("stiff arm") to be responded to would be the same as those that would have to be suppressed for the second suggestion ("loose arm"). For example, there is the attitude of skepticism (a cognitive stimulus), the suppression of which would facilitate both responses.

This paradox of reversal learning has been explained in a similar manner by Harlow and his Error-Factor theory. As Harlow points out, "discrimination learning and reversal are not antithetical processes. So far as EF elimination is concerned, the two learning problems have more in common than at variance." [Harlow, 1959, p.522] If we under stand Harlow correctly, error-producing factors (EF) can be looked upon as what we refer to as incompatible stimuli or stimuli that lead to wrong responses. Thus, he is saying that the discrimination and discrimination reversal problems, although involving apparently opposite responses, require suppression of many similar incompatible stimuli.

Instant Hypnosis

Very often we see or hear of a hypnotist putting a subject under hypnosis instantaneously by merely snapping his finger or whispering certain key words in the subject's ear. This type of instant hypnosis most likely occurs as a result of the subject being previously hypnotized and given the post-hypnotic suggestion that he will go back under hypnosis upon a given cue. The inhibitory set developed during the hypnotic induction is conditioned to this cue by means of this post-hypnotic suggestion (see section below on post-hypnotic suggestion) and thus re-evoked when the cue is presented.

Instant hypnosis, however, can also occur without the subject directly being hypnotized previously by the hypnotist. This can happen when the subject has heard of or seen the hypnotist's great effectiveness and believed it. In a sense, this subject is already hypnotized or conditioned by having heard or seen the positive responses achieved by the hypnotist, (This is discussed further in the section on prestige, below.)


A self-hypnotic induction can be explained in much the same way as the hypnotic induction above. The only major difference in this case is that the subject plays the part of both the hypnotist as well as the subject. To put it in terms of the conditioning paradigm, the only major difference is that in self-hypnosis the CS is the following stimulus situation: "the subject giving himself a suggestion during a hypnotic induction" rather than "the hypnotist making a suggestion during a hypnotic induction".

There is also instant self-hypnosis. It is explained in the same way as the first type of instant hypnosis, except that the cue the subject is given is a self-induced one. He is also told that in this state he will be able to give himself suggestions in the same way as the hypnotist has, including the suggestion to awaken.

Explanation of HI in Three Hypotheses

The above explanation of hypnotic induction can be condensed into three major hypotheses. In this section we shall look at the evidence in support of these hypotheses, as well as suggest further tests.

Hypothesis I: Hypnotic induction is a conditioning process.

Hull and his coworkers provide a good portion of the evidence bearing on the validity of this hypothesis. In his chapter, "Hypnosis Regarded as Habit", Hull [1933] proposes that if hypnosis results from a conditioning process, or is a "habit phenomenon" as he prefers to refer to it, it should display certain characteristics common of habit phenomena. (We are making the assumption that habits are types of conditioned responses.) Hull points out six such common characteristics:

  1. practice in an act facilitates its subsequent performance;
  2. the rate of facilitation in the practice activity is more rapid early in the practice than later;
  3. a period of disuse is followed by a partial loss of the facilitation resulting from practice;
  4. other things being equal, the loss of facilitation following disuse is less where the original repetitions were widely spaced than where closely spaced;
  5. with the resumption of practice this is recovered, the new practice curve showing a picture of negative acceleration; and
  6. the amount of loss resulting from disuse is recovered with less practice than was required for its original acquisition.

In two experiments by Kreuger [1931], one of Hull's associates, it was found that hypnotic induction did indeed display the above characteristics.

Hull also presents as corroborative evidence a number of experiments showing that waking suggestion (which in the present paper is felt to be essentially equivalent to hypnotic suggestion) has "habituation effects parallel to those shown by Kreuger to be characteristic for the process of hypnotization" [Hull, 1933, p.343]. These include the experiments of Hull and Huse [1930], Williams [1930], and Patten, Switzer and Hull [1932].

In concluding the chapter "Hypnosis Regarded as Habit", Hull states:

"Such a remarkable and detailed conformity of the phenomena hypnosis to the known experimental characteristics of ordinary habituation can hardly be accidental and without significance. The indications would seem to be that whatever else hypnosis may be it is – to a considerable extent, at least – a habit phenomenon and that quite possibly this hypothesis may furnish the basis for an ultimate understanding and explanation of its hitherto largely inexplicable characteristics."
[Hull, 1933, p.347]

The reader might question the use of Hull's evidence in support of Hypothesis I on the grounds that the latter hypothesis is referring to the conditioning of "an inhibitory set", not a particular response as proposed in Hull's postulate. Actually, there is no real discrepancy here. The author is of the opinion, like Harlow and Osgood, that

"Simultaneous with every increment in excitatory habit tendency in the association of a given stimulus with a given reaction, there is also generated an equal increment of inhibitory habit tendency in the association of the same stimulus with the directly antagonistic reaction."
[Osgood, l948, p.150]
and, thus, that both are directly related.

Hypothesis II: The response conditioned during the HI is an inhibitory set, a set which tends to inhibit stimuli incompatible with the response suggested by the hypnotist.

This hypothesis is, of course, founded to a great extent on Corollary 4 of the postulates, which states that inhibitory responses can be conditioned. There are a number of corollaries to Hypothesis II; we shall look at three of them.

Corollary 1: Reactivity to competing stimuli should decrease as the HI progresses.

This is what we would expect if HI involves the conditioning of an inhibitory bet, and consequently the gradual increasing of the inhibitory set as the HI progresses. The results of an experiment by Hernández-Peón , et al [1960] appears to support this corollary. These investigators found that the HI resulted in the diminution of the forearm flexor response elicited by tactile stimulation. (This tactile stimulation can be looked upon as a competing stimulus.) It is interesting to note their interpretation of the results:

"All procedures for inducing hypnosis require the focusing of attention upon the experimenter's verbal stimuli with obliteration of irrelevant stimuli (competing stimuli) It does not seem unreasonable that during the hypnotic state itself, the thresholds of perception for modalities other than the auditory are raised by the postulated centrifugal sensory inhibition at lower levels of the central nervous system...

"Our results fall in line with the above mentioned hypothesis although sometimes the size of the skin reflex remained unchanged during hypnosis, in most of the experiments a more or less intense diminution was induced by the hypnotic state per se...  Our results agree with those of Scars [1932], Dynes [1932], and Doupe et al. [1939], who recorded during hypnotic anesthesia partial, but significant reduction of non-voluntary physiologic reactions to pain such as the psycho galvanic reflex, respiratory, cardiac, and vasomotor responses."
[pp.37 and 40]
Corollary 2: Increasing the inhibitory response paired with the CS on each trial should facilitate the conditioning (facilitate the hypnotic induction).

This corollary would help explain how sleep suggestions facilitate the hypnotic induction. Sleep suggestions would produce an added inhibitory response which would summate with the one produced by a positive response to a suggestion, resulting in stronger inhibitory responses with each suggestion.

The point that sleep suggestions could evoke sleep-like responses was brought out by Pavlov [1960] when discussing methods of hypnotic induction:

"At present the more usual method consists in the repetition of some form of words, describing sleep, articulated in a flat and monotonous tone of voice. Such words-are, of course, conditioned stimuli which have become associated with the state of sleep. In this manner any stimulus which has coincided several times with the development of sleep can now by itself initiate sleep or a hypnotic state."
[pp.404, 405]

It should be mentioned that there are at least two possible difficulties with the use of sleep suggestions in a hypnotic induction. First, there is the possibility that the sleep suggestions might be too effective; i.e., the subject might literally fall asleep, which would mean a loss of contact between the hypnotist (H) and the subject (S). One possible way to avoid this is for H to suggest that S will continue to hear his voice even though going into a "deep sleep". The other difficulty is that the sleep-like state produced might interfere with the responsiveness to other suggestions, especially any requiring an alert, awake state. This difficulty might be avoided if prior to such suggestions H suggests that S will be wide awake and very much alert although in a deep state of hypnosis.

Corollary 2 would also explain why devices which focus S's attention will facilitate the induction. As Hernández-Peón [1959] and others have shown, a major component of the attention response is the inhibition of competing stimuli. Any means, then, whereby H can get S to focus his attention (such as having S look at a shiny crystal or a spinning spiral) should add to the inhibitory response and thus aid the HI.

Corollary 3: The more stimuli brought under the control of the inhibitory set with each succeeding trial (with each suggestion), the more effective the HI and thus the greater the increase in general suggestibility.

From this we can make at least two predictions:

  1. The greater the overlap among the stimuli incompatible to two suggestions, the greater will be the increase in response to the second suggestion when the first one has been positively responded to. This, of course, can be stated another way – the more similar the second suggestion to the first, the greater the increase in responsiveness to the second suggestion when the first one has been positively responded to.

  2. A hypnotic induction which only involved getting a positive response to the same suggestion over and over again, or to a series of similar suggestions would not lead to as great an increase in general suggestibility as a hypnotic induction which involved getting a positive response to a series of different suggestions. (Increase in suggestibility would be measured by the increase in responsiveness to a set of suggestions different from those used in the induction.)

Parenthetically, it should not be inferred from this corollary that for a suggestion to be positively responded to, all the stimuli competing with suggested response must be covered by the inhibitory set. One should not forget that it is a combination of the inhibitory set and the strength of the cognitive stimulus evoked which determines the probability of a positive response to a suggestion. Also, there can be degrees of positive response, depending on how many competing stimuli have been inhibited.

Hypothesis III: A positive response to a suggestion will induce within the responding person a more or less generalized increase in the normally existent tendency to respond to succeeding suggestions.

Hull [1933, pp.313-314] points out two studies in support of this hypothesis [Caster and Baker, 1932; and Jennes, 1933]. Caster and Baker

"took with a stop-watch the suggestion times required to produce lid-closure in ten subjects,
  1. preceding positive response to arm suggestion, and
  2. following such response.
Under these conditionsm the hypothesis demands that the trance induction time following the arm movement should upon the whole be less than that preceding the response to arm suggestion."
This hypothesis was confirmed in seven out of ten subjects. Jennes [1933], repeating this experiment with some improvements, confirmed the hypothesis in eight of nine subjects.

The above findings seem to fit in with the reports of a number of investigators:

"Schilder and Kauder [1927], for instance, have pointed out that a number of earlier investigators as well as they themselves had found it efficacious to give their subjects some waking suggestions easy of execution before attempting to hypnotize them. This appeared to facilitate the process."
[Weitzenhoffer, 1953, p.40]
Opposed to Hypothesis III we find the results of two of Hull's own experiments [Hull et al., 1933; Fatten et al., 1932]. In these experiments he attempted to see if a positive response to one waking suggestion would increase the responsiveness to a second waking suggestion. His results indicated no such increase, thus contradicting the hypothesis. In commenting on the apparent contradiction between these two experiments, and the Caster and Baker and Jennes experiments, Hull states:
"It is difficult to reconcile the results of the two groups of investigations. The problem involved is of such central importance that the experiments should be repeated with judicious variations to make certain whether or not some hidden defect of technique may not have produced this seeming inconsistency, particularly between the study of Jennes [1933] and that of Hull et al. [1933]...–No question in the whole subject of hypnosis and suggestibility is in such urgent need of critical experimentation."
[Hull, 1933, p.393]

A thorough analysis of the two negative experiments by hull et al. [1933] and Patten et al. [1932] uncovers what well might be a "hidden defect of technique". Both experiments used the following two suggestions:

  1. that the head would fall forward on the chest, and
  2. that the arm which was suspended by a special device, would sway forward (from side to front).

The order of these suggestions was, of course, alternated. In both the Hull et al. and Patten et al. experiments, it was found that giving the head suggestion first actually seemed to interfere with the response to the subsequent arm suggestion rather than facilitate it. This, of course, was in direct contradiction to the hypothesis that positive response to direct suggestion, as such, evokes a generalized hypersuggestibility. The possible defect of technique that might explain such contradictory results is the distinct possibility that arm movement might naturally be interfered with more when the head is in the bent-over-on-chest position than when upright. To illustrate this the reader need only try it himself. If he holds his arm extended then lets his head come down, he will notice that as the head comes down, the arm will also come down a bit as the center of gravity is shifted. This greater pull downward on the arm might explain the longer time for it to sway forward in response to the arm suggestion. This possible explanation is given further support when we see that no such interference with the head suggestion was reported by Hull et al. [1923] when the arm suggestion was given first, and what is more, Batten found that when the arm suggestion was given first, the head suggestion tended to be facilitated (seven out of ten S's).

The following are a number of corollaries to Hypothesis III, a test of which would be a further test of the hypothesis:

Corollary 4: If a positive response to a suggestion will increase responsiveness to the next suggestion, then a negative response should decrease responsiveness.

This can easily be tested by giving the subjects in one group one or more very difficult suggestions (i.e.. suggestions with a low probability of being initially evoked) prior to a test suggestion and comparing their response to this test suggestion with that of a second group that does not first get the difficult suggestions.

Corollary 5: If a series of suggestions are given to the subject, the probability of inducing a state of hypnosis will be greater if the suggestions are given in a gradually increasing order of difficulty than if they are given in random order.

This would be predicted since the probability of a positive response to any individual suggestion, according to the theory, is greater the greater the number of positive responses to previous suggestions, and this is maximized if the suggestions are given in gradually increasing order of difficulty. Each positive response increases the probability of success in the following suggestion.

It is interesting to look at some of Barber's results in this light [1961, 1965]. In a number of studies he has shown that a good many hypnotic phenomena can be gotten by means of waking suggestion, without the need of a "hypnotic induction". (What he considers a hypnotic induction is mainly the giving of sleep or relaxation suggestions.) However, if one looks at Barber's studies, one finds that he usually gives his waking suggestions in a series (sans the "hypnotic induction") usually in gradually increasing order of difficulty. It might very well be that part of the reason for the high percentage of hypnotic-like responses he reports is that he has actually put some of the S's through a hypnotic induction, at least according to its definition in the present paper. In other words, responding positively to the previous suggestion(s) increases the probability of responding to the next one.

Corollary 6: A hypnotic state can be facilitated if, along with each of the first few suggestions given in the hypnotic induction, the actual sensory stimuli which would ordinarily evoke these suggested responses accompany the suggestions without the subject's knowledge.

This should be done in such a way that the subject is not aware of the "artificial" cause of the responses so that he will ascribe them solely to the hypnotist's suggestions. It is felt that this can be done, if the sensory stimulus is kept close enough to threshold, almost the point of being sub-threshold. The response to the sensory stimulus would summate with that to the cognitive simulus (evoked by the suggestion) to lead to the overall response.

The response could be "artificially" induced in a number of ways. For instance, the suggestions that the eyes are going to get tired can be helped if a slight strain is placed on them by having S look at an object at a difficult angle. Or the suggestion that S was going to feel cold could be reinforced by actually lowering the temperature of the room somewhat.

It is interesting to know that an experiment somewhat along these lines has already been run. This is the "abstract conditioning" experiment of Corn-Becker et al. (1949), where a series of "artificially reinforced" suggestions led to a definite positive response to a subsequent "nonreinforced" suggestion. The main difference between Corn-Becker's experiment and the type proposed by the corollary is that Corn-Becker did not attempt to keep the reinforcing stimuli close enough to threshold, and thus the S's were quite aware of the external cause of the responses and did not think that they were due solely to E's suggestions. One might say that in Corn-Becker's experiment any conditioned inhibitory set was conditioned to the stimulus pattern "Experimenter plus External Stimulation", whereas in the experiment proposed by the corollary the CS would be just "Experimenter" (or Hypnotist). The likely reason that Corn-Becker did get a positive response to the "nonreinforced" suggestion at the end of the series was because of stimulus generalization; i.e., having been conditioned to a stimulus pattern (Experimenter plus External Stimulation), S will still respond to a component of that pattern (Experimenter). However, it would be predicted that such a generalized response would not be as strong (both in terms of amplitude and ease of extinction) as one where the CS is the same one used during the conditioning. Nevertheless, this approach to facilitating HI bears further investigation. It might have the advantage over the "sub-threshold approach" of not attempting to deceive the subject. If deception were detected by a subject, it could produce irrevocable damage to his attitude toward hypnosis.

Individual Difference Factors Influencing Hypnotic Induction

In this section we shall list some of the individual difference factors that have been purported to influence HI and attempt to show how their role can be explained in terms of the theory.

Voluntary Attention (Concentration)                                   

That people differ in their ability to voluntarily suppress irrelevant stimuli (ability to concentrate) is fairly obvious. We also know that the greater the ability of S to concentrate on H’s suggestions the greater the probability of responding positively to them.  Thus, we would predict that the greater a subject’s ability to concentrate, the greater the probability of success of the HI.


It is fairly well accepted that the more prestige the H has in the eyes of the subject the better his chances of success.  It is felt that this is so because the statements, commands or suggestions of a person with prestige tend to be questioned less; i.e., such a person evokes a greater inhibitory set to begin with.  People, in general, have been previously conditioned to accept at face value the statements of someone who is an authority in his field.  That is, an inhibitory set which inhibits contradictory stimuli has been previously conditioned (much in the same way as in the hypnotic induction process).  This is so because what the authority says has usually turned out to be true.  The more a person is known to be an expert the greater the unquestioning acceptance of his statements (i.e., the greater the inhibitory set).  Thus, when a person hears of a particular hypnotist’s great successes (or witnesses them), this prestige factor or inhibitory set is evoked by the hypnotist and thus aids the HI.  This could be tested by seeing the effect of hypnotizing a shill or previously determined good subject in front of the prospective subject.  At least one experiment has supported the contention that prestige can increase the chances of success in hypnotic induction, that of Das [1960]. It is felt that if a waking suggestion is effective, it is often, to a great extent, dependent on this prestige factor.

Related to the prestige effect is the concept of transference as it is used by psychoanalysts in explaining hypnosis.  The latter feel that one of the major factors involved in a hypnotic induction is a transference effect; that is, the subject in going under hypnosis is actually reacting to the hypnotist as he would to his parents when a child.  The prestigeful influence of the parent is "transferred" to the hypnotist during the process of hypnotic induction according to this view.  It could be that certain characteristics of the hypnotist and his hypnotic induction might, through generalization, evoke the child-parent type of prestige effect which could then interact with the hypnotic induction to facilitate it.  How big a factor it will play in the hypnotic induction will be determined by how similar to his parents the subject perceives the hypnotist as being.  It should be pointed out that this transference could also hurt the hypnotic induction if the patient had developed a negative reaction pattern to his parents.

Subject’s Expectation

The expectation of being hypnotized can have a number of different effects on the hypnotic induction.  First of all, there may be a negative effect produced if the subject has a great fear of hypnosis.  Or, conversely, if he is an adventuresome person and enjoys exploring the "unknown", it may have a positive effect.  (These effects will be discussed below in the section on fears.)

A second effect that expectation can have is that of focusing S’s attention on the appropriate CS-CR contingency.  That is, as a result of the expectancy of being hypnotized, the S may be more likely to correctly ascribe the occurrence of the "strange" phenomena to the hypnotist than to some external cause.  For example, if the suggestion of coldness is given in a non-hypnotic setting, S is more likely to ascribe any subsequent feeling of coldness to the possibility that someone has lowered the temperature of the room.  If he doesn’t associate the positive response (and thus the inhibitory response) with the hypnotist (the CS), then the desired conditioning does not take place.  The idea that focusing attention on the contingency would facilitate conditioning is given considerable support by recent studies on the effect of awareness on learning [e.g., Eriksen, 1962].  These studies indicate that the more the subject is aware of the correct contingency the better the conditioning if his attitude is positive.

 Parenthetically, it might be mentioned that if it is important that the responses be ascribed to the appropriate CS (the hypnotist or "Experimenter", we should then expect that a subject will be less likely to be hypnotized if he performs the suggested response voluntarily just to please the hypnotist.  If he does this, he would not ascribe the responses to the hypnotist but to his own volition, and thus would defeat the purpose of the HI.

A third way expectation can influence the hypnotic induction is if S has a certain preconceived notion of what hypnosis is like.  One example of how this can have a negative effect on the hypnotic induction is where the subject incorrectly expects that hypnosis is some state of unconsciousness or sleep.  When he finds himself still aware of things or awake, he thinks that he is not hypnotized.  This negative thought can, of course, have a braking effect on the hypnotic induction.  Such a negative effect can be eliminated, it would seem, by means of a pre-induction talk where S is told that hypnosis is not a state of sleep nor unconsciousness and where he is given some idea of what to actually expect.  Another expectation that can hurt a hypnotic induction is one where the subject believes that being hypnotized means responding to every suggestion in the HI.  If a person fails to respond to one for some reason, even after he has successfully responded to the previous ones, he may decide that he is no longer hypnotized and at that point may stop responding to all succeeding suggestions.  To eliminate this, as part of the pre-induction talk, the subject should be told that because of individual differences, there may be some suggestions that work very well for some people but not for others, and, therefore, it should not bother him if he doesn’t respond to a suggestion; in such a case, he should just wait for the next one.

A subject’s expectation of what hypnosis is like can influence the hypnotic induction in other ways.  For example, if a subject is told that catalepsy of the dominant hand occurs when one goes under hypnosis [Orne, 1959], then as S feels himself being put under, he is also lndirectly being given the suggestion of catalepsy of the dominant hand. This response can, in turn, influence the hypnotic induction, as do any positive responses to previous suggestions.

This last aspect of expectation has played a major role in the "goal-directed" [White, 1941] and "role-playing" (Sarbin, 1950) theories of hypnosis.  These theories state in essence that hypnotic behavior is a meaningful, goal-directed striving, its most general goal being to behave like a hypnotized person as this is continuously defined by the hypnotist and understood by the subject.

In two recent experiments by Barber and Calverley [1964b, 1965], the overall effect of expectation of hypnosis was found to have a positive effect. They found that higher scores on the Barber Suggestibility Scale are obtained when the experimental situation is defined as hypnosis rather than as control.


Fear of the unknown, mistrust of the hypnotist, fear of revealing "inner secrets", etc., are all examples of fears that can interfere with the hypnotic induction. Because of such fears, any positive response to suggestion during the hypnotic induction would be fear-producing and, thus tend to be avoided or suppressed, and in this way, interfere with the conditioning of the inhibitory set. (A pre-induction talk aimed at allaying S's fears should facilitate the hypnotic induction.) Conversely, an attitude of adventuresomeness on the part of the S would probably lead to facilitation of the hypnotic induction. That this seems to be the case is given support by the work of Hilgard [1967]. He reports a positive correlation between adventuresome and hypnotic susceptibility.


It is fairly obvious that if a subject is not very much interested in being subjected to a hypnotic induction, his chances of positively responding are much less. In one study by Barber and Calverley [1966] it was shown how important the factors of boredom and disinterest are in affecting S's responsiveness to a hypnotic induction. By repeating the same monotonous HI procedure over a period of eight days, the experimenters produced a significant drop in suggestibility.


The fact that certain attitudes can influence the HI is fairly well accepted [Weitzenhoffer, 1953, p.283]. For example, in a recent study [Dorcus, 1963] it was found that indications of fear are considered a good sign by some hypnotists [e.g., Gindes, 1951] since this would probably indicate a high degree of expectancy of being hypnotized. However, all things being equal, i.e., if people are matched on degree of expectancy of being hypnotized, it would be predicted that the greater the fear, the greater the interference. And that by means of a pre-induction talk aimed at eliminating the interference of certain known negative attitudes (these can be thought of as competing cognitive stimuli), the experimenter was able to hypnotize the six presumably unsusceptible subjects. In this regard, we can also consider the work of Barber. In one study, for example, Barber and Calverley [1964a] found that by manipulating S's attitude towards the task at hand (a suggestibility test) they were able to significantly affect S's suggestibility score in the expected direction. Pre-test instructions aimed at producing a negative attitude led to a considerable drop (almost down to 0) in responsiveness.

We will now discuss three typical attitudes that might influence the induction and attempt to explain how in terms of the theory.

One negative attitude which often interferes with hypnotic induction can be called, for want of a better term, "non-submissiveness", or a strong desire to be in control of oneself all the time. This would lead to a decrease in the probability of a positive response to the suggestions of the hypnotist, and thus interfere with the HI, since the desire for self-control would be incompatible with giving up control to the hypnotist. This interference can best be eliminated if the hypnotist does not take the authoritarian attitude as his approach. Many people have the misconception that hypnosis invariably means losing control to someone else. This need not be the case. It can be worked so that the only person gaining greater control over the subject is the subject himself. This can be done by instructing the subject during the pre-induction talk that the hypnotist should be looked upon as an instructor who is merely showing him a technique whereby he can gain greater control over himself, greater control over his involuntary side. He is told that he can reinterpret each suggestion given by the instructor by replacing "you" with "I". For example, if the instructor tells the subject "you will now feel your arm being pulled up", the subject can be telling himself, "I will feel my arm being pulled up." He can also be told that he does not have to respond to a particular suggestion if it annoys him for some reason, and that he can always feel free to "awaken" at any point if he so wishes.

Still another way of eliminating this problem might be to let the subject hypnotize himself completely. Instead of the H giving S the series of suggestions, S could give them to himself after learning what the series is. This method might have other advantages as well in that S could set his own pace. He would go on to the next suggestion only after succeeding on the previous one. Often, a hypnotist may not spend enough time on a particular suggestion, or, conversely, he might spend too much time on it. (In self-hypnosis, as in hetero-hypnosis, a response is usually considered positive only if it occurs involuntarily.)

Another common interfering attitude is what can be called the "rational" attitude. Too strong a desire to have a rational explanation for everything can lead the S to ascribe any positive response to "more rational causes" than the hypnotist's suggestions; e.g., "My arm came down because it would naturally get heavy being in such a position for so long", or "My eyes closed because they would naturally do so after staring so long", etc. And as already pointed out, it is important that the S be focused on the correct CS-CR contingency; that is, he should ascribe a positive response to the hypnotist's suggestions. This negative attitude can best be eliminated if the subject is given a reasonable explanation for hypnosis in the pre-induction talk. For example, it should be stressed that no trickery need be involved in hypnosis; that words can actually evoke such phenomena naturally; that hypnosis is a state where words have a much greater effect because of the highly increased concentration, etc.

A third common attitude which can interfere with the hypnotic induction is an attitude of skepticism. Skepticism can be thought of as a conditioned set to inhibit response to suggestions and is felt to be a result of a conditioning process where suggestions have been negatively reinforced somewhere in the past of the subject. Such an inhibitory set would naturally interfere with a positive response to H's suggestions and thus interfere with the conditioning process taking place during the HI. Elimination of this skeptical attitude can be helped by making sure that the initial suggestions in the HI have a high probability of success (e.g., the "chevreul pendulum" suggestion [Weitzenhoffer, 1957]).


We would expect that the greater a subject's imagination, i.e., the greater the ease of evoking vivid imagery to begin with, the greater the probability of responding to suggestions and, therefore, the greater the susceptibility to hypnotic induction. This seems to be supported by the available evidence. Hilgard [1967], for example, reports positive correlation's between childhood fantasy and involvement in reading, and hypnotic susceptibility.

 However, there may be limits to how great a part an initial vivid imagination may play, for a subject may conclude that the positive response to a suggestion is due to his own vivid imagination and not the hypnotist's doing. He might remember that he has been able to evoke similar responses when using his own imagination. And, as has been mentioned, it is important that S ascribe the positive responses to the hypnotist.


According to Weitzenhoffer's review of the area, suggestibility and hypnotic susceptibility at first increase with age, reaching a peak at the ages of seven to eight, then decreases gradually to the age of twenty, when it begins to level off [Weitzenhoffer, 1953, p.76]. The reason that suggestibility varies in this way with age can be traced to certain factors that vary with age. One of these is language ability. Since hypnosis is dependent to a great extent on the conditioned response evoked by words, we can understand why very young children whose language ability is not yet well developed would make very poor subjects, and thus, why we would expect an initial gradual increase in suggestibility with increasing age.

An explanation for the gradual decline in suggestibility after the age of eight is that with continued increasing age the number of cognitive stimuli competing with a suggestion increases (i.e., knowledge increases with age), and a corollary to the Reciprocal Inhibition or Stimulus Dominance Hierarchy postulate is that the more stimuli in the hierarchy, the lower the probability of any one of them being reacted to. These competing cognitive stimuli develop in a number of ways. First, we know that with increasing age the number of possible cognitive stimuli evoked by a word increases.  For example, the word "house" has been associated with more and more houses as the person gets older; thus, if the suggestion is given, "You will now see a house", there will be competition among the many "house" images, with a resultant weakening of any one finally singled out.  Also, with increasing age there will be a greater number of possible contradictory cognitive stimuli evoked by a suggestion; i.e., the S has more information available with which to verify or contradict the suggestion.  Finally, there is the fact that with increasing age there is the development of skepticism.  As pointed out above, skepticism can be thought of as a conditioned process where suggestions have been negatively reinforced; i.e., people learn in time that not everything anyone says is true.


To summarize the above briefly, an attempt has been made to explain hypnotic induction as the conditioning of an inhibitory set – the set to inhibit stimuli (sensory or cognitive) incompatible with a suggestion given by the hypnotist. The hypnotic induction procedure, defined as the giving of two or more suggestions such that a positive response to one will increase the probability of responding to the next one, was placed in the conditioning paradigm, with the CS, UCS, UCR, CR, trials and reinforcement clearly delineated.

The explanation of hypnotic induction was put in the form of three major hypotheses:

  1. hypnosis is a conditioning phenomenon;
  2. the response conditioned during hypnosis is an inhibitory one; and
  3. a positive response to one suggestion increases the probability of responding to the next one.

Evidence was presented in support of these hypotheses and further experimentation proposed. Numerous suggestions, deducible from the theory, for improving the success of HI were interspersed throughout.

Finally, some of the major individual factors that can influence the HI were discussed included concentration, prestige, expectation, fears, attitudes, imagination, and age.


An attempt will be made to show that all hypnotic phenomena can be explained in terms of the following two hypotheses: Hypotheses IV – a suggestion leads to the desired response by first evoking a cognitive stimulus which is connected with that response; and Hypothesis V – the inhibitory set evoked by a suggestion facilitates the suggested response by inhibiting stimuli competing with the cognitive stimulus. This explanation holds for all suggestion phenomena, whether the suggestion is given in the hypnotic state or in the normal state. The reason that hypnotic suggestions lead to stronger responses than normal suggestions is that the inhibitory set is made more effective after a hypnotic induction.

In what follows we will first analyze a number of hypnotic phenomena in terms of this explanation and then look at the evidence in support of it. (Most of the examples of hypnotic phenomena that will be given are referred to in Weitzenhoffer's [1953] extensive review of the field.) This, of course, will not be an exhaustive list of hypnotic phenomena, but should be enough so that the reader will be able to apply the explanation to other hypnotic phenomena not covered.

Hypnotic Phenomena


A hallucination is herein defined as a highly vivid image which is incongruous with the present sensory and/or cognitive environment ("reality"). An image is defined as a cognitive stimulus emanating from an engram or recording of a previously experienced sensation or combination of sensations. Since engrams can be of any sense modality, images can also be in any sense modality. This is important to remember, as one often associates the term image with only visual images.

A hallucination occurs as a result of a suggestion because

  1. an image (a cognitive stimulus) is evoked, and
  2. an inhibitory set is also evoked, one which inhibits competing stimuli in the Stimulus Dominance Hierarachy (one which inhibits "reality") sufficiently so that the image rises to the dominant position in the hierarchy.

As an example of how the explanation works, let's say the suggestion were given to a subject that there was a watermelon on an obviously empty table. First of all, an image of the watermelon on the table would be evoked by this suggestion, since words can act as conditioned stimuli triggering engrams. However, this image ordinarily would tend to be quickly suppressed by the more dominant incompatible stimuli present. The very sight of the empty table, i.e., the incompatible sensory stimulus "empty table", would most likely be enough to suppress the image. In addition, there might also be cognitive stimuli in contradiction, such as the knowledge that watermelons weren't in season, or the remembrance that no watermelons had been carried into the room, etc. If, however, these incompatible, competing stimuli could be suppressed or eliminated, then, according to the theory, a highly vivid image of the watermelon on the table would occur (an image so vivid that if one were to eat a piece of this imaginary melon, we would find all the responses associated with the eating of a real melon – salivation, gastric secretions, enzyme production, etc.). Hypnosis facilitates the production of hallucinations because its strong inhibitory set helps in the suppression of those contradictory stimuli.

Age Regression

Age regression induced through hypnotic suggestion is a phenomenon very similar to hallucinations induced through hypnotic suggestion. There are two main differences, however. First, whereas age regression involves the evoking of a specific image or related set of images which are recordings of an actual event that has taken place in S's past, a hallucination can involve combinations of such images forming a new and not previously experienced precept (such as the image of an animal with a lion's body and a giraffe's head). Second, a hallucination is often projected onto the present sensory environment (e.g., "You will see a red number 7 on the wall." The "7" is imaginary, (but the wall is real), whereas in age regression, the present environment tends to be replaced entirely by the set of images evoked.

Control of Physiological Responses

Hypnosis can be looked upon as a state where one has greater control over involuntary responses. Most physiological responses are considered in this class. Among the physiological responses reported to have been controlled by hypnosis we find the following: basal metabolism [Platinov, 1959, p.110]; blood sugar level [Platinov, 1959, p.113; Weitzenhoffer, 1953, p.135]; enzyme secretion, gastric acidity, and secretion of bile [Weitzenhoffer, 1953, p.135]; water metabolism and temperature regulation [Platinov, 1959, pp.161, 169]; blister formation [Hadfield, 1917; Schneck, 1953, p.263; Ulman, 1947]; hunger contractions [Lewis and Sarbin, 1943; Scantlebury et al., 1942]; and heart rate [Schneck, 1953, p.262].

An example of how such mechanisms can be controlled by means of hypnosis has already been given above. We saw how the suggestion of eating a hallucinatory watermelon would be expected to lead to certain gastronomical responses.

Another example would be the control of secretion of pepsin. Here we could suggest the eating of a steak, since pepsin is the enzyme secreted when proteins are eaten.

In these examples we see that in order to evoke the particular involuntary response(s) in question we must first evoke a cognitive stimulus to which the particular response is attached. The inhibitory set produced by the HI suppresses competing stimuli, thereby increasing the probability that this cognitive stimulus will rise to a dominant position in the hierarchy and thus increasing the probability that the response will be made.


Emotions, another type of involuntary response, can be evoked by means of a suggestion that evokes a cognitive stimulus which leads to the desired emotional response. The inhibitory set raises this cognitive stimulus in the stimulus dominance hierarchy leading to a strong emotional response. As an example, the emotion of fear could be produced by suggesting that S was going to be burned with a hot poker. The emotion of anger could be produced by suggesting something that would be know to evoke anger in the subject. In a similar manner, emotions can also be quelled by hypnosis. Fear, for example, can be eliminated by focusing on a cognitive stimulus incompatible with the fear. A dentist using hypnosis to eliminate fear in his patient might suggest a pleasant image of listening to relaxing music. The combined inhibitory effect of this cognitive stimulus plus the overall inhibitory set of the hypnosis will act to inhibit the fear response.

Motor Phenomena

Among the hypnotic phenomena that can be classified as motor phenomena, we would list such things as deep relaxation, involuntary movement of a limb, and increased work capacity (paralysis could also be listed here, but it will be discussed in the section on hypophenomena).

A relaxation response would be produced by evoking a cognitive stimulus connected with the relaxation response. This could be done by simply suggesting that the subject were going to feel more relaxed, since the word "relax" has been associated often enough previously with the response of being relaxed. Or, the S could be told that he were lying relaxed in a nice warm tub of water or lying on a beach on a nice warm sunny day.

Involuntary movement of the limb could be produced by evoking the image of the limb moving or being moved. For example, if we wanted the arm to rise, we would suggest that there was a rope lifting the arm.

In all these cases of motor control, the inhibitory set would act to increase the strength of the cognitive stimulus by inhibiting interfering stimuli. For example, in the case of increased work capacity, one type of interfering stimulus inhibited would be pain.


The phenomenon of hyperacuity of the senses produced through hypnotic suggestion can be explained as follows:

  1. The suggestion of increased acuity evokes a cognitive stimulus which mediates attentive responses; e.g., in the case of hypervisual acuity, the head turning in the appropriate direction, the eyes focusing, the increased activity of the reticular activating system, etc.;
  2. The suggestion evokes an inhibitory set which inhibits any stimuli that would tend to compete with this cognitive stimulus; to continue the example, any stimulus that would distract the subject or compete with the visual stimulus. Such competing stimuli would include the "negative" thought likely to occur to the subject that he will only be able to see as well as he has in the past.

Among the hypophenomena we would place such things as suggested blindness, anesthesia, amnesia and paralysis. Suggestion leads to hypophenomena:

  1. by evoking a cognitive stimulus which is incompatible with the stimulus or response we wish to suppress, and
  2. by evoking an inhibitory set which tends to inhibit stimuli incompatible with this cognitive stimulus.

Following from this explanation the suggestion of blindness, for example, must first evoke a particular "blindness" cognitive stimulus (which, of course, is incompatible with responding to visual stimuli).  Put in other words, the subject must have some idea of what the word blindness means.  Although few people will have experienced actual blindness in their past, they can still give some meaning to the term.  The subject could, for instance, interpret "blindness" to mean "unable to see as if one were in a dark room." He has experienced dark rooms and thus a cognitive stimulus or image of a dark room can be evoked.  Since reaction to this image is incompatible with seeing, visual stimuli will be inhibited.  (From this we would predict that one’s chances of producing blindness through suggestion would be better if one actually suggested that S would find himself in total darkness, rather than directly suggesting blindness.  The advantages of the use of this "indirect" suggestion over direct suggestion will be discussed more fully in a subsequent section.)

Actually, whenever any visual hallucination is suggested while S’s eyes are open he becomes blind to the visual stimuli incompatible with the hallucination.  For example, if S is told that when he opens his eyes he will be in his first grade classroom, he will see this and will not see anything but the image; i.e., he will be blind to everything but the image.  The reason he does not appear blind at the time, of course, is that he is "seeing" his classroom.

As another example of how a hypophenomenon occurs under hypnosis, we can take an interesting finding of Hernández-Peón, et al. [1960]. He discovered that he could get complete anesthesia to a cutaneous stimulus by suggesting that S was experiencing a different stimulus, a burning sensation.  Here the hallucination of a burning sensation inhibited the cutaneous stimulus.

In explaining hypophenomena one might think that it would not be necessary to evoke a specific cognitive stimulus, that the inhibitory set by itself should suppress the stimulus to be inhibited.  It is true that the inhibitory set may include the particular stimulus in its scope and thus cause inhibition to a certain extent.  However, evoking a cognitive stimulus incompatible with this stimulus increases the amount of inhibition, suppressing the stimulus still further.

The mechanism involved in the production of amnesia through suggestion is similar to that involved in causing blindness and anesthesia.  The main difference is that whereas the suggestion of blindness or anesthesia involves inhibition of a sensory stimulus, the suggestion of amnesia involves inhibition of a cognitive stimulus.

There is another type of amnesia associated with hypnosis.  It is referred to as "spontaneous amnesia."  This often occurs upon awakening from the hypnotic state, without amnesia directly being suggested.  One way that this can occur is that the S might have expected amnesia as a natural consequence to hypnosis.  In this case one can say that amnesia had been suggested, but indirectly.  Another way is as follows:  In order to recall a previous event, cues associated with the event are needed.  Spontaneous amnesia occurs because when a person awakens from a deep hypnotic state the cues around him are no longer the ones he was focusing on in his trance.  During the trance he has been completely oblivious of his surroundings as he focused on the imaginary world evoked by the hypnotist; thus, the cues in the surroundings were not connected with this "imaginary world".  This same type of explanation can be used to explain why it is so difficult to recall dreams upon awakening from sleep.

Paralysis produced by suggestion occurs when the cognitive stimulus evoked is one incompatible with the motor response being inhibited.  For instance, the suggestion that S will be unable to move from his chair might evoke the dominant thought that he is too comfortable to move; or the suggestion that S will not be able to lift his arm might evoke the image of a very heavy weight on the arm.

Evidence in Support of the Explanation of Hypnotic Phenomena

In this section we will look at the evidence in support of Hypotheses IV and V, as well as suggest further tests of them.  In doing so, we will not only be presenting evidence in support of the theory, but we will also be suggesting methods whereby suggestions can be made more effective; i.e., whereby we can increase the probability of getting a positive response to a suggestion.

Hypothesis IV:  A suggestion produces the desired response by first evoking a cognitive stimulus which is associated with that response.

This, among other things, means that the suggestion must have meaning for the subject, or no response will result.  For example, if the experimenter suggests to the subject that he will secrete the enzyme pepsin (the protein enzyme), no response is likely to occur since most people would not know what pepsin is.

From the hypothesis, we can deduce a number of corollaries:

Corollary 7: The Higher the cognitive stimulus is in the stimulus dominance hierarchy (SDH) to begin with (i.e., the height before the cognitive stimulus is aided by the inhibitory set), the greater the response to the suggestion.

This corollary would predict that hypnosis would more greatly facilitate the recall of meaningful material than nonsense material; this, because recall of meaningful material would involve evoking stronger cognitive stimuli than nonsense material. In support of this contention are a number of studies. For example, White, Fox and Harris [1940] tested recall for nonsense material, meaningful verbal material, and meaningful nonverbal material (scenes from movies).  They reported no gain in hypnosis for nonsense material and a gain of around 80 percent for nonverbal meaningful material.

This corollary would also imply, for example, that if we wanted to increase the chances of inducing age regression, we would be wise to first suggest some incidents that are likely to have made a deep impression at the particular age (i.e., left a strong engram) such as a birthday party or graduation.

This corollary would also predict that indirect suggestion would be more effective than direct when attempting to control involuntary responses.  To illustrate what is meant here we will look at a number of examples: 

Since the amount of responsiveness to individual suggestion will be a definite factor affecting hypnotic induction, these methods might also be felt to be of value in making a hypnotic induction more effective as well.

In the above pepsin example, it was implied that the word pepsin was never associated with the eating of protein.  But for some people there might be some association between the two.  Would the suggestion of pepsin secretion lead to pepsin secretion in such people?  Probably, to a certain extent; but from this second corollary we would predict that this direct suggestion of pepsin secretion would be much less effective than the indirect suggestion of eating a steak.  This is because the cognitive stimulus of a protein food evoked by the word "steak" would be higher in the stimulus dominance hierarchy than the cognitive stimulus evoked by the word "pepsin."  The word "pepsin" has most likely not been present very often during the eating of a protein meal (a person is much more likely to think of the word "steak" while eating a steak than the word "pepsin"), nor is it likely to have been associated that much with protein-type words which could act as mediators.

Similarly, if we wanted a person to salivate it would be wiser to use the indirect suggestion of tasting salt, sucking a lemon, or eating a delicious meal rather than the direct suggestion to salivate.  How often do we actually think of salivating when we are salivating? Also, if we wanted to increase heart rate, we would suggest something fearful. If we wanted to decrease heart rate, we would suggest something relaxing.

The evidence fairly well supports the contention that indirect suggestion is more effective than direct in controlling involuntary responses.  The conclusion reached by Weitzenhoffer [1953, p.138] summarizing his extensive review of this area was that the involuntary functions

"appear to be most susceptible to indirect influences arising from the direct evocation of emotional states and hallucinations.  Direct evocation of the changes themselves is least effective.  In fact, it is rare that involuntary responses are directly altered by suggestion.  It is of considerable significance for a theory of hypnosis that the available information appears to show that in nearly every reported instance for which alterations of reflex and reflex-like responses were produced by suggestions, the reflex arc was most certainly one that involved higher centers in the cortical and subcortical regions."
Corollary 8: The more (compatible) cognitive stimuli associated with the response evoked by the suggestion, the stronger the response to the suggestion.

Thus, if we wanted to induce a vivid regression, it would be wisest to suggest as many things known to be associated with the particular age as possible, as opposed to merely suggesting that S will regress to a particular age.  For instance, the experimenter could get considerable information about a particular day in the patient’s past from his parents and use this in his age regression suggestions. Also, if we wanted to increase the probability of producing an involuntary response, it would probably help to add considerable garnishing to the suggestion. For example, instead of merely suggesting that S was eating a steak, we might suggest that he was eating a thick, juicy steak, smothered in onions.

Hypothesis V:  The inhibitory set facilitates the suggested response by inhibiting stimuli competing with the cognitive stimulus.

This hypothesis is, of course, founded to a great extent on Corollary 5 of the postulates which states that if a set to inhibit incompatible stimuli is conditioned to a given CS, the presence of the CS will facilitate the occurrence of any response that would be interfered with by such incompatible stimuli.

We shall now look at three corollaries to this hypothesis.

Corollary 9: Suggestibility should be increased if sensory stimulation is curtailed.

This corollary would predict, for example, that if the eyes are shut, the lights are dim, proprioceptive stimulation is kept down (by lying still), noises are eliminated, etc., suggestibility should be increased.  (Anyone familiar with the area of hypnosis will recognize these sensory curtailing procedures as part of the usual procedure followed by most hypnotists.)  Curtailment of sensory stimulation decreases the number of stimuli in the stimulus dominance hierarchy (and this includes cognitive stimuli since sensory stimuli can evoke cognitive stimuli) and thus increases the responsiveness to any cognitive stimuli focused on.

In partial support of this prediction are the sensory deprivation studies already mentioned above, which report an apparent increase in suggestibility under sensory deprivation conditions.

Similar to the sensory deprivation evidence are the clinical reports on patients with damaged sensory organs.  This includes the visual sense [Colman, 1894; Wagener, 1948; Bartlet, 1950; and Weinberger and Grant, 1940]; the auditory sense [Colman, 1894].  A high incidence of hallucinations have been reported in such studies, which would lead one to suspect that suggestibility is also increased.  At least one report [Sternberg, 1964] does indicate this to be so.  In this report the hallucinations were shown to be induced through self-suggestion.

Corollary 10: Drugs that act as stimulus inhibitors should lead to a state of heightened suggestibility.

In support of this prediction are the numerous studies indicating that such drugs as LSD and Sernyl, which have been shown to act as stimulus inhibitors, do indeed produce states of hypersuggestibility [Barrios, 1965: Sjoberg, 1965; Solursh and Rae, 1966].

Similarly, anesthetic type drugs, such as sodium pentothal, which induce a sleep-like state, have been reported to increase suggestibility when light doses are used; i.e., when doses are not heavy enough to induce complete unconsciousness [Weitzenhoffer, 1953, pp.52-54].

Corollary 11: Suggestibility should be greater when the number of potentially conflicting cognitive stimuli are kept to a minimum.

It has already been pointed out how the elimination of negative attitudes towards accepting suggestions would be expected to increase suggestibility.  Also, from this corollary, we would expect that responsiveness to a suggestion would be greater the more unfamiliar the subject is with the area of the suggestion, or as put by Lindzey [1954, p.27], people will accept suggestions more readily "if they are relatively unfamiliar with a topic, unaccustomed or unable to check up on the suggestion offered to them... "


To sum up briefly, response to suggestion (whether it be normal or hypnotic suggestion) occurs because of two properties of a suggestion.  The words of a suggestion can act as conditioned stimuli which

  1. trigger the suggested responses (via the appropriate cognitive stimuli), and
  2. evoke an inhibitory set which increases the strength of the suggested response by suppressing any stimuli (both sensory and cognitive) which would be incompatible with the suggested response. The reason that hypnotic suggestion is more effective than normal suggestion is that the inhibitory set is greater in the  state of hypnosis.

All responses produced in the hypnotic state can be carried over into the normal "waking" state. That is, they can be made to re-occur on cue after the hypnosis is terminated. This includes the control over all the involuntary functions mentioned, including habits, attitudes, fears, etc. This "carry-over" is done by means of what is referred to as post-hypnotic suggestion. The purpose of the present section is to explain how post-hypnotic suggestion produces such results.

The first step in this explanation is to show that the phenomena under the heading of post-hypnotic suggestion can be explained as a form of higher-order conditioning, a form that Mowrer [1954] has called sentence or sign-sign conditioning, and which the present writer refers to as cognitive-cognitive conditioning. The second step is to present evidence that hypnosis can facilitate this type of conditioning.

In what follows we shall first define cognitive-cognitive conditioning and post-hypnotic suggestion.  Next, we shall condense the explanation of post-hypnotic phenomena into two major hypotheses and present evidence in support of them.


Cognitive-Cognitive (C-C) Conditioning

Cognitive-Cognitive Conditioning is defined as a form of higher order conditioning resulting from the pairing of two cognitive stimuli.

It differs from Pavlovian or first-order conditioning in that the CS and UCS are cognitive rather than sensory.

As an illustration, let's say we wanted to condition salivation to the ringing of a bell by means of cognitive-cognitive conditioning. Rather than pair a real bell with real food, as in Pavlov's classic example of conditioning, we should be able to establish an association between bell and food by pairing the words "bell" and "food". (Because of previous conditioning, the word "bell" has come to evoke the cognitive stimulus "bell" and the word "food" the cognitive stimulus "food".)

We find that Hebb [1949] has proposed a similar model to explain learning in the mature organism. According to Hebb,

"The characteristic adult learning (outside of psychological laboratories) is learning that takes place in a few trials, or in one only. It seems always to involve a recombination of familiar perceptions and familiar patterns of movement... Adult learning is thus a changed relationship between the central effects [cognitive stimuli] of separate stimulations, and does not concern the precipitating stimulus or, primarily, the motor response whose control is imbedded in the central activity... That is, the central effects of sensation are what enter into an association, rather than the comparatively simple sensory event itself. This seems especially true of the most efficient learning – the kind that is established most easily and persists longest."

The type of cognitive-cognitive conditioning resulting from suggestion differs from Pavlovian (or sensory-sensory) conditioning in still another way. A suggestion which pairs the words "bell" and "food" involves more than just saying "bell" and "food", "bell" "food", "bell" "food" over and over. The form of suggestion usually used is more like, "Whenever you hear a bell you will find the taste of food in your mouth." This suggestion does two things, it evokes the cognitive stimuli "bell" and "food", but, in addition, it evokes an inhibitory set (as do all suggestions) which tends to suppress any stimuli which would interfere with the association of these stimuli.

Post-Hypnotic Suggestion (PHS)

Post-hypnotic suggestion can be defined as suggestion given during hypnosis producing C-C conditioning that affects later, post-hypnotic behavior.

Not all suggestions producing C-C conditioning during the hypnotic state will affect later post-hypnotic behavior. Whether the post-hypnotic behavior is affected will depend on the wording of the suggestion and on how the hynotic state is terminated.  For example, the suggestion, "When I ring a bell you will taste food", given during the hypnotic state will probably not affect later, post-hypnotic behavior. This is because when bringing the subject out of the hypnotic state the hypnotist either directly or indirectly suggests that the subject will come back to normal; i.e., that all suggestions given during the hypnotic state will no longer hold. That this suggestion of return to normality can so quickly extinguish the conditioning that has taken place is given some support by the work done on the effect of cognitive factors on conditioning. For instance, Spence [1963] found that when subjects in a conditioning experiment were led to believe that the experiment was over, presentation of the CS was suddenly found to no longer evoke the CR.

Such effects of trance termination on C-C conditioning can be gotten around by means of appropriate wording of the suggestion. For example, we would word the suggestion to read, "Whenever I ring a bell you will taste food, "or better yet, "After you have awakened, whenever I ring a bell you will taste food."

Explanation of PHS in Terms of Two Hypotheses

Hypothesis VI: Post-Hypnotic Suggestion leads to behavior change by a form of higher order conditioning called cognitive-cognitive conditioning.

In strong support of this hypothesis are Mowrer's theoretical formulations on language and behavior, presented in his 1954 presidential address to the American Psychological Association [Mowrer, 1954] and later expanded in his book, Learning Theory and the Symbolic Processes [Mowrer, 1960]. In his discussion of the role of language in conditioning, Mowrer postulates that the sentence (a form of suggestion) can act as a means of conditioning. As he puts it,

"The notion under examination in this chapter is... that the sentence is, pre-eminently, a conditioning device, and that its chief effect is to produce new associations, new learning, just as any other paired presentation of stimuli may do...

"The essence of the argument advanced up to this point is that the subject-predicate complex which we call a sentence is, in effect, simply an arrangement for conditioning the meaning reaction produced by the predicate to the interoceptive stimulation aroused by the meaning reaction elicited by the sentence subject."
[Mowrer, 1960, pp.141-142, 147]

Although in this quote Mowrer refers to "meaning reaction" rather than cognitive stimuli, it will be readily apparent to anyone reading Mowrer that he would consider the two terms practically synonymous [see pp.163-207].

Mowrer goes on to "put this hypothesis about language function into a broader, more systematic perspective" by subsuming sentence conditioning under what he calls "sign-sign" conditioning (what we refer to as cognitive-cognitive conditioning). He points out that signs need not be words only (as in sentence conditioning) but other stimuli and cues with acquired meaning as well.

What Mowrer is essentially saying, then, is that contiguous cognitive stimuli, whether elicited (suggested) by words or by other stimuli, can bond together forming a new cognition, a new conditioned association, leading to new behavior. Mowrer’s arguments in favor of such a contention are quite persuasive, and, as he points out, the experimental evidence in support of it is already beginning to come in [e.g., Staats et al., 1959].

Hypothesis VII: Hypnosis facilitates the C-C conditioning produced by suggestion.

It must, of course, be obvious to anyone that under ordinary circumstances suggestions are not always readily accepted, thus C-C conditioning does not always take place after the appropriate suggestion. Why is this so? We will find that the answer to this question will begin to throw some light on the part hypnosis plays in facilitating C-C conditioning.

Osgood [1963] perhaps best answered this question in his 1963 presidential address to the American Psychological Association when discussing Mowrer’s concept of the sentence as a conditioning device. According to Osgood, if the assertion made by the sentence (the suggestion) is incongruent with the subject’s previously held beliefs and attitudes (the cognitive environment) or his present perceptions (the sensory environment), it will tend to be suppressed.

The interference of incongruent stimuli with C-C conditioning is understandable in terms of the conditioning paradigm if we recall Postulate II (Conditioning). It will be remembered that a corollary to this postulate stated that anything interfering with the contiguous occurrence in the focus of attention of the stimuli being associated would interfere with the conditioning. Since incongruent or incompatible beliefs, attitudes, perceptions, etc., would tend to suppress the cognitive stimuli to be paired, they would thus interfere with the conditioning. Therefore, we would hypothesize that anything that would eliminate such interfering stimuli should facilitate the C-C conditioning. (This hypothesis can be tested by first producing a situation where competing stimuli were eliminated or suppressed and then seeing if this facilitates the C-C conditioning.)

This leads us to the part that hypnosis plays in the facilitation of the conditioning. Hypnosis, it is felt, provides an especially effective means (the inhibitory set) whereby interfering stimuli can be readily inhibited. That the writer is not alone in this approach to explaining the part hypnosis plays in conditioning is seen from the following quote of Leuba's:

"I attributed the quickness and the ease of conditioning during hypnosis to the relatively complete concentration achieved on the conditioned and unconditioned stimuli, and the consequent absence of conflicting and inhibitory responses at the time of conditioning. I envisioned hypnosis as providing ideal circumstances for conditioning to occur. It provided the experimenter with the means for excluding distracting psychological variables – interfering thoughts and experiences."
[1955, p.10]

In discussing the possible mechanisms whereby hypnosis facilitates C-C conditioning we should not only explain why hypnosis facilitates the making of the associations, but also why the CR produced can be such an enduring, "functionally autonomous" response. Hull [1933] seems to be aware of this characteristic in his section on post-hypnotic phenomena where he discusses the results of Patten's [1930] study of the effect of repetition on the strength of the post-hypnotic response:

"The composite graph of these results shows that the vigor of the [post-hypnotic] response while slightly variable, displays no tendency whatever to fall, but, if anything, a slight tendency to rise. Patten believes that with a daily practice post-hypnotic sug­gestion might persist indefinitely without renewal of the suggestion. However this may be, it is evident that the repeated performance of the post-hypnotically suggested act characteristic of clinical practice would seem to be favorable for maintaining its strength."

It is felt that the functionally autonomous nature of the post-hypnotic conditioned response can best be explained if we assume an interference theory explanation of extinction. This theory states that in order for a res­ponse to become extinguished, another incompatible response must become conditioned to the CS. An implication from this interference theory would be that if the CR is stronger than a potentially interfering response, the latter will be the one inhibited. Thus, as long as we have a strong enough CR to begin with, it can keep itself from being extinguished. And what’s more, if we have such a strong conditioned response, it not only will inhibit the competing responses, but will itself become conditioned to the potentially interfering stimuli. (For example, we know that if we attempted to extin­guish a strong conditioned fear response by feeding an animal in the direct presence of the feared object, we could very well find that the animal soon becomes afraid of eating.) Not only would the CR become associated with the competing stimuli, but, of course, neutral stimuli as well. All this would serve to strengthen the CR in that it would now be associated with many more stimuli than just the original CS.

It should be mentioned that in therapy there is probably still another reason for a post-hypnotic response becoming functionally autonomous--it can become self-reinforcing from the relief or new pleasure experienced whenever the new response occurs.

We have shown how a strong CR can become functionally autono­mous, but now the question is, "Why is the CR established through PHS so strong in the first place?" In order to explain this, we propose, first of all, that in the process of conditioning, in general, there are two components of the UCS which become associated with the CS, an excitatory and an inhibitory one. This inhibitory component, or set, as we refer to it, is the same one suppressing the competing stimuli at the time of the association.

This is close to the position held by Harlow [1959], who considers learning to involve the transfer of the learned inhibition of the error-producing factors(the competing stimuli) operating during the learning, to the particular situation (the CS).

In the case of the conditioning taking place when a PHS is given, the inhibitory set conditioned to the CS is the same one developed by the hypnotic induction. It is because this set is so strong that the CR is so strong.

Understandably, a learning theorist might hesitate in accepting the possibility that it is a process of conditioning that underlies the dramatic changes produced in hypnotherapy. One-trial conditioning and functional autonomy are not commonly encountered in the laboratory. However, such phenomena are more prevalent outside the laboratory. Because theses phenomena are difficult to fit into the learning theorists' present scheme of things – because they don't seem to fit the usual gradual acquisition curves and the declining extinction curves – does not mean that they should be rejected as conditioning phenomena. Rather, the learning theorists should reject their outmoded theories, or at least revise them so as to better encompass these phenomena as well. The inhibitory set approach which has been stressed in this paper is felt to be one direction learning theorists could take in arriving at a more comprehensive theory of learning.

Now let us see if we can find evidence to show that hypnosis does indeed facilitate cognitive-cognitive conditioning.

Evidence that Hypnosis Facilitates C-C Conditioning

There are at least four sources of evidence that we might use to support the hypothesis that hypnosis facilitates C-C conditioning. One comes from the use of PHS to facilitate psychotherapy. Another source comes from the experiments which have shown hypnosis to facilitate first-order conditioning. A third source comes from its use in medicine. Finally, there is the experimental work that has been done on post-hypnotic suggestion.

Facilitation of Therapy Via PHS

As pointed out by Barrios [1969, 1970], post-hypnotic suggestion has been shown to be a highly effective means for producing therapeutic behavior changes. Three large-scale studies were cited in support of this contention [Richardson, 1963; Chong Tong Mun, 1964, 1966; and Hussain, 1964].

Richardson reported an improvement rate of 94.7 percent of 76 cases of frigidity. The average number of sessions was 1.53. The percentage of orgasms (the criterion for judging improvement) rose from a pre-treatment average of 24 percent to a post-treatment average of 84 percent.

Chong Tong Mun's study covered 108 cases. These included patients suffering from asthma, insomnia, alcoholism, dysmenorrhea, dermatitis, anxiety state and impotence. The percentage of patients reported improved (removal or improvement of symptoms) was 90 percent. The average number of sessions was 5. The average follow-up period was 9 months.

Hussain's study reports on 105 patients of varying diagnostic categories. This included patients suffering from alcoholism, sexual promiscuity, impotence and frigidity, sociopathic personality disturbance, hysterical reactions, behavior disorders of school children, speech disorders, and a number of different psychosomatic illnesses. The percentage of patients reported improved was 95.2 percent. The number of sessions ranged from 4 to 16. The criterion for judging improvement was complete or almost complete removal of symptoms. The follow-up ranged from 6 months to 2 years.

Facilitation of First-Order Conditioning

Two studies seem to indicate that hypnosis facilitates first-order conditioning. Scott [1930] found that he could establish a conditioned finger withdrawal response much more rapidly and effectively in his hypnotic subjects. Whereas only five of nine control subjects were conditioned, in an average of 26.6 trials, all of the hypnotized subjects were conditioned and in an average of only 14.2 trials. The remaining four controls had not been conditioned after an average of 30.3 trials.

Leuba [1940] found that he could establish conditioned sensations in his hypnotic S's in an average of six trials, often in only one trial. During deep hypnosis, two stimuli, such as the ringing of a bell and a pin prick on the hand, were applied simultaneously for about six pairings. Before awakening, S's were given post-hypnotic amnesia for what had occurred. A few minutes after awakening, one of the two stimuli was presented whereupon S automatically reacted as if the other stimulus had also been presented. The conditioned sensations were frequently so intense and vivid as to be mistaken for actual sensations. Unfortunately, Leuba does not report using a control group of non-hypnotic S's.

Use of PHS in Medicine

Post-hypnotic suggestion has been used very successfully with hospitalized patients who were ill due to traumatic injury and/or chronic disease [Cangelo, 1961; Crasilneck et al., 1955; Fogelman and Crasilneck, 1956; Kroger and DeLee, 1943; Marmer, 1956; Mason, 1955; Raginsky, 1951; and Schneck, 1953]. In these studies we find post-hypnotic suggestion serving a number of different uses. For example, it is of great use in the reduction of pain and the need for narcotics. This includes post-operative pain, the pain resulting from severe burns, and the pain of terminal cancer. It has also been used, for example, to induce a greater appetite in patients whose previous refusal to eat was endangering their lives [Crasilneck, et al., 1955].

A criticism that might be leveled at the use of the use of the above clinical reports as evidence in support of the contention that hypnosis facilitates C-C conditioning is that in most cases no appropriate comparison control group was run. That is, matched patients were not treated with waking suggestion rather than hypnotic suggestion. Some people might feel that such results could have been achieved on the basis of waking suggestion alone.

Experimental Work Done on Post-Hypnotic Suggestion
"Despite a wealth of anecdotal material and case reports, there have been few experimental investigations of the performance of post-hypnotic behavior."
[Fisher, 1954]

Although fisher made the above statement in 1954, for the most part, it continues to hold true.

The following is a summary of most of the studies indicating the effectiveness of post-hypnotic suggestion:

    Lundholm [1928] was able to produce deafness and blindness by means of PHS. Hammer [1954] found that post-hypnotic suggestions of increased ease, concentration, motivation and ability led to significant increases in various learning tasks. Gladfelter and Crasilneck [1960] found that they could increase S's vocabulary skill by means of post-hypnotic suggestions aimed at inducing certain emotions, fear having the greatest effect. Rosenberg [1960] used PHS to effectively change subjects' attitudes.

    A number of studies have been done on duration of the PHS [Kellogg, 1929; Patten, 1930; Weitzenhoffer, 1950; Edwards, 1954 and Orne, 1963]. In general, these studies indicated that, although there was an overall gradual decay of the response, in many cases it continued to be effective for long periods of time, even years; and in some cases there was no decay.

    There are also a number of studies done on investigating other characteristics of the post-hypnotic suggestion. Erickson and Erickson [1941] investigated the "spontaneous self-limited post-hypnotic" trance produced in performance of the PHS. Marcuse [1945] studied the effect on PHS of conscious awareness of the post-hypnotic signals and responses. Weitzenhoffer [1950] discussed the effect of difficulty of task on PHS. Levitsky [1960] summarized various techniques for giving the post-hypnotic suggestion.

    A study by Barrios [1969, 1973] was more specifically aimed at testing the hypothesis that hypnosis facilitates cognitive-cognitive conditioning. The experimental design was such as to eliminate certain methodological shortcomings associated with most of the previous hypnosis experiments. Among other things, this included using an appropriate control group as well as using the subjects as their own controls; a tape recorder was used to eliminate any possibility of experimenter biasing due to changes in tone of voice; a more appropriate measure of hypnotic depth was used; and an involuntary response (salvation) was used to measure the conditioning rather than the usual voluntary type of response used in most previous PHS experiments. The results from the experiment supported the three predictions made from the hypothesis. That is, it was found that

    1. the hypnosis group showed significantly greater conditioning than the control group;
    2. the strength of the conditioned response for the hypnosis group was positively correlated with hypnotic depth; and
    3. the conditioned response once formed was a strong one, as evidenced by no significant extinction.

To briefly sum up this final section, the phenomenon of post-hypnotic suggestion, whereby responses produced in the hypnotic state can be carried over into the normal state, was explained as occurring through a process of higher-order conditioning. It was also pointed out that it is the inhibitory set produced by the hypnotic induction that facilitates this conditioning. This overall explanation was condensed into two hypotheses and evidence was presented in support of them.

(back to top)