Barber (1964e) concluded that the research failed to demonstrate that hypnosis produces auditory or visual hallucinations that are the same as per ceptions or different from imagination. Erickson (1938a; 1938b), on the other hand, took the position that often hallucinations are quite real and reported that suggestions of negative auditory hallucinations, or deafness, could not be distinguished from organic deafness by ordinary means. His subjects dis played no startle response to an unexpected loud sound, failed to raise their voices in speaking when background noise was increased, or failed to blush to auditory stimuli that would normally produce such a response in a particular subject. He also found that a conditioned finger withdrawal response to an auditory-conditioned stimulus disappeared during hypnotically suggested deafness and reappeared after the hypnosis. Black and Wigan (Barber, 1964c) found a similar result with an autonomic nervous system response not under conscious control as a finger flexion is. Pattie (1935) reported the failure to produce uniocular blindness in a small group of subjects as disclosed by ster eoscopes, filter, and Flees box tests. To reconcile these conflicting views, it will be necessary to sample a number of lines of research.

Barber and Calverley (1964;) report that suggestions of deafness were ef fective in 15 hypnotized and 15 nonhypnotized subjects. However, if these subjects were subjected to delayed auditory feedback where the sound of their own voices was delayed slightly, they reacted as do typical subjects with normal hearing by stuttering, mispronouncing words, increasing vocal inten sity, and talking more slowly.

Barber (1964c) reports that in hypnotically suggested deafness in one ear, subjects who display positive results still report hearing a beat note if stimu lated with slightly different frequencies in each ear. Weitzenhoffer criticized this study on the grounds that the frequency applied to the "deaf" ear could have reached the other by bone conduction, but it is interesting to note that the one subject who did not experience the beat note was a physics major presumably familiar with the phenomenon of beat notes. In a study providing results analogous to the common finding that hypnotic pain control has little effect on physiological measures correlated with pain, Sabourin, Brisson, and Deschambault (1980) found that hypnotically in duced deafness did not influence a conditioned heart rate response or the response time in a key-pressing task to an auditory stimulus in subjects re porting a positive subjective effect.

Spanos, Jones, and Malfara (1982) found that high-susceptibility subjects reported greater deafness than low-susceptibility subjects in response to suggestions of unilateral deafness but did not differ objectively in impairment from the latter as measured by responses to words presented in dichotic pairs. Crawford, MacDonald, and Hilgard (1979) found that reduction in hearing in response to hypnotic suggestion correlated 0.59 with hypnotic suscepti bility but the "hidden observer" technique (see p. 116) disclosed that covert hearing was at least 20% greater than reported overtly by the subjects. Subjects who are instructed to hallucinate a background (which normally produces an optical illusion effect) over a figure do experience such an illusion but not as strongly as they would with a real picture of the background added and no more than nonhypnotized subjects instructed to imagine the back ground (Barber, 1964e). Miller, Hennessy, and Leibowitz (1973) found that if such an illusion-producing background was negatively hallucinated away, the Ponzo illusion did not disappear. Hypnotic subjects capable of negatively hallucinating portions of visual stimuli showed varying degrees of ability to attenuate the Tatchner-Ebbin-ghaus circles illusion posthypnotically (Blum, Nash, Jansen, and Barbour, 1981). Miller and Leibowitz (1976) found that a hypnotically produced re-striction of the visual field produced behavior no different from that obtained from a group of simulators. Similar results were reported by Leibowitz, Lundy, and Guez (1980). Leibowitz, Post, Rodemer, Wadlington, and Lundy (1980) found that the amount of visual field narrowing occurring in response to in-structions to simulate such narrowing was a function of the method of mea surement, with direct measurement by perimetry yielding the most effect. Dorcus (1937) found no pupillary reflex in response to suggestions of light intensity change. He also found that the postrotational eye movement (nystagmus) produced in four subjects after hypnotic suggestions that the subject was rotating in a chair were voluntary and not the same as the eye movement produced by the same subjects when actually rotated. Also, falling responses following rotation suggestions did not appear unless the subjects had prior experience actually being rotated, and when produced under these circum stances, they were in the wrong direction for the rotation direction suggested.

Wallace (1980) reports that perceived autokinetic movement of a hyp notically hallucinated light was a function of hypnotic susceptibility as mea sured by the HGSHS. Since the subjects were all psychology students, it is not possible to confirm the present author's opinion that performance was also a function of the subject's knowledge of psychology. The suggestion of a hallucinated light in a dark room is an indirect suggestion to produce au- J tokinetic motion to a knowledgeable subject. Erickson (1939b), using very deeply hypnotized subjects, produced some degree of color blindness as measured by the Ishihara plates. Barber and Deeley (1961) report producing color-blind responses in nonhypnotized sub jects by instructing them to "concentrate away from red and green." Cunningham and Blum (1982) and Harvey and Sipprelle (1978) found significant differences between the subjective experience reported by subjects success fully experiencing hypnotically suggested color blindness and the behavior of people with congenital defects in color vision.

Some subjects who are instructed to hallucinate colors either under hyp-nosis or task-motivational instructions report the occurrence of negative af-terimages. Barber (1964c, 1959b) suggests that such reports do not occur in subjects who are naive concerning the phenomena of negative color after-images, but if they do occur, the afterimage colors reported are those com monly described in elementary psychology texts—that is, the complementary color of the one hallucinated (e.g., red-green, blue-yellow) instead of the somewhat different (more pastel) colors usually reported in actual negative afterimages. Similarly, if an actual color was shown and the subject was told it was different, the actual color, not the hallucinated one, determined the nature of the afterimage (Barber, 1964d).

In view of the foregoing studies, the question arises about which view point, Barber's or Erickson's, is correct concerning the reality or validity of positive and negative hypnotic hallucinations. In the view of the author, both are correct. Erickson is right that these are real experiences; Barber is right that hallucinations are different from ordinary sensations. Hypnotic blindness or deafness is not the same as organic blindness or deafness any more than hysterical blindness or deafness is. Of course, negative afterimages do not occur in subjects not familiar with this phenomenon. How could they? A negative afterimage produced by a real external stimulus is a retinal phenomenon produced by the differential fatigue of different visual receptors. A hallucinated color does not result from retinal activity but from suggestions reaching the cerebral cortex. Sensations or physiological responses in sense organs are not modified in hypnotic hallucinations; perceptions or higher-level mental processes are.

An afterimage produced to a hallucinated color is as much a suggested effect as the color Itself. It is an excellent example of an indirect suggestion. This does not mean that it is not experienced. The real question asked when we inquire about the reality of a hypnotically Induced hallucination is, How vivid is it, or How similarly does the subject experience it to a real external stimulus? This is an unanswerable question. Trying to render the question answerable by equating "real" with similarity to a sensory experience in a physiological sense only introduces confusion. In spite of their rather divergent views, the work of Barber (1958d) and Erickson (1944) seems to support the general conclusion that subjects given hypnotic suggestions of deafness or blindness for a particular person or object behave as though they are trying to avoid perceiving that person or object. Subjects try to avoid focusing or looking at the subject of the negative hal lucination or report perceiving it vaguely. A similar result is reported by Hil-gard and colleagues for negative hallucinations of pain in that a subjectable to ignore the suffering aspects of pain will still report experiencing the sen sations in some manner if he is instructed that there is a hidden observer who can report these sensations (Hilgard, Morgan, MacDonald, 1975).

Barber claims that to get a subject not to experience the object of a negative hallu cination, it is necessary to convince him of the objective truth of the exper imenter's statement that the object is no longer present. Thus, if a subject is told that a chair is no longer present, he will try to look away from it but will not bump into it if it is directly in his path of travel. If, on the other hand, noises are made simulating the removal of the chair while the subject's eyes are shut, he will act as though he really does not see it at some level and will walk directly into it. Erickson reported a similar effect when a subject acted at though he really did not see one negatively hallucinated person but did show some signs of perceiving another for whom the suggestions were made more recently. He ascribed this difference in reactions, in accordance with his characteristic view that a very deep trance is required for this effect, to the fact that it takes time for the suggestions to become fully effective.

Although it seems clear that a positive hallucination of a complex sense modality like vision originates in the cortex, not in a sense organ, some of the easier-to-elicit tactile hallucinations may, partially at least, involve paying attention to a certain amount of dermal stimulation normally present and customarily ignored.

A hallucination is defined as a perception in the absence of a real external stimulus. Usually the occurrence of a hallucination is a symptom of a psychotic disorder, but under certain circumstances, normal people may hallucinate. These situations include conditions of sensory deprivation, extreme hunger or thirst, fever, drugs, REM sleep (nocturnal dreams), and, in some cases, scrying (crystal ball gazing). Normal people may also hallucinate under the influence of suggestions, hypnotic or otherwise.

Psychotic hallucinations in general have both a characteristic sensory mo dality and a characteristic content that vary between diagnostic categories. For example, schizophrenic hallucinations are predominantly auditory and have a characteristic obscene or self-critical content. Most psychotic hallucinations are accompanied by a delusional belief in their objective reality that is often absent in the hallucinations of normal people. This phenomenon is not intrinsically unreasonable, for psychotic hallucinations tend to be consis tent with past experience. For example, a hallucinated image will usually ob scure parts of real images lying in back of it in the visual field, and it will cast a reflection in a mirror. All of his life the patient has been correct in believing the information his sense organs have communicated to him about the external world, and there is no reason why he should not believe in the ver-idicality of these images when they are hallucinated.

In the case of the hypnotically suggested hallucination, the modality and the contect of the hallucination are functions of the suggestions made. Hallucinations can be suggested in any sense modality; the ones most commonly used are vision, audition, olfaction, gustation, touch, heat, and cold. : hallucinations may be suggested in specific modalities or it may produce multimodality effects. For example, the hallucinated fly in the Stanford test (SHSS:A) may produce visual, auditory, fects. In addition to positive hallucinations, negative hal-the subject fails to perceive some real external stimulus, ted. These are analogous, if not identical, to the everyday >erson is looking directly at an object that he is searching :e it. Those portions of the external environment that are lotic suggestions are generally perceived accurately if not (Orne, 1962d).

Hallucinations in general are difficult to elicit under hypnosis, and there is ilty between the various sensory modalities. Tactile hallucinations are comparatively easy to produce. Suggesting to a group of nons that they should notice that their noses are beginning to in effect in many of them. So will reading a paragraph iing that might be used in making such a suggestion. The rises of olfaction and gustation are also more amenable to ! more highly developed senses of audition and vision.

Orne (1962d) points out that when a visual hallucination is suggested In a subject he may react in a variety of ways. He may act as though he sees 1 what has been suggested or seem disturbed because he does not experience it. In the former case, if he is questioned after the experience, he may say that (1) he saw nothing but felt compelled to act as if he did; (2) he expeFrienced a visual image but knew it was unreal; (3) he experienced a real external image but it had illogical aspects to it (e.g., he could see a chair j through a hallucinated person); or (4) he experienced an image indistinguish able from reality. Thus Orne categorized the subject's subjective experience into one of four categories. He considers only the last two as true halluci nations. In actuality, there is probably an infinite series of gradations of subject responses, and individual investigators differ in what they define as a positive | response to a suggested hallucination.

Another point needs to be made: there is no way for an experimenter to observe the subject's hallucination directly. Hence he or she must rely on the subject's verbal report of his or her experiences. Thus it is possible, and in deed probable, that one subject who experiences a hallucination more vividly than another may report it as less vivid because of individual differences in the use of language and subjective standards of what the term vivid means. This is the same problem experienced by dream researchers who purport to be studying dreams but are actually studying verbal reports of dreams. The only hallucinations that an investigator can observe directly are his or her own, and these are necessarily individual and atypical. The question about the relative subjective experiences of two different subjects reporting their own hallucinations, no matter how similar or different their verbal descriptions, is as unanswerable by observation as is the question of whether two subjects describing the same stimulus as blue are having the same or radically different subjective experiences.

Such questions are philosophical, not scientific, ones. Although the degree of the apparent reality of a hallucination can only be estimated by a verbal report, Orne (1962d) attempts to distinguish effects that are actually experienced from those that are simulated by subjects motivated to produce what the experimenter wants them to by the use of stimulating subjects. These are subjects who have not been hypnotized but have been instructed to act as if they have been and to attempt to deceive the experi menter making the behavioral observations (who is not told which subjects are actually hypnotized). Simulators are usually informed that if the experi menter discovers that they are simulating, he or she will halt the procedure; hence, its continuation lets the simulator know he is successful in efforts at deception. The logic behind the use of simulating subjects is that both hyp notized subjects and simulators are equally motivated to produce the sug gested behavior, but if only the hypnotic subjects actually experience the ef fects suggested, their behavior may be different to some degree from that of the subjects who are faking an effect. The lack of knowledge on the part of the experimenter of the real or simulating status of a subject eliminates experimenter bias and prevents any unconscious systematic differential treatment of the two types of subjects.

There are behavioral differences between real subjects and simulators. If a subject is told to hallucinate the experimenter sitting in a chair and is then told to turn around and look at where the experimenter really is, he will often appear surprised and report seeing him twice. He may not know which image is real. (Some subjects will distinguish the real from the hallucinated image by having the hallucinated one raise his hand.) Simulating subjects will usually deny seeing the experimenter when looking at him because they believe they are not supposed to.

If a negative hallucination is suggested so that the subject is told he can no longer see a chair and then is asked to walk in a direct line with the chair, hypnotized subjects will avoid bumping into the chair, while simulators will Usually walk into it. Spanos, Churchill, and McPeake (1976) found that a cooperative attitude toward hypnosis and involvement in everyday fantasy were each positively correlated with the ability of a subject to experience visual and auditory hal lucinations. Visual hallucinations were more difficult to produce than auditory hallucinations, but they found that the abilities to produce these two types of hallucinations were correlated. They reported no sex difference in the ability to hallucinate. A large majority of their subjects reported their experiences as Imagined rather than seen or heard. Ham and Spanos (1974) report that with 60 male and female subjects equally assigned to hypnotic and task-motivational groups, the task-moti vated subjects performed better in response to suggestions of visual or au-ditory hallucination. Spanos, Mullens, and Rivers (1979) in a 2 x 3 factorial study compared hypnotic and task-motivated subjects in performance of vis ual and auditory hallucinations in response to brief suggestions, long sug gestions, and suggestions providing an imaginary context. Task-motivated subjects performed better than hypnotic subjects on auditory hallucinations, I and the authors report a "trend toward significance" in this direction on visual hallucinations. Both long and image-involving suggestions were equally more effective than short suggestions for auditory hallucinations but were not sig nificantly different for visual hallucinations.