CHAPTER 2 p 1 The domain of , 1 revisited I John F. Kihlstrom

Hypnosis is a process in which one person, desig- question has nothing to do with sleep. Although nated the hypnotist, offers to another the deep roots of hypnosis may reach back to person, designated the subject, for imaginative the ancient temples of Aesculapius (Ellenberger, experiences entailing alterations in perception, 1970; Gauld, 1992; but see Stam and Spanos, memory and action. In the classic case, these 1982), the immediate historical predecessor of experiences are associated with a degree of subjec- hypnosis is the of Franz tive conviction bordering on delusion, and an Anton Mesmer (for a definitive biography, see experienced involuntariness bordering on com- Pattie, 1994). Or was it? Peter has recently pulsion. As such, the phenomena of hypnosis argued that the true ancestor of hypnosis is the reflect alterations in consciousness that take place exorcism practiced by Johann Joseph Gassner in the context of a social interaction. (1729-1779), a Catholic priest who performed exorcisms throughout Europe right before 2.1. Introduction Mesmer came onto the scene (Peter, 2005). Although the similarity between some of I have employed one or another close variant of Gassner's practices and Mesmer's practices is the above definition of hypnosis at least since undeniable, the prize probably should remain 1982 (Kihlstrom, 1982, 1985), and it has also with Mesmer. After all, Gassner offered a super- served as the starting-point for the 'consensus' natural theory of illness, while Mesmer at least definition and description of hypnosis crafted embraced the principle that disease had natural by the American Psychological Association's causes and cures. But Mesmer's theory-that Division 30, the Society for Psychological his effects were mediated by a physical fprce Hypnosis, in 1993 (Kirsch, 1994a,b; for later analogous to magnetism-was wrong too, and definitional developments, see Killeen and from a scientific perspective we can now under- Nash, 2003; Green et al., 2005a). In this chapter, stand both his cures and Gassner's cures as the I reflect on each element of the definition, its product of 'imagination'-but no less genuine historical evolution, and its current status. for that. It is not Mesmer's fault that, in the late 2.2. Hypnosis nineteenth century, psychology was not permit- ted the status of a true science. The term hypnosis itself is, of course, something Mesmer's theory was discredited in 1784, by of a misnomer, because the phenomenon in a French royal commission chaired by Benjamin 22 - CHAPTER 2 The domain of hypnosis, revisited

Franklin (Bailly, 178412002; Franklin et al., hypnosis, involving monoideism-the concen- 178412002; for recent commentaries, see tration of attention on a single object. With this Forrest, 2002; Kihlstrom, 2002; Laurence, 2002; one-two punch, renaming the phenomenon Lynn and Lilienfeld, 2002; McConkey and Perry, and offering a new, more scientifically accept- 2002; Perry and McConkey, 2002; Spiegel, able theory to explain it, Braid buried Mesmer 2002), but his practices lived on-first as mes- and animal magnetism once and for all (Kravis, merism, a term introduced to English in 1784 by 1988).Although the term hypnosis had appeared Benjamin Franklin himself (Pepper, 191 l), and before Braid's time (Gravitz and Gerton, 1984; which lasted long enough to be used by Pattie, 1990), Braid gave us the whole vocabu- Elliotson (184311977) and Esdaile (184611977). lary for hypnosis as we understand it today, But while Mesmer's practices were revived in the offering clear definitions of the terms hypnotic, early nineteenth century, their identification hypnotize, hypnotized, hypnotism, dehypnotize, with the man himself must have created a public dehypnotized and hypnotist (Kihlstrom, 1992b). relations problem; a new label was needed. According to the Oxford English Dictionary, the Puysegur (1807), Deleuze (1813) and others word hypnosis entered the medical dictionary in among the second generation of mesmerists 1882-although the same entry flirted with proffered the master's own preferred label, ani- Braidism. mal magnetism, a term which persisted almost These terms have stuck with us, through what into the modern era (Binet and Fere, 1888). But Gauld (p. xi) aptly termed 'the heyday of hypno- this was unsatisfactory-not least because the tism' in the run-up to the twentieth century Franklin Commission had disproved the notion (Charcot, Janet, Liebeault, Bernheim, James, that magnetism had anything to do with the Prince and Sidis), through the first systematic phenomenon. experimental work by Young (1925, 1926) and Of course, a major transition in conceptions Hull (1933), and into the post-war revival of of hypnosis had begun in 1784, even before the clinical and experimental hypnosis at the hands Franklin Commission had completed its work, of Weitzenhoffer (1953), Gill and Brenman when Puysegur magnetized Victor Race, a (1959), Orne (1959), Sutcliffe (1960, 1961), young shepherd on his estate. Instead of under- E. R. Hilgard (1965), Barber (1969), and Sarbin going a magnetic crisis, Victor fell into a sleep- and Coe (1972). Although some authorities were like state in which he was nonetheless responsive once so disturbed by the term that they felt com- to instructions, and from which he awoke with pelled to enclose it in scare quotes (e.g. Barber, amnesia for what he had done. Art$cial som- 1964), that is all over now (Barber, 1999). nambulism-a term introduced by Puysegur Unfortunately, the word hypnosis has also himself-had a double advantage, in eliminat- been appropriated by anesthesiologists, to refer ing references to both Mesmer and magnetism, to the loss of consciousness that is one of the and also eliminating the convulsive seizures that three aspects of 'balanced anesthesia' (the others were the hallmark of the mesmeric crisis. Here, are areflexia, or the paralysis of the skeletal mus- I think, we have the true predecessor of modern culature, and analgesia, or the lack of pain sen- hypnosis: a sleep-like state in which the subject sation; see Kihlstrom and Cork, 2005). We now is responsive to the-to the what? Mesmerist know that hypnosis is not anything like sleep, won't do, nor will magnetizer, for the reasons physiologically (Evans, 1979); nor is general stated. Perhaps somnambulizer?Not likely. anesthesia, for that matter. And although sleep is The problem was solved once and for all by a frequent metaphor used in hypnotic i5duc- Braid, who coined the terms neurypnology and tions, and relaxation is a typical accompaniment neuro-hypnotism-which quickly-and, as to hypnosis, we now know that neither is neces- Gauld (1992, p. 281) remarks, 'mercifully'dis- sary for hypnosis to occur (Banyai and Hilgard, solved into hypnotism. Although Puysegur had 1976; Alarcon et al., 1999).At this point, then, it offered a psychological theory of animal mag- is clear that the term hypnosis has become func- netism, in terms of the influence of the magne- tionally autonomous of its origins (Allport, tizer's will on the subject, Braid (1855) also 1937). Nevertheless, the term has long served to offered the first psychophysiological theory of label a particular set of phenomena that we are Hypnotist - 23 interested in, of human mental function, and we and women be trained as hypnotists and then should just stick with it. turned loose on an even larger group of subjects (Coe, 1976; D'Eon et al., 1979; Banyai, 1991, 2.3. Process 1998). It's all a very daunting task. So far as the laboratory is concerned, we In the first sentence of previous versions of the assume that the hypnotist functions much like a definition given at the outset, I described hyp- coach, or a tutor, whose job is to help subjects to nosis as a 'social interaction'-which indeed it become hypnotized, and to experience hypnotic is. However, hypnosis is not just a social interac- suggestions. The coach-tutor analogy breaks tion: it also involves certain changes in experi- down, however, in that it is arguably helpful if a ence, thought and action. As such, one could tennis coach can play tennis, or if a math tutor just as easily write that 'hypnosis is an altered knows mathematics. But it does not seem to be state of consciousn~ss. . .'. Describing hypnosis important whether the hypnotist is hypnotiza- as a social interaction acknowledges the role ble. Ted Barber was, by the accounts of those that social influence plays in the process, but it who knew him, highly hypnotizable; Jack also may privilege the social over the cognitive. Hilgard was not. It is tempting to speculate on Accordingly, I now prefer the somewhat more the role these individual differences might have neutral term process ('Hypnosis is a process in played in designing their programs of hypnosis which one person .. .'), in the dictionary sense research, with Barber (1969) believing that of a dynamic series of ongoing actions or events. hypnotic phenomena were possible for anyone Certainly hypnosis is a process in this sense, who tried hard enough, and E. R. Hilgard beginning with an induction procedure, contin- (1965) developing a whole battery of scales for uing with whatever transpires while the subject measuring hypnotizability. Surely the most is hypnotized, and ending with the termination experienced hypnotist of all time is the person of hypnosis and, perhaps, the testing of post- whose voice is heard on the standard recording hypnotic suggestions. What goes on in this of the Harvard Group Scale of Hjrpnotic process-both interpersonally and intrapsychi- Susceptibility, Form A-a professional radio cally-is what interests us as hypnosis researchers announcer without any training in psychology and as practitioners. or hypnosis (L. Dumas, 1964; Orne, 1964). In the clinic, things may be different. But 2.4. Hypnotist even there, the empirical literatureas opposed to folklore-on hypnotist effects is sparse The mythical image of the hypnotist is, arguably, (Diamond, 1984). While anyone can be a hyp- the character Svengali in George du Maurier's notist, it takes special training to use hypnosis 1895 novel Trilby, and the 1931 film made from appropriately in medicine, dentistry or psy- it, starring John Barryrnore and Marian Marsh chotherapy. Some of that training is in hypnosis, (Kihlstrom, 1987): the eyes, the passes with the and in an age when hypnotists advertise in the hands, the rapport with the subject, the under- Yellow Pages, consumers naturally wonder how current of sexuality (of course, Svengali is male they can ensure that a practitioner is appropri- and Trilby female), the risk of harm and the inti- ately trained. Fortunately, there are organiza- mations of the paranormal. More importantly, tions in the USA such as the Society of Clinical though, there is the idea that the hypnotist pos- and Experimental Hypnosis which provide sesses a particular personality, which affords workshops in various skill areas, and agencies him a special power to control others-for good such as the American Board of Clinical Hypnosis and for evil. In stark contrast to this image, the to certify competence in the use of hypnosis in literature on hypnotist characteristics, and their medicine, dentistry, psychotherapy and clinical effects on hypnosis, is remarkably thin. Partly, social work. Similar organizations exist around I think, this is due to methodological considera- the world, including in Australia and the UK. tions. Simply to investigate the effect of the hyp- These organizations and agencies have their notist's gender on hypnotizability would require place, but when it comes to evaluating a clinical that a fairly large, representative sample of men hypnotist the heuristic offered by Martin Orne 24 . CHAPTER 2 The domain of hypnosis, revisited

(personal communication, frequently repeated) measurement of hypnotizability on a firm is probably the best: nobody should treat a con- quantitative basis (E. R. Hilgard, 1965). By dition with hypnosis that they are not qualified introducing a standardized procedure for sub- to treat without hypnosis. ject selection, the Stanford scales made it possi- The idea that hypnosis involves two people, a ble for different laboratories to replicate and hypnotist and a subject, would seem immedi- extend each other's studies. Moreover, the ately contradicted by the phenomenon of self- Harvard Group Scale of , hypnosis, in which there does not seem to be Form A (HGSHS:A; Shor and Orne, 1962) any hypnotist at all. At one level, we can say that injected substantial economies of scale into the there is no contradiction, because in a very real assessment procedure. sense all hypnosis is self-hypnosis. The hypno- The availability of reliable and valid instru- tist can recite an induction procedure and make ments for measuring hypnotizability makes suggestions for various experiences, but it is the it possible to examine the personality and cog- subject who must actively participate in the nitive correlates of this cognitive skill. Here, process; without that active participation, noth- it must be said, the yield has been relatively ing happens. Comparisons of self-hypnosis with light. One of the most frustrating aspects of the more traditional 'hetero'-hypnosis show that post-war revival of hypnosis research was they are highly correlated (Shor and Easton, the discovery that, while hypnotizability could 1973; Orne and McConkey, 1981; L. S. Johnson be reliably measured, hypnotizability simply - et al., 1983). It may be more difficult to give one- did not appear to correlate substantially with self a that there is a fly buzzing the sorts of personality characteristics measured annoyingly around one's head, and easier to by the major personality inventories, such as believe that the fly is there when one does not the Minnesota Personality Inventory (MMPI) already know that it is simply a suggestion from and the California Psychological Inventory a pre-printed script. But it is clear that the (CPI; E. R. Hilgard, 1965). Nor does hypnotiz- overlap between the two forms of hypnosis is ability appear to be related to individual considerable. differences in 'cognitive style', such as field At the same time, it must be said that much of dependenceindependence (A. H. Morgan, 1972). what passes for 'self-hypnosis', especially in the The mystery was gradually resolved by the clinic, more closely resembles a relaxed state of discovery that hypnotizability does correlate reverie, and imagery, in which subjects are with scales measuring the tendency to have allowed to construct their own experience with- experiences such as those of hypnosis, but out- out any particular direction from the hypnotist side the hypnotic situation (Shor, 1960; As, (Fromm et al., 1981; Olness, 1981). The rela- 1962,1963; As et al., 1962; Shor et al., 1962; J. R. tionship of this form of hypnosis to the more Hilgard, 1965). The strongest correlate of hyp- traditional 'hetero' form is open to question notizability discovered so far is absorption, or (Kahn et al., 1989)-as is the question of the tendency to become absorbed in various whether this form of 'self-hypnosis' should sorts of sensory, cognitive and imaginal experi- really be called hypnosis at all. ences (Tellegen and Atkinson, 1974; for a review, see Roche and McConkey, 1990). Absorption 2.5. and imaginative involvement simply were not Subject represented on the scales of the MMPI and CPI; Surely the main reason that the hypnotist has so put another way, the individual diffeiences little impact on what goes on in hypnosis is, measured by these inventories fall outside the simply, that virtually all the action is in the sub- domain of hypnosis. However, absorption is ject. The importance of individual differences in related to at least some facets of openness to hypnotic susceptibilityhas long been recognized experience, one of the 'Big Five' dimensions in (Friedlander and Sarbin, 1938), and the devel- the structure of personality (Glisky et al., 1991). opment of the Stanford Hypnotic Susceptibility Even the correlation between hypnotizability Scales, Forms A, B and C (SHSS:A, B and C; and absorption is still too small to permit indi- Weitzenhoffer and Hilgard, 1959, 1962) put the vidual levels of hypnotizability to be predicted Subject . 25

with any confidence from personality measures. Moreover, the determinants and correlates of But at least they tell us that individual differ- this developmental trend remain largely unex- ences in the specific domain of hypnosis are plored (J. R. Hilgard, 1970). For example, little connected with individual differences in the has been done to connect the development of wider domain of personality. Still, even open- hypnotizability with the development of the ness is multifacted: while hypnotizability is theory of mind in young children (Welchross, correlated with absorption, it is not correlated 1999). Surely, the child's capacity to be hypno- with intellectance or liberalism (Glisky and tized must be related to his or her ability to Kihlstrom, 1993). appreciate mental states as such, as well as the Other than this, we know remarkably little distinction between reality and imagination. about hypnotizable individuals. Despite the Along the same lines, response to direct and implications of the Svengali myth, there is no challenge ideomotor suggestions may follow appreciable gender difference in hypnotizability different developmental trajectories, possibly (Weitzenhoffer and Weitzenhoffer, 1958), and related to the development of the general capac- any difference there might be should not be ity to inhibit action; a similar point may apply to taken too seriously, as the literature on gender positive versus negative hallucinations, and to differences presents a host of interpretive diffi- age regression as opposed to amnesia. culties (Maccoby and Jaddin, 1974; Tavris, 1992; Once established, hypnotizability seems to be Hyde, 2005). Stereotypically 'feminine' individ- about as stable as other cognitive skills, such as uals are no more hypnotizable than stereotypi- intelligence (Morgan et al., 1974; Piccione et al., cally 'masculine' ones (Kihlstrom, 1980). 1989). But how is it established? Some research Part of the difficulty in establishing meaning- seems to indicate a genetic contribution to ful correlates of hypnotizability may be method- individual differences in hypnotizability ological. Individual differences in personality (A. H. Morgan et al., 1973), possibly mediated are commonly measured by questionnaires, through the attentional system (Raz et al., while hypnotizability is measured by work sam- 2004~).Evidence for the acquisition of hypnotiz- ples of actual performance. Such heteromethod ability through experience comes mostly from correlations (Cronbach and Meehl, 1955; studies of childhood imaginative involvements Loevinger, 1957; Campbell and Fiske, 1959) are (J. R. Hilgard, 1970). Perhaps the most interest- usually low. Nevertheless, the typical personal- ing developmental finding to date is of an age- ity-hypnosis correlation-including the corre- by-gender interaction, such that women of lation between absorption and hypnotizability childbearing age tend to be more hypnotizable -is low even by the standards of Mischel's than their male counterparts (A. H. Morgan and (1968) 'personality coefficient'. It may be that Hilgard, 1973). On the assumption that most of more interesting results will be obtained when these women were in fact stay-at-home moth- both predictor and criterion are measured by ers, one interpretation of this finding is that the same performance-based methods. As an a capacity for hypnosis is sustained in parents- example of the sort of work that might be done, male or femalewho participate in their chil- Dixon and his colleagues found that hypnotiz- dren's imaginative involvements (J. R. Hilgard, ability was correlated with individual differ- 1970). Given the cultural changes that have ences in automatic processing, as measured by occurred since the early 1970s, a more balanced the Stroop test (Dixon et al., 1990; Dixon and test of this hypothesis is now possible. Laurence, 1992). Research of this sort may help Although modern theories of individual-dif- link the domain of hypnosis to the wider ferences generally assume that personality char- domain of attention, at both the psychological acteristics are distributed along continuous and neuroscientific levels of analysis. dimensions (such as the 'Big Five personality Hypnotizability shows an interesting develop- traits; Wiggins and Trapnell, 1990) as opposed mental trend across the lifespan (Cooper and to discrete categories (such as the classical typol- London, 1971; A. H. Morgan and Hilgard, 1973), ogy of melancholics, cholerics, sanguines and but the results of these cross-sectional studies phlegmatics; Kant, 179811978), the n~tionthat remain to be confirmed by longitudinal research. individual differences may resolve into more or 26 . CHAPTER 2 The domain of hypnosis, revisited less discrete types has currently regained a Eysenck and Furneaw (1945) distinguished foothold in the literature (Gangestad and between primary ideomotor suggestibility,involv- Snyder, 1985; Waller et al., 1996). Perhaps hyp- ing direct verbal suggestions for bodily move- notic virtuosos constitute a discrete type of per- ments, and secondary suggestibility, involving son, qualitiatively different from those who do indirect, nonverbal suggestions for sensory- not possess a particular talent for hypnosis perceptual experiences. Although this distinc- (Balthazard and Woody, 1989; Oakman and tion makes some sense, conceptually, it has not Woody, 1996). For that matter, White (1937) always held up empirically (Evans, 1967). once suggested that there was more than one Moreover, it seems clear that there are forms type of hypnosis. Perhaps the existence of more of suggestibility that go beyond the primary- or less discrete profiles of hypnotic ability secondary distinction. For example, Eysenck (E. R. Hilgard, 1965; Brenneman and Kihlstrom, proposed a category of tertiary suggestibility, 2006;) suggests that there may be several differ- to cover persuasion effects in attitude change ent kinds of hypnotic virtuosos. (Eysenck, 1947). Social psychologists have studied aspects of tertiary suggestibility involv- 2.6. ing conformity, persuasion and other forms Suggestions of social influence (Zanna et al., 1987; Zimbardo Suggestion is central to hypnosis. On the and Leippe, 1991; Forgas and Williams, 2001); HGSHS:A and SHSS:C, hypnosis is induced by these appear to be largely unrelated to hypnotiz- suggestions for relaxation, focused attention ability (Moore, 1964). There is also the placebo aid eye closure; and hypnotizability is measured effect, and related effects of suggesting to by response to suggestions for arm catalepsy, patients that they are receiving medical treat- age regression, auditory hallucination, post- ment (Harrington, 1997; Shapiro and Shapiro, hypnotic amnesia, and the like. The connection 1997; Kihlstrom, 2003). Although placebo between hypnosis and suggestion is so strong responses are mediated by expectation (Kirsch, that the two domains have been concatenated 2004), and expectation plays some role in hyp- throughout the modern history of the field nosis (Council et al., 1986), hypnotic analgesia (Bernheim, 188611889; Hull, 1933; Weitzenhoffer, appears not to be mediated by the placebo 1953; Braffman and Kirsch, 1999; Kirsch and response (Evans, 1967, 1976; McGlashan et al., Braffman, 2001). The concept of neutral hypno- 1969). Furthermore, expectancies turn out to sis, as a distinct psychological state independent play only a modest role in hypnotic responsive- of the subject's response to suggestions, has ness (Shor et al., 1984; Benham et al., 2006). had its advocates (e.g. Ludwig and Levine, 1965; Finally, there is interrogative suggestibility, Kihlstrom and Edmonston, 1971; Edmonston, assessed in terms of the responses of eyewit- 1977, 1981). Still, the fact remains that what nesses, patients and others to leading questions is perennially interesting about hypnosis is (Gudjonsson, 1984; Doris, 1991; Eisen et al., how the hypnotized individual responds to 2002; Shobe and Kihlstrom, 2002). This also suggestions. appears to be independent of hypnotizability It has to be said, however, that not all sugges- (Register and Kihlstrom, 1988). The domain tions, or forms of suggestibility, belong in the of hypnosis will be defined more clearly as domain of hypnosis (for a detailed analysis, the relationships among the various forms of see Tasso and Perez, Chapter 11, this volume). suggestibility, and their respective underlying The suggestion that there is a fly in the room, mechanisms, receive further study (Ghemghiu buzzing annoyingly around the subject's head, et al., 1989; Schumaker, 1991). or that there is a voice coming over a loud- Based on present knowledge, hypnotic sug- speaker asking the subject questions, is not the gestions most closely resemble primary ideomo- same as Joseph's suggestion, recorded in Genesis tor suggestions, suggesting that they belong (xl, 14) that Pharaoh let him out of prison; to overlapping domains (E. R. Hilgard, 1965). or the suggestion, written on a slip of paper Still, the overlap is far from complete: in the first deposited in a box provided for this purpose, place, indirect suggestions such as those that that the library install wireless Internet access. characterize secondary suggestibility are not Suggestions - 27 unknown in hypnosis (Yapko, 1983; McConkey, Hilgard, 1963, 1967). Unfortunately, interpret- 1984; Lynn et al., 1993). Both hypnosis and pri- ing these three factors is complicated by the fact mary ideomotor suggestibility entail responses that their constituent items differ in difficulty to direct suggestions, but close analysis reveals levels. Direct suggestions tend to be relatively that hypnotic suggestions are directed toward easy, while challenge and cognitive suggestions perceptual-cognitive experiences, rather than tend to be relatively difficult. Accordingly, some overt motor behaviors. We do not suggest to theorists (e.g. Coe and Sarbin, 1971) have subjects simply that their outstretched arms are argued that the apparent factor structure of rising or falling; rather, we suggest that they are hypnotic suggestions is actually an artifact of weighted down by a heavy object or being lifted item difficulty, and that, appearances to the wn- up by helium balloons. In hypnosis, the overt trary, a 'single role-relevant skill' (p. 1) runs motor behavior follows from the subjective through the matrix of item correlations. experience-a point to which we shall return. Untangling content from difficulty level is not There are further complexities even within easy, because difficulty levels of the items of the this expanded domain of primary suggestion. In Stanford and Harvard scales are fixed by the the first place, ideomotor suggestions can be standardized procedures established for their further classified into two types: direct sugges- administration and scoring. tions for the facilitation of some motor behav- Some progress can be made, however, once it ior, such as eye closure or arm catalepsy; and is understood that the difficulty levels of various challenge suggestions for the inhibition of motor items, as fixed in the standardized scales of behavior, such as arm rigidity or arm immobi- hypnotic susceptibility, are largely arbitrary. For lization. Moreover, hypnotic suggestions can example, the hand-lowering suggestion of affect perceptual and cognitive experiences as HGSHS:A, generally considered a very easy well as motor behaviors. Interestingly, these per- item, can be made more difficult by requiring ceptual-cognitive suggestions also come in two that the subject's hand drop further, faster, than forms that roughly parallel direct and challenge what is specified in the standardized scale. ideomotor suggestions. Some cognitive sugges- Similarly, post-hypnotic amnesia, generally con- tions involve the production of percepts and sidered a difficult item, can be made easier by memories, such as the dream and positive audi- adopting a looser criterion for initial forgetting tory hallucination ('loudspeaker') items of or subsequent reversibility (Kihlstrom and SHSS:C; others involve the inhibition of per- Register, 1984). When item difficulty levels are cepts and memories, such as the negative visual adjusted in this manner, the three-factor struc- hallucination ('three boxes') and post-hypnotic ture still emerges (Tellegen and Atkinson, 1976; amnesia items of the same scale. In every case, Kihlstrom et al., 2006). Moreover, a cluster however, there is a more or less direct suggestion analysis of the original standardization data for that a state of affairs exists that does not accord SPSHS:I and I1 uncovered 12 distinct profiles of with objective reality. The Stanford scales, which hypnotic ability, independent of overall level of assess the subject's response to such suggestions, hypnotic response (Brenneman and Kihlstrom, thus represent a prototype that defines the 2006). A recent, highly sophisticated multivari- domain of hypnosis. ate analysis confirms that the three basic factors The distinctions between direct and challenge of hypnotizability-direct, challenge and cogni- ideomotor suggestions, and between ideomotor tive suggestions-are not an artifact of item and cognitive suggestions, has been repeatedly difficulty (Woody et al., 2005). The situation is confirmed by multivariate analyses of the thus roughly analogous to the structure of intel- Stanford and Harvard scales (e.g. E. R. Hilgard, ligence, with various group factors collected by a 1965; Spanos et al., 1980; for a review, see single overarching factor of general hypnotiz- Balthazard and Woody, 1985). Discovery of the ability. Hypnosis is closely related to primary multidimensional nature of hypnotic sugges- suggestibility, but the domain of primary tion led to the development of the Stanford suggestibility must be expanded to include Profile Scales of Hypnotic Susceptibility,Forms I perceptual-cognitiveeffects as well as ideomotor and I1 (SPSHS:I and 11; Weitzenhoffer and responses. 28 . CHAPTER 2 The domain of hypnosis, revisited

2.7. Imaginative experiences is defined as a mental image that is out of con- trol, hypnotic experiences are closer to halluci- Hypnotic experiences take place in the realm of nations than to ordinary mental images. In this imagination-there isn't really a balloon lifting respect, just as hypnosis should not be solely up the subject's hand, or glue holding the sub- identified with suggestion, so hypnosis should ject's hands together, or a loudspeaker on the not be narrowly identified with imagination. wall; nor does the age-regressed subject grow Accordingly, to relabel hypnotic susceptibility as smaller in the chair. Nevertheless, the relation- imaginative suggestibility (Braffman and ship between hypnosis and mental imagery is Kirsch, 1999; Kirsch and Braffman, 2001) would rather vexed. For example, hypnotizable indi- be to risk making a category mistake. viduals have no better mental imagery abilities than the rest of us-though here, admittedly, the fault might lie with the scales used to meas- 2.8. Perception, memory ure mental imagery (Kearns and Zamansky, 1984; K. S. Bowers, 1992; Glisky et al., 1995). Of and action course, it may simply be that better, more per- The phenomena of hypnosis are mostly cogni- formance-oriented, methods of assessing tive in nature, involving alterations in percep- imagery (Kosslyn et al., 1984) will yield substan- tion and memory. Hypnotized subjects perceive tial correlations with hypnosis, of a sort that things that aren't there, and fail to perceive have so far eluded empirical investigation. things that are there. They experience them- On the other hand, examination of the struc- selves as if they were young again. They dream ture of hypnotic suggestions, and of the phe- even though they are not asleep. They cannot nomenology of hypnotic experience, suggests remember what happened to them while they that hypnotic experiences bear only a tangential were hypnotized, and they remember things relationship to mental imagery as we typically that did not actually happen to them. Even the experience it. Note, first, that while hypnotic ideomotor phenomena of hypnosis are cogni- experiences surely take place in imagination, tive in nature, because the motor behaviors fol- hypnotic suggestions do not typically ask sub- low from suggestions for perceptual change. In jects to imagine anything. Consider, for exam- the direct suggestions, subjects perceive balloons ple, the Fly Hallucination item (#9) on tied to their wrists, or heavy objects in their HGSHS:A. Subjects are not asked to imagine hands, and then the limbs move accordingly. In that there is a fly buzzing around the room. the challenge suggestions, they feel their hands Instead, they are informed-it is suggested- glued together, and have difficulty taking them that there is one. And hypnotic subjects respond apart; they feel their outstretched arms stiffen- accordingly. Hypnotic experiences take place in ing, and then have difficulty bending them. So imagination, but they do not have the same too, for post-hypnotic suggestions, the essence experiential qualities as ordinary mental of which is not that subjects touch their ankles imagery. Mental images are deliberately, con- when the experimenter taps, but that they feel sciously constructed, while hypnotic experi- an urge to do so, do not know why and do not ences are generally accompanied by an remember the suggestion. The behavioral experience of involuntariness. It is this experi- responses by which we measure hypnotizability ence of involuntariness, not the vividness of follow from suggestion-induced changes in mental images that gives hypnotic experiences perception and memory. their hallucinatory quality. Perceptual-cognitive changes lie at the heart There are exceptions, of course, even on of most hypnotic phenomena, but cognition is HGSHS:A itself. But in general, when hypnotic not all there is to mental life. As Kant suggestions use the word imagine, they do so in put it, 'there are three irreducible faculties of much the same way that they use the word sleep. mind: knowledge, feeling, and desire' (Kant, If a mental image is defined as a percept in the 178111929, p. 14). So what of the other two absence of a stimulus, then hypnotic suggestions elements in the 'trilogy of mind' (E. R. Hilgard, definitely involve imagery. But if a hallucination 1980)? What role do emotion and motivation Perception, memory and action - 29 play in the phenomena of hypnosis? These have dissociate them-so that, for example, the mostly gone unexamined, but there are hints in hypnotized patient can feel pain but not suffer the literature that hypnosis can affect emotional from it. Interestingly, these suggestions produce and motivational processes as well. different patterns of brain activity: sensory pain For example, Damaser et al. (1963) employed activates the primary somatosensory cortex, hypnotic suggestion to investigate the physio- while suffering activates the anterior cirigulate logical correlates of various emotional states. gyrus (Rainville et al., 1997, 1999; Price et al., Levitt and his colleagues observed the effects of 2002). Emotional numbing may operate .in the suggested anxiety, covered by a further sugges- manner of a suggestion for analgesia, or amne- tion for post-hypnotic amnesia, on various sia, at the level of conscious awareness. Just aspects of test performance (Levitt, 1967; Levitt as Hilgard's 'hidden observer' studies showed and Chapman, 1979). Similarly, Blum and his that pain can be represented subconsciously colleagues explored the effects of hypnotically despite the experience of analgesia (Knox et al., suggested arousal on various aspects of per- 1974; E. R Hilgard et al., 1975,1978), so it may formance (e.g. Blum and Porter, 1972). This be that hypnotically suggested emotional numb- research, as well as Watkins's clinical work ing impairs explicit, or conscious, affective expe- on the 'affect bridge' (Watkins, 1971), set the rience but spares implicit, or unconscious, stage for Bower's use of hypnotically suggested representations of emotional state (Kihlstrom emotions in his pioneering studies of mood- et al., 2000). congruent and mood-dependent memory (Bower So far as motivation is concerned, it should be et al., 1978, 1980; Bower, 1981). The memory noted that Hilgard's attraction to the field was effects originally reported by Bower have partly an outgrowth of his longstanding interest since been replicated using nonhypnotic meth- in problems of motivation, and in particular the ods (e.g. Eich, 1995; Bower and Forgas, 2000). distinction between voluntary and involuntary However, difficulties in replicating the original acts (E. R. Hilgard, 1964).Theorists of a psycho- hypnosis experiments have perhaps contributed dynamic bent have also made use of hypnosis to to a declinein research on hypnotically elicited study problems of unconscious motivation-an emotion. This is an area that warrants further enterprise which E. R. Hilgard (1961,1964)also investigation-especially as interest in emo- endorsed. In one prominent line of research, tion is being revived within psychology, and based on a paradigm originally devised by Luria researchers seek reliable means of manipulating (1932), subjects received suggestions for a subjects' emotional states (Larsen and Sinnett, paramnesia, or false memory, that they had 1991; Westermann et al., 1996; Nummenmaa committed some socially undesirable act-the and Niemi, 2004; Goritz and Moser, 2006). suggestion itself was covered by a further sugges- Many of the cognitive effects of hypnosis tion for amnesia (Huston et al., 1934; Reyher, come in contrasting forms, such as the positive 1967; Sommerschield and Reyher, 1973). and negative hallucinations, and this may be the Alternatively, subjects received suggestions for a case for the emotional effects as well. In the variant on age regression, in which they were Bower studies, the hypnotic suggestion appears asked to relive a conflictual, ego-threatening to operate in the manner of a positive hallucina- event from childhood (Blum, 1967, 1979). tion-subjects experience an emotion in the Unfortunately, the demand characteristics of absence of the appropriate stimulus. By analogy such experiments are fairly clear, making inter- to negative hallucinations, Bryant and his col- pretation of the results difficult (Reyher, J969; leagues have employed suggestions for 'emo- Sheehan, 1969, 1971b). Nevertheless, the work tional numbing', which may reduce subjects' deserves to be remembered as an important line of conscious awareness of their emotional states experimental research on psychodynamic theory. (Bryant and Mallard, 2002; Bryant, 2005). Motivation is also relevant to a question Although hypnotic analgesia reduces both the that has intrigued investigators ever since the sensory-perceptual and affective components of beginning of experimental hypnosis research: pain (E. R. Hilgard, 1967; Knox et al., 1974), whether hypnotic suggestions of various kinds can appropriately worded suggestions can also enhance human performance-what Marcuse 30 . CHAPTER 2 The domain of hypnosis, revisited

(1959) called 'the generation of hypers' (as in One motivational application of hypnosis hyperpraxia, or enhanced muscular perform- is in sports psychology (Unestahl, 1979; ance, hyperesthesia, or enhanced sensory acuity, W. P. Morgan, 1980, 2002; Mairs, 1988; Taylor and hypermnesia, or enhanced memory) and et al., 1993). Unfortunately, field studies the 'wantos' (as in 'Want to make the weak as in this area have rarely taken advantage of strong as a lion?'). A vast experimental literature the analytical power of paradigms such as on the hypnotic transcendence of normal vol- the London-Fuhrer design. As a result, it is untary capacity, mostly with variants on the often unclear whether any improvements London-Fuhrer paradigm (London and Fuhrer, in performance are actually mediated by 1961) in which hypnotizable and insusceptible increased motivation-as opposed to self- subjects were tested in baseline, nonhypnotic distraction, analgesia-like pain relief or even and control conditions, yielded largely negative absorption in the athletic performance results (e.g. Evans and Orne, 1965; Orne, 1966; (i.e. 'flow'; Csikszentmihalyi, 1990; Grove and Sheehan and Perry, 1976). 'In general, these Lewis, 1996; Pates and Maynard, 2000). studies found that appropriately motivated insusceptible subjects performed as well or bet- 2.9. Classic instance ter on measures of muscular strength and endurance than hypnotizable subjects. Similarly, The effects of hypnotic suggestions-the classic there is no experimental evidence that hypnosis case or instance-are not experienced by, or reliably enhances learning or memory (Council observed in, everyone who is hypnotized. The on Scientific Affairs, 1985; Kihlstrom and Eich, phenomena that have enticed theorists from 1994; Whitehouse et al., 2005). Finally, provoca- James, Freud and even Pavlov to today are most tive evidence that hypnosis could improve visual likely to occur in those subjects who are most acuity in myopic subjects (Graham and highly hypnotizable (Heap et al., 2004). And, Leibowitz, 1972) has been called into question as a corollary, there is little point in studying by more recent analyses (Raz et al., 2004b). hypnosis in subjects who cannot experience Most of these studies employed direct sugges- it. Accordingly, hypnosis researchers devote tions and exhortations for improved perform- an extraordinary amount of time, effort and ance. However, Slotnick et al. (1965) obtained resources to assessing the hypnotizability of the different results when they added involving subjects who participate in their experiments instructions to the usual exhortations. In these (for a comprehensive summary of assessment instructions, subjects were asked not simply to procedures, see Barnier and McConkey, 2004). improve their performance, but also to think of The optimal screening procedure for hypnosis themselves as 'stronger and more capable'. Under research is to begin with HGSHS:A, which these conditions, exhortation plus involvement allows subjects to familiarize themselves with yielded a marked improvement in performance hypnotic procedures, and also provides a first over exhortations alone, and this was true even approximation of their hypnotizability. Then, when the involving instructions were adminis- high-scoring subjects can be invited to return tered in a nonhypnotic control condition. for a final assessment with SHSS:C. The Because all the subjects in this experiment were Stanford group developed the Stanford Profile hypnotizable, it seems likely that the effects of Scales of Hypnotic Susceptibility, Forms I and I1 the involving instructions were mediated by the (SPSHS:I and 11), to permit more fine-grained capacity for absorption and imaginative involve- assessments of hypnotizability, particularly with ment that is correlated with hypnotizability. respect to different patterns of ability within the Although it would be useful to replicate this high range (E. R Hilgard et al., 1963; Weitzenhoffer experiment with insusceptible subjects as well, it and Hilgard, 1963, 1967; for a review, see is possible that imagining oneself stronger can in McConkey and Barnier, 2004), but these never fact make one (a little) stronger, in the manner entered into common use. Instead, SHSS:C can of the self-fulfilling prophecy (Snyder, 1984), be 'tailored' for special assessment purposes, whether that imagination takes place in hypno- without losing its value as a standard psychome- sis or not. tric instrument (Hilgard et al., 1979). Classic instance . 31

Nowhere else in psychology, except perhaps in It is not clear why this is the case. Some clini- neuropsychology, is so much effort devoted to cians, especially those working in the tradition subject screening and selection prior to formal of Milton H. Erickson, may discount the impor- experimental research. In this respect, hypnosis tance of individual differences in hypnotizabil- may serve as a model for individual differences ity (Frankel, 1985). They may fear. that research throughout psychology. Unfortunately, a finding of low hypnotizability will reduce a the sheer economics of such a rigorous assess- patient's motivation for treatment, but the scant ment scheme naturally provides an incentive for evidence available suggests that this is not the investigators to take shortcuts. One of these is a case (Frankel et al., 1979). It would seem obvi- group-administered version of SHSS:C, the ous that candidates for , or for Waterloo-Stanford Group Scale of Hypnotic hypnosis as an adjunctive treatment, should be Susceptibility: Form C (WSGC; K. S. Bowers, screened to determine whether they are, in fact, 1993)-which, while offering some of the same hypnotizable. While it is true that such assess- economies as HGSHS:A, eliminates the individ- ment takes time, clinicians often take time to ual administration that permits detailed inquiry administer other instruments, such as the into the subject's experience. Another trend, Rorschach and the Thematic Apperception Test, more disturbing, is reliance on HGSHS:A as the that are less relevant to treatment. In any event, sole screening instrument for hypnosis research. the Stanford group developed a set of abbrevi- The problem is that HGSHS:A does not contain ated scales that afford valid assessment of hyp- enough difficult and cognitive items to provide notizability in clinical contexts (A. H. Morgan a good assessment of hypnotic abilities; as a and Hilgard, 1978-1979a,b). These should be result, its ability to identlfy 'hypnotic virtuosos' more widely used in clinical research and prac- is relatively poor (Register and Kihlstrom, tice than they seem to be. 1986). In this respect, SHSS:C remains the 'gold Can anyone become a hypnotic virtuoso? standard' of hypnotizability measures. Hull thought of hypnosis as a habit phenome- Another unfortunate trend is the prolifera- non, but, while practice may help subjects tion of alternative scales to measure hypnotiz- become hypnotized more readily, it does not ability, including the Barber Suggestibility Scale seem to make them more hypnotizable. We now (BBS; Barber, 1965), Carleton University think of hypnotizability as a cognitive skill, and Responsiveness to Suggestion Scale (CURSS; we generally think of skills as things that can be Spanos et al., 1983a), Creative Imagination Scale acquired, refined and perfected through experi- (CIS; Wilson and Barber, 1978) and Hypnotic ence. Nevertheless, attempts to modify hypnoti- Induction Profile (HIP; Spiegel, 1972; Orne zability have mostly produced ambiguous et al., 1979). When different laboratories use results (Diamond, 1974, 1977a,b, 1982; Perry, different scales to measure hypnotizability, 1977). More recently, a package known as the the virtues of standardization are lost, and it Carleton Skills Training Program (CTSP) has becomes all the more difficult for one labora- been proposed as a means for enhancing hypno- tory to replicate and extend another's work. tizability (Gorassini and Spanos, 1986; Gorassini Some of the newer scales are shorter than et al., 1991; Gorassini, 2004). Here, too, however, HGSHS:A and SHSS:C, and thus arguably more the enhancement of hypnotizability appears to economical to use, but abbreviation comes at be heavily laced with demands for overt behav- the expense of content validity. The CURSS does ioral compliance (Bates et al., 1988; Bates and include assessments of subjective experience Brigham, 1990; Bates and Kraft, 1991; Bates, and experienced involuntariness, but these can 1992). Although this point has been vigorously easily be added to the Stanford and Harvard debated by advocates of the CSTP (e.g. Spanos scales without compromising their essential et al., 1989-1990; Gorassini, 2004), few if any properties (P. Bowers, 1982; Register and laboratories have abandoned subject selection Kihlstrom, 1986). based on formal measurements of hypnotizabil- As valuable as the standardized scales have been ity, such as HGSHS:A and SHSS:C, in favor of for research purposes, it is sadly the case that creating virtuoso subjects wholesale out of uns- hypnotizability is rarely measured in the clinic. elected subjects with the CSTP. 32 . CHAPTER 2 The domain of hypnosis, revisited

The phrase classic instance can refer to depth behavioral compliance. This point was made of hypnosis as well as to hypnotic ability. In the early on in a series of studies of Barber's 'task nineteenth century, Braid attempted to charac- motivation' paradigm of hypnosis, which puts terize the various stages of hypnotic sleep, very strong pressure on subjects for overt behav- Charcot described catalepsy, lethargy and som- ioral compliance with suggestions (Barber, nambulism as representing three stages of hyp- 1969, 1972). In one study, Barber and Calverley nosis, and both Liebeault and Bernheim offered (1964) reported that subjects in a task motiva- criteria for diagnosing the various stages of hyp- tion control condition gave reports of the nosis (for a review, see E. R. Hilgard, 1965). 'reality' of suggested hallucinations that were Indeed, some of the very first scales of hypnotiz- comparable with those given by hypnotic sub- ability were actually labeled as scales of hypnotic jects. K. S. Bowers (1967) replicated this finding, depth (Friedlander and Sarbin, 1938; LeCron, but found that the reality ratings of task motiva- 1953). However, modern scales of hypnotic tion subjects returned to baseline levels follow- depth have usually been offered as supplements ing demands for honesty in reporting. Next, to, rather than substitutes for, the assessments of Spanos and Barber (1968) confirmed this find- hypnotic ability provided by the conventional ing, but found that the hallucination ratings of standardized scales (e.g. Tart, 1970). Perhaps the hypnotic subjects were not corrected by honesty simplest measure of hypnotic depth is a 1-10 demands. Finally, K. S. Bowers and Gilmore rating of 'how deeply hypnotized' subjects feel (1969) found that honesty ratings corrected the themselves to be (O'Connell, 1964; Register and hallucination reports of simulating, but not real, Kihlstrom, 1986). hypnotic subjects. The entire cycle of research Such global depth ratings, based on purely just summarized underscores the importance of subjective (and probably idiosyncratic) criteria, subjective conviction in distinguishing what is are likely to reflect little more than the subject's hypnosis from what is not and who is genuinely involvement with the hypnotic procedure, and hypnotized from who is not. are to be taken with a grain of salt. However Orne's real-simulator design also provides a depth is measured, there is a definite conceptual means to this end. The real-simulator paradigm difference between depth and ability. Presumably, was intended, largely, to serve as a means to highly hypnotizable subjects can experience verify the ecological validity of laboratory 'lighter' or 'deeper' stages of hypnosis, in much experiments on hypnosis and other phenomena the same way that a virtuoso pianist can play (Orne, 1959,1962, 1969, 1970, 1972, 1973). But more or less well, depending on the circum- because Orne's hypnosis research was part of a stances. Someone who lacks the ability to play larger interest in the objective study of subjec- the piano, however, can never play really well. tive states, including sleep and the detection of Insusceptible subjects, on this model, lack the deception, the real-simulator design was also a ability to become deeply hypnotized, no matter vehicle for careful post-experimental inquiry how hard they try. The classic instance, the pro- into subjects' private experiences of hypnosis. totype or defining example of the domain of Similarly, the Experiential Analysis Technique hypnosis, is a hypnotizable person who is deeply was developed as a means for systematically hypnotized. inquiring into subjects' (retrospective) impres- sions of hypnotic procedures and their response 2.10. Subjectlve conviction to them (Sheehan et al., 1978; Sheehan and McConkey, 1982; McConkey and ~arnier, Subjective experience lies at the heart of hypno- 2004). Both procedures have their place in sis. It is not interesting that a hypnotized subject hypnosis research, but they--especially the will lower his outstretched arm when told that real-simulator design--can also be expensive to it is becoming heavy. What is interesting is that implement. the arm actually begins to feel heavy. It is Although the Stanford-type hypnotizability the subject's conviction that the suggested scales were deliberately constructed with behav- event is really happening that distinguishes ioral as opposed to subjective measures of a genuine hypnotic experience from overt response, assessments of subjective experience Experienced involuntariness . 33

can be easily added to them. For example, must be fit at all costs. For example, we now subjects can simply be asked whether each sug- know that hypnosis need not be relaxing, and gestion succeeded in producing its intended hypnotized subjects need not be drowsy effect (Register and Kihlstrom, 1986). One com- (Vingoe, 1968; Banyai and Hilgard, 1976; Malott parative study found that ratings of subjective and Goldstein, 1981; Alarcon et al., 1999). experience corrected the behavioral scores of Whether subjects are relaxed or active, alert or the BSS strongly downward, but had signifi- drowsy, what really belongs in the domain of cantly less effect on scores of the SHSS:A (Ruch hypnosis is subjective conviction in the experi- et al., 1974). Perhaps the correlations between ences that are suggested to them. As Sutcliffe subjective experience and objective response are (1960,1961) put it, the hypnotized subject is, in high on the Stanford-type scales because the some sense, deluded about the actual stimulus scales themselves put so much emphasis on state of affairs. subjective experience, despite their behavioral scoring. Still, direct assessment of subjective experience makes the point that subjective 2.11. Experienced experience lies at the heart of the domain of Involuntariness hypnosis, and that behavioral responses flow from subjective conviction. Whether subjects are relaxed or active, alert or Of course, the subjective experience of hyp- drowsy, another element in the subjective expe- nosis can go beyond subjective conviction in the rience of hypnosis is the experience of involun- suggested effects (Sheehan and McConkey, tariness in response to hypnotic suggestions. 1982; McConkey and Barnier, 2004). Following The outstretched arm does not just feel heavy: it in the tradition of nineteenth-century descrip- appears to become heavier all by itself, without tions of the depth or stages of hypnosis, a num- the subject deliberately constructing the image. ber of modern investigators have proposed that The experience of involuntariness is part and the subjective experience of hypnosis can be parcel of subjective conviction: one cannot assessed along a number of different dimen- believe that one's arm has become light, being sions. For example, Shor (1962) proposed that pulled up by helium balloons, if one is deliber- hypnotic depth be evaluated along three con- ately imagining that it is so, or voluntarily rais- ceptually independent dimensions: the loss of ing the limb. The experience of involuntariness the generalized reality orientation (Shor, 1959); is what distinguishes a suggestion from an nonconscious involvement; and archaic involve- instruction (Weitzenhoffer, 1974, 1980); non- ment. Later, he suggested five additional dimen- conscious involvement (Shor, 1959,1962,1979) sions on which the phenomenal experience is also what distinguishes hypnotic experience of hypnosis should be assessed (Shor, 1979; from mere behavioral compliance. Kihlstrom et al., 1989): drowsiness; physical and In view of the centrality to hypnosis of the mental relaxation; mental imagery; absorption experience of involuntariness, it is somewhat in the ongoing experience; and access to nor- surprising that it is so seldom considered in the mally unconscious ideas and memories. Along assessment of hypnotizability. Both HGSHS: the same lines, Field (Field, 1965; Field and A and SHSS:C rely exclusively on observed or Palmer, 1969; Kihlstrom et al., 1989) and Pekala self-reported behavioral response, as does the (Pekala et al., 1985; Pekala and Kumar, 2000; BSS. This general failure to include an inquiry Pekala, 2002) have developed questionnaire into the experience of involuntariness was criti- techniques for assessing a number of subjec- cized by Weitzenhoffer (1980a) himself, as part tive experiences thought to be associated with of his analysis of the classic suggestion efject hypnosis. (Weitzenhoffer, 1974). Following the argument To some extent, these proposals obviously of Bernheim (188611889, p. 125), who asserted reflect their originators' theoretical preconcep- that 'The most striking feature of a hypnotized tions concerning the nature of hypnosis. As subject is his automatism', Weitzenhoffer such, they risk constituting a kind of Procrustean asserted that only involuntary responses to bed into which subjects' experience of hypnosis suggestion should count as truly hypnotic in 34 . CHAPTER 2 The domain of hypnosis, revisited nature (see also Weitzenhoffer, 1980). Thus, at these are easily inserted into the procedure. least in principle, assessments of hypnotizability Again, perhaps the simplest procedure is a rating that do not assess involuntariness may be con- scale, with opposite poles labeled 'Deliberate, taminated by mere behavioral compliance. In Effortful, Voluntary' and 'Automatic, Effortless, reply, E. R. Hilgard (1981) argued that the vast Involuntary' (P. Bowers, 1982). However, ratings majority of subjects experience their response to in the midrange of a continuous dimension of hypnotic suggestions as involuntary, so that the involuntariness remain somewhat ambiguous. degree of contamination may not be great. While the meanings of high and low ratings on Of course, the degree of such contamination such a scale are fairly clear, intermediate ratings is an empirical question. K. S. Bowers (1981), might mean that the subject's response was per- examining response to an abbreviated version of ceived as partially voluntary and partially invol- SHSS:A, found that 80 percent of passed items untary; or a response that began voluntarily were experienced as involuntary behaviors, and might have continued involuntarily. Accordingly, only 20 percent as voluntary. Moreover, subjects P. Bowers et al. (1988) introduced a categorical who experienced their response as involuntary rating system, which allowed for such alterna- scored higher on SHSS:A, and on a subsequent tives as well as the more extreme alternatives SHSS:C, than those who did not, regardless of of deliberate versus involuntary responding to whether they passed the item according to the suggestion. behavioral criterion. Subsequent studies of It is one thing to assess the experience of invol- scales of the Stanford type also found low rates untariness; it is quite another to explain it. Early of voluntary response (P. Bowers, 1982; Farthing authorities, such as Bernheim (1886/1889), really et al., 1983; P. Bowers et al., 1988). Such studies seem to have believed that the hypnotic subject indicate that most positive responses to hyp- was some sort of automaton. Similarly, Arnold's notic suggestions are, in fact, associated with the (1946) theory of ideomotor action held that experience of involuntariness. behavioral responses to hypnotic suggestions Still, the fact that some positive behavioral occurred automatically whenever the subject responses are experienced as voluntary suggests vividly imagined some suggested state of affairs. that ratings of experienced involuntariness can Beginning in the mid- 1970s, cognitive be of value in the assessment of hypnotizabil- psychologists began to elaborate a technical ity-a point with which E. R. Hilgard (1981) concept of automaticity, couched in the frame- concurred. For example, in the normative study work of limited-capacity models of attention of the CURSS, which includes subjective and and information processing (LaBerge and involuntariness scores as well as the usual objec- Samuels, 1974; Posner and Snyder, 1975; tive score, requiring subjects to pass both the Schneider and Shiffrin, 1977). According to this objective criterion and rate the response as at view, automatic processes share four character- least moderately involuntary, cut the mean score istics in common: (1) they are inevitably evoked in half and shifted the distribution of scores by the presence of certain stimuli in the envi- dramatically to the left (Spanos etal., 1983a,b,c). ronment; (2) once evoked, they are incorrigibly While such findings suggest that the CURSS executed, in a ballistic fashion; (3) they are behavioral scores might be heavily contami- effortless, in the sense that they do not consume nated with compliance, Spanos and his col- cognitive resources; and (4) they are processed leagues have argued that contamination extends in parallel, so that they do not interfere with to the Stanford scales as well (Spanos et al., other ongoing cognitive processes. So defined, 1986a,b). However, this was clearly not the case automatic processes are involuntary in the same in the studies of Kenneth and Patricia Bowers, way that reflexes and instincts are involuntary. or in Hilgard's own studies, where honesty Based on this technical definition, however, demands and involuntariness ratings had little it appears that hypnotic experiences are not effect on SHSS scores. involuntary after all (Lynn et al., 1990). For Although the published versions of the example, response to post-hypnotic suggestion Harvard and Stanford scales do not contain is sensitive to the context in which the cue is assessments of experienced involuntariness, presented (Spanos et al., 1987); even highly Altered state of consciousness . 35 hypnotizable subjects can resist hypnotic have revived Arnold's theory of ideomotor suggestions (E. R. Hilgard, 1963); and execution responding, suggesting that hypnotic experi- of a post-hypnotic suggestion consumes cogni- ences are an automatic consequence of positive tive capacity, so that post-hypnotic responses response expectancies. It will take some time to can interfere with other resource-demanding sort all of this out, but at present it appears that processes (Hoyt, 1990). Accordingly, many the theoretical battleground in hypnosis has modern theories of hypnosis incline toward the shifted, from explaining response to hypnotic view that hypnotic experiences are not actually suggestions in general, and debating the mecha- automatic in nature, even though they may be nisms of various hypnotic suggestions in partic- experienced as involuntary. In E. R. Hilgard's ular, to accounting for subjects' experience (1977) neodissociation theory of divided con- of involuntariness in response to hypnotic sciousness, for example, automatic processes suggestions. running in parallel serve to illustrate the idea of divided consciousness. But the theory itself suggests that the experience of involuntariness 2.12. Altered state of occurs because the cognitive module that exe- cutes the suggestion does so outside of phenom- enal awareness (Kihlstrom, 1992a, 1998). Over the years, much ink has been spilled over As another example, Spanos's (1986a) social- whether, and in what respects, hypnosis repre- cognitive view of hypnosis explains the experi- sents an altered state of consciousness (Ludwig ence of involuntariness in terms of self-deception and Levine, 1965; Chaves, 1968; Spanos, 1970, on the part of the subject, who mistakenly 1986b,1987a,b;Spanos and Chaves, 1970; Barber, attributes his or her response to external rather 1972; Sarbin and Coe, 1972; E. R. Hilgard, than internal factors-a mistake encouraged by 1973a,b, 1992; Blum, 1978; Kihlstrom, 1992a, various features of the social situation in which 2007; Sarbin, 1992; Kirsch and Lynn, 1998c; hypnosis takes place. In both Hilgard's and Oakley, 1999a, 1999b; Kallio and Revensuo, Spanos's theories, the experience of involuntari- 2000, 2003). Sometimes the debate has been ness has some of the qualities of an illusion- couched in terms of the validity of trance or although for Hilgard the source of the illusion is special-process theories of hypnosis. Sometimes to be found in the communications among cog- it has been framed as a conflict between nitive subsystems, while for Spanos the source is alternative paradigms for the investigation of to be found in the structure of suggestions and the phenomenon. Sometimes, it seems to be other features of the social milieu. It should be a local manifestation of a broader conflict over noted, in passing, that Spanos has elsewhere whether mentalistic entities such as conscious- suggested that reports of involuntariness stem ness have any place in a scientific explanation from subjects' strategic attempts to create an of behavior. impression that they are deeply hypnotized Nevertheless, the conclusion that hypnosis (Spanos et al., 1985). In this view, the experience reflects an altered state of consciousness seems of involuntariness is no such thing. unavoidable (Kihlstrom, 2005). After all, con- In stark contrast, some modern approaches sciousness has two principal aspects (Kihlstrom, seem to hold that hypnotic experiences actually 1984): monitoring ourselves and our environ- occur automatically. For example, Woody and ment, so that objects, events and our internal Bowers (1994; see also Woody and Sadler, 1998) mental states are accurately represented in phe- drew on neuropsychological theories to suggest nomenal awareness; and controlling ourselves that hypnosis alters the functioning of executive and the environment, through the voluntary control systems associated with the prefrontal initiation and termination of thought and cortex, with the result that hypnotic responses action. And hypnosis alters both of them: hyp- are truly involuntary even if they are not techni- notized subjects see things that are not there, cally automatic (see also Haggard et al., 2003). and fail to see things that are there; they fail to And, somewhat paradoxically, Kirsch and Lynn remember things that they just experienced, and (Kirsch and Lynn, 1998b,c, 1999; Kirsch, 2001) they remember things that didn't happen; they 36 - CHAPTER 2 The domain of hypnosis, revisited cannot control their bodily movements, and hypnosis was hypersuggestibility; but while hyp- they execute post-hypnotic suggestions without nosis may-may--enhance suggestibility, at the knowing why they are doing so. From this point very least it is clear that suggestibility is some- of view, it would seem that the only way to deny thing that also occurs in the normal waking state that the phenomena of hypnosis reflect alter- (Weitzenhoffer and Sjuberg, 1961; Braffinan and ations in consciousness would be to deny that Kirsch, 1999; Kirsch and Braffman, 2001). More the phenomena themselves are genuineto recently, Orne (1959) suggested that hypnosis assert, for example, that hypnotic subjects really was characterized by trance logic, which he do feel pain, and really do remember, despite described (informally, to colleagues; but, alas, what they say after they have been given sugges- never in print) as a kind of 'peaceful co-existence tions for analgesia and amnesia. between illusion and reality'. For example, Perhaps, though, the problem lies in the way Orne reported that some hypnotized subjects, that altered state is defined. If we believe that hallucinating a companion, saw through their every state of consciousness is associated with hallucination to the back of the chair on which some unique physiological signature, much as they sat; or they also saw the real person, sitting sleep is associated with the absence of alpha nearby; simulators, Orne reported, never did activity in the electroencephalogram (EEG) and these things. Everyone who has ever worked with dreaming with the occurrence of rapid eye a hypnotized subject has observed trance logic; movements (REM), then the lack of a physio- but the implication that trance logic was unique logical indicator for hypnosis may be taken as to hypnosis-it's 'essence'-was vigorously evidence that hypnosis is not an altered state of challenged by R. F. Q. Johnson (1972; for a cri- consciousness after all. But of course, this puts tique, see E. R. Hilgard, 1972; for a rejoinder, see the cart before the horse. Physiological indices R. F. Q. Johnson et al., 1972). Later studies are validated against self-reports, as when employing a battery of test items confirmed Aserinsky and Kleitman (1953) awakened their that hypnotizable subjects were more likely to subjects up during periods of REM and non- display trance logic than insusceptible subjects REM sleep to ask them if they were dreaming. (e.g. Peters, 1973; Obstoj and Sheehan, 1977; For this reason, physiological variables have no Perry and Walsh, 1978), but it also became privileged status over introspective self-reports clear that trance logic was observed in other as indices of consciousness. situations as well, such as nonhypnotic imagina- Arguably, it would be useful if states of con- tion (McConkey et al., 1991). sciousness had distinct physiological correlates. Despite the human species' deep epistemic But our present knowledge of mind-body rela- desire to carve nature at its joints, and slot tionships is simply not sufficient to make such different states of consciousness into discrete correlates a necessary part of the definition. categories, in the final analysis it may be best to After all, cognitive neuroscience has made very treat hypnosis and other altered states of con- little progress in the search for the neural corre- sciousness as natural concepts, represented by a lates of ordinary waking consciousness prototype or one or more exemplars, each con- (Metzinger, 2000; Coltheart, 2006a,b). It is very sisting of features that are only probabilistically difficult to infer from a particular pattern of associated with category membership, with no brain activity just what the subject is doing clear boundaries between one altered state and (Poldrack, 2006). How far in the future do the another, or between altered and normal con- neural correlates of altered states of conscious- sciousness (E. R Hilgard, 1969; Kihlstrom: 1984, ness, such as hypnosis, await? And even when 2005). And because we cannot have direct knowl- they become available, how reliably will we be edge of other minds, altered states of conscious- able to determine that subjects have been hyp- ness must also remain hypothetical constructs, notized, just by examining their brains? inferred from a network of relationships among Even at the psychological level of analysis, it variables that are directly observable (Garner may not be possible to find any unique cognitive et al., 1956; Campbell and Fiske, 1959; Stoyva and or behavioral change associated with hypnosis. Kamiya, 1968),much in the manner of a psychi- Hull (1933) thought that the hallmark of atric diagnosis (Orne, 1977). From this point Altered state of consciousness - 37 of view, the diagnosis of an altered state of a consequence of the subject's altered subjective consciousness can be made with confidence to experience, and is of no interest in the absence the extent that there is convergence among four of corresponding subjective experience. There is kinds of variables: an induction procedure; really no getting away from self-reports; the alterations in subjective experience; associated methodological trick is to collect them under changes in overt behavior; and physiological circumstances where subjects believe it is legiti- correlates. mate for them to reflect accurately on their Operationally, an altered state of conscious- experiences. ness can be defined, in part, by the means Because both self-reports and overt behaviors employed to induce it-or, alternatively, as the are under voluntary control, and thus subject to output resulting from a particular input distortion by social influence processes, hypno- (Barber, 1969). Operational definitions of this sis researchers have long been interested in psy- sort are a residue of functional behaviorism in chophysiological indices of response. Over psychology, but the role of an induction proce- the years, a number of such indices have dure in hypnosis remains open. Certainly an been offered, including galvanic skin conduc- induction procedure helps to define the situa- tance (O'Connell and Orne, 1968), EEG alpha tion as hypnosis, as opposed to something else, (R. A. Dumas, 1977) and theta (Crawford so that the subject has some sense of what to and Gruzelier, 1992) activity, and increased expect and what to do. But it is not sufficient to activation of both right (MacLeod-Morgan produce hypnosis: the subject must also be hyp- and Lack, 1982) and left (Maquet et al., 1999) notizable--and then there is the nontrivial fact cerebral hemispheres, but these have often that any effect elicited while the subject is hyp- proved to be artifacts of confounding variables notized can also be elicited outside hypnosis, by such as relaxation, or otherwise not intrinsic to means of post-hypnotic suggestion. And an hypnosis. The ambiguities and controversies induction procedure may not be necessary surrounding hypnosis were not resolved by either: highly hypnotizable subjects may not psychophysiology, and they are unlikely to be benefit much from an induction, and highly resolved by neuroscience. experienced subjects may not need the ministra- Because subjects can have a wide variety of tions of the hypnotist in order to enter hypnosis. experiences while they are hypnotized, it was In contrast, introspective self-reports of probably a mistake to expect that there would be changes in subjective experience would seem to any neurophysiologicalcorrelates of hypnosis in be central to the definition of any altered state of general, following an induction procedure but consciousness. After all, the domain of hypnosis in the absence of any specific suggestions. is defined by suggested changes in perception, Investigators who are interested in the neural memory and the voluntary control of behav- correlates of hypnosis are more likely to find ior-analgesia, amnesia, the experience of something interesting when they focus on the involuntariness, and the like. If the hypnotist correlates of specific hypnotic suggestions-as gives a suggestion-for example, that there is an in brain imaging work that shows specific object in the subject's outstretched hand, getting changes in brain activity corresponding to heavier and heavier-and the subject experi- hypnotic auditory (Szechtman et al., 1998) and ences nothing of the sort, it is hard to say that he visual (Kosslyn et al., 2000) hallucinations, anal- or she has been hypnotized. gesia (Rainville et al., 2002) or agnosia (Raz Of course, self-reports have always made psy- et al., 2005). Note, however, in these cases, the chologists nervous, even in the heyday of intro- brain signature associated with the hypnotic spectionism (Boring, 1953; Robbins, 2004). effect was not unique to hypnosis. In the Accordingly, another residue of behaviorism is a Kosslyn et al. (2000) study, for example, hypno- methodological choice to focus on overt behav- tized subjects received a suggestion to perceive ior. If a subject hallucinates an object in his out- a grayscale stimulus as colored, and a colored stretched hand, and feels it grow heavier and stimulus in grayscale. A positive response to heavier, eventually his arm ought to drop down these suggestions was associated with changes in to his side. Note, however, that overt behavior is the 'color area' of the occipital cortex, but these 38 - CHAPTER 2 The domain of hypnosis, revisited changes were the same as those observed when and the hypnotist; in the case of self-hypnosis, nonhypnotized control subjects perceived col- one person takes on both social roles. Then there ored or grayscale stimuli, or when they simply is the situation in which hypnosis takes place, imagined the stimuli as such. The brain changed including the physical environment (laboratory, with the experience, but the origins of the clinic), as well as the whole socio-cultural matrix experience-whether in stimulation, hypnotic that surrounds the transaction- Mesmer, the suggestion or vivid imagination-did not much Svengali myth, , Saturday-morning matter. cartoons, The Manchurian Candidate, the Setting aside the issue of how altered states of listings of hypnotists in the telephone directory, consciousness can be defined in general, how can advertisements in the newspapers and other we characterize the alterations in consciousness media, and all the rest. Hypnosis is linked in the observed in hypnosis? The clearest answer is that popular mind with persuasion, compliance and the core phenomena of hypnosis-the ones that other aspects of social influence, including sub- really matter, and that distinguish the domain of liminal influence-a link that can reach mytho- hypnosis from that of other forms of suggestibil- logical proportions. A recent Google search on ity-entail a division in consciousness affecting the terms Hitler and hypnotist yielded 48 000 hits, percepts, memories and other mental contents including 'The George W. Bush Hypnosis File'. that are normally accessible to conscious aware- And not just in the popular mind: George ness and are instead processed subconsciously Estabrooks, a leading authority on hypnosis (E. R. Hilgard, 1977; Kihlstrom, 1984, 1992a, before its current revival, said of Hitler: 'We can, 1998,2007; K. S. Bowers and Davidson, 1991). I think, make out a very convincing case that In post-hypnotic amnesia, the phenomenon that basically Hitler's emotional domination of the gave hypnosis its name, explicit memory or con- crowd ... was only the attack of the stage hypno- scious recollection is impaired, but priming and tist one step removed' (Estabrooks, 194311957, other expressions of implicit or unconscious pp. 120-12 1). But we do not have to go as far as memory (Schacter, 1987) are spared. Post- Estabrooks to acknowledge that hypnosis pro- hypnotic suggestion can likewise be construed vides much grist for the social-psychological mill. as a failure of conscious prospective memory This was true even before there was a social (Einstein and McDaniel, 1990; Graf and Uttl, psychology. The Franklin Commission's studies 2001; Zimmer et al., 2001); the post-hypnotic of the role of imagination in mesmerism are rec- response is, in this view, an implicit expression of ognized today as the first experiments in psychol- memory for the post-hypnotic suggestion itself. ogy (Kihlstrom, 2002). In his chapter on hypnosis Analgesia suggestions disrupt explicit perception in 'Principles of Psychology', James (189011980) (Kihlstrom et al., 1992) of the pain stimulus, but underscored the role of the subject's expectations leave implicit expressions of pain, such as psy- and the hypnotist's skill in developing rapport, chophysiological responses, intact. Hilgard's and suggested that 'the bodily symptoms of the 'hidden observer' is a metaphor for the continu- Salpetriere patients', which Charcot attributed to ing subconscious representation of the pain neurological changes, were 'all of them results of stimulus. Dissociations between explicit and expectation and training' (p. 1198). Suggestion, implicit memory, and between explicit and as exemplified by hypnosis, was one of the 'simple implicit perception are not a unique signature of and sovereign' concepts (the others were syrnpa- hypnosis: they are also observed elsewhere, in a thy and imitation) by which pre-experimental wide variety of normal and pathological condi- social psychologists sought to explain interper- tions. But they do appear to be the signature of sonal behavior (Allport, 1954). Ross's 1908 text- the kind of alteration in consciousness that book, the first to have 'social psychology' in its occurs within the domain of hypnosis. title, attempted to explain all social behavior in terms of suggestion and imitation-terms which 2.13. Social lnteractlon he used interchangeably(Ross, 1908).McDougall's text, following only a few months later, offered a At the very least, hypnosis entails a dyadic rela- more extensive set of principles, but suggestion tionship between two individuals, the subject and submission still played a large role in his Social interaction . 39

approach (McDougall, 1908). Both made exten- 'socio-cognitive' theory described hypnosis as sive references to the literature on hypnosis, a strategic enactment shaped by the subject's especially the work of the Nancy School of understanding of task demands, as negotiated Liebeault and Bernheim (Gauld, 1992). with the hypnotist, in the context of specific his- A fully fledged social-psychological approach torical circumstances. For example, Spanos to hypnosis had to wait until after the Second argued that reports of experienced involuntari- World War, when the emergence of social psy- ness were in part misattributions shaped by the chology as an experimental discipline coincided structure of the suggestions administered to sub- with a revival of research interest in hypnosis. jects (Spanos and DeGroh, 1983), as well as a Interestingly, Sarbin's (1954) role theory, strategy for subjects to present themselves as intended as a general theoretical framework for deeply hypnotized (Spanos et al., 1985). These understanding social behavior, found its most elaborations of role theory, coupled with a popular application in hypnosis (Sarbin, 1950; debunking tendency (e.g. Spanos et al., 1982), Sarbin and Andersen, 1967; Sarbin and Coe, also left the impression that hypnotic subjects 1972; Coe and Sarbin, 1991). Unfortunately, the were engaged in something akin to faking. theory's reliance on a dramaturgical metaphor Despite the similarity in names, a rather for behavior led some to conclude that hypnosis different perspective on hypnosis is found in the was somehow akin to faking. Sarbin and his 'social cognitive' approach to hypnosis offered associates repeatedly disavowed this interpreta- by Lynn, Kirsch and their colleagues (Kirsch, tion-although, to be fair, the theory's reliance 1991; Lynn and Rhue, 1991; Kirsch and Lynn, on a definition of role-playing as as-if behavior 1995, 1998b,c)-sometimes with a hyphen, certainly encouraged the idea that hypnotized sometimes without. Partly rooted in Rotter's subjects weren't analgesic, amnesic, and so on- (1954) cognitive-social learning theory of per- they were only behaving as if they were. Still, sonality, the theory emphasizes the importance with such concepts as role perception, role enact- of response expectancies as determinants of ment, role location, self-role congruence, role both behavior and experience (Kirsch, 1985; expectations, role skills, role demands, role prepa- Kirsch and Council, 1989). The antecedent ration and the audience, role theory certainly expectancies are shaped by the usual sorts of offered a rich vocabulary for the analysis of the interpersonal processes, including all the sorts interpersonal aspects of hypnosis. of interpersonal influence that social psycholo- Role theory emerged from a sociological gists study. But once these expectancies are social psychology, which emphasizes explana- formed, their causal effect on hypnotic respond- tory concepts (such as role) that reside in the ing is mediated by a process of ideomotor action institutional, societal and cultural context of very similar to that described by Arnold (1946). individual behavior, and that rejects mentalistic In this social-cognitive model, social processes constructs. As their labels indicate, two other shape expectancies and other cognitions; and social-psychological approaches are more explic- response expectancies generate responses by a itly allied with the cognitive traditions in psycho- mechanism similar to the self-fulfilling logical social psychology, which place consider- prophecy (Merton, 1948) and other expedancy- able weight on the individual's internal beliefs, confirmation processes (Snyder and Swann, attitudes and explanations. Spanos's socio-cognitive 1978; Darley and Fazio, 1980; Snyder, 1984; perspective (Spanos, 1991) began as a revision of Jones, 1986). Barber's (1969) task-motivational approach to The difference between Kirsch and L+n's hypnosis, which emphasized the role of attitudes social-cognitive approach and Spanos's socio- and expectancies, as well as the subject's willing- cognitive approach can be seen in the analysis of ness to think and imagine with the themes of experienced involuntariness. Expectancies, suggestions (Barber et al., 1974). shaped by suggestions and other aspects of the Spanos's theory then spent time as a 'cognitive- social context function like ideas; and the idea of behavioral perspective' (Spanos and Chaves, an action leads automatically to its execution. 1989a,b) and as a 'social-psychological interpre- Thus, involuntariness is neither a misattribution tation' (Spanos, 1986a). In its final form, Spanos's nor an element of strategic self-presentation; 40 - CHAPTER 2 The domain of hypnosis, revisited instead, it is a subjectively convincing phenome- twenty-first century, we are reminded nal experience that follows from the mechanism that, as Hull prepared the monograph summa- that links suggestions to responses. Hypnosis rizing his research program, he also left the is simply a more general case of these basic field detailed descriptions of 102-not 100, nor phenomena of suggestion and ideomotor 101, but 102-hypnosis experiments that had action (Braffman and Kirsch, 1999; Kirsch not been done, and which were, in his view, and Braffman, 2001). As another contrast with well worth doing (Hull, 19930a,b) and some Spanos's position, Kirsch and Lynn, while assurn- 40 studies of waking suggestibility as well ing an appropriate stance of scientific skepti- (Hull, 1929). Most of these experiments cism concerning various claims about hypnosis, remain undone, and most of these remain well apparently feel no need to engage in a program worth doing. And how much more remains of debunking. Hypnotic effects can be accepted to be done, given that we know so much as reflecting genuine subjective experiences, more about both mental processes and social even if these effects are to be attributed to sug- interactions! gestion, not to hypnosisper se (Raz et al., 2006). For most of its recent history, the social- Still, the social-cognitive approach to hypno- psychological approach to hypnosis has defined sis shares some undesirable features with con- itself in opposition to those approaches that temporary work in social cognition, and indeed focused on alterations of consciousness social psychology generally-which is, frankly, occurring in hypnosis (Ludwig and Levine, that it is not very social (Carlson, 1984). For a1 1965; Chaves, 1968; Spanos, 1970, 1986b, the talk about social-psychological approaches 1987a,b; Spanos and Chaves, 1970; Barber, 1972; to hypnosis, relatively little experimental work Sarbin and Coe, 1972; E. R. Hilgard, 1973a,b, has been devoted to core topics in social psy- 1992; Blum, 1978; Kihlstrom, 1992a, 2007; chology. What is the relationship between atti- Sarbin, 1992; Kirsch and Lynn, 1995, 1998a,b,c, tudes towards hypnosis and hypnotic behavior Kirsch and Lynn, 1998~;Oakley, 1999a,b; Kallio (McConkey, 1986; Capafons et al., 2004)?What and Revensuo, 2000, 2003). The result has actually transpires between the hypnotist and been to give the literature on hypnosis some of the subject (Sheehan, 1971a; McConkey and the features of a zero-sum game, in which Sheehan, 1976, 1980, 1982; Baker and Levitt, evidence for the involvement of some social- 1989)? How do hypnosis and other forms of psychological process, such as variations in the suggestibility relate to susceptibility to other wording of suggestions, is taken as evidence forms of social influence (Moore, 1964; Orne against the involvement of some cognitive and Evans, 1965; Evans, 1967)? How do individ- process, such as divided consciousness (e.g. ual subjects influence each other in group hyp- Spanos and Hewitt, 1980; Laurence et al., 1983; nosis, such as the HGSHS:A? (Evans and Spanos, 1983; Green et al., 2005b,c; Kihlstrom , Mitchell, 1986)? What cognates of hypnosis can and Barnier, 2005). But it does not necessarily be found in non-Western cultures (Mischel and follow that, because interpersonal processes Mischel, 1958; Kirmayer, 1992)?Consider hyp- shape hypnosis, hypnosis cannot also involve an nosis as something that happens between two alteration in consciousness. people, and then scan the table of contents of There is a third way, and it has been available any introductory survey of social psychology: to us from the beginning. Wiiam James, con- we have only scratched the surface of the sidering the competing claims of the Salpetriere domain of hypnosis. and Nancy schools concerning the nature of hypnosis, concluded that 'The suggestion-theory may therefore be approved as correct, provided we 2.14. Two ways In hypnosis grant the trance-state as its prerequisite (James, 189011980, p. 1201, italics original). Fifty years and a thlrd way later, at the dawn of the modern era of hypnosis Actually, this is true for both aspects of hypno- research, R. W. White asserted that 'The theory sis-the alterations of consciousness and the of hypnotism will never prosper until, outgrow- social interactions. As hypnosis enters the ing the dialectic dichotomy of "striving" and References . 41

"state", it considers the possibility of interaction' Acknowledgments (White, 1941, p. 502). Martin T. Orne (1959), White's protege as both an undergraduate and The point of view represented in this chapter is a graduate student at Haward, famously tried based on research supported by Grant MH- to distinguish between artifact and essence of 35856 from the National Institute of Mental hypnosis (Orne, 1959), but a careful reading Health. I thank Patricia A. Register, Leanne of his work makes it clear that the demand Wilson, Paula Niedenthal, Ernest Mross, Jeanne characteristics that surround hypnosis are as Sumi Albright, Martha Glisky and- Susan important as any 'trance logic' that might arise McGovern for their contributionsto that program within hypnosis. of research. I also thank Amanda Barnier for her R. E. Shor, writing in the first edition of this thoughtful and thorough editorial comments. volume, noted that 'The fundamental problem in hypnosis research is that it is faced with two References dangers, which, like the rock and whirlpool of Scylla and Charybdis, are so situated that they Alarcon,A., Capafons,A., Bayot, A. and Cardena, E. (1999) must be encountered together, as if they are one' Preference between two methods of active-alert hypnosis: not all techniques are created equal. American (Shor, 1972, p. 15). Shor thought that the prob- Journal of Clinical Hypnosis, 41: 26%276. lem was that of simultaneously 'maintaining Allport, G. W. (1937) The functional autonomy of motives. both the disciplined skepticism of the scientist American Journal of Psychology, 50: 141-156. and the confident persuasiveness of the hypno- Allport, G. W. (1954) The historical background of social tist' (p. 15). Today, we can rephrase the problem psychology. In G. Lindzey and E. Aronson (ed.) Handbook of SocialPsychology,Vol. 1, pp. 146. Random as follows: that of simultaneously maintaining House, New York. an interest in the cognitive processes by which Arnold, M. B. (1946) On the mechanism of suggestion and consciousness is divided in hypnosis, and an hypnosis. Journal of Abnormal and Social Psychology, 41: interest in the social context in which hypnosis 107-128. takes place. Tracing the in As, A. (1962) Non-hypnotic experiences related to hypnotizability in male and female college students. four stages from Mesmer to Hull, Shor asked: Scandinmian Journal of Psychology, 3: 112-121. 'How well have modern investigators learned to As, A. (1963) Hypnotizability as a function of nonhypnotic sail between Scylla and Charybdis? To what experiences. Journal ofAbnormal and Social Psychology, extent will modern viewpoints be seen through 66: 142-150. time as true advances-perhaps to a fifth stage As, A., O'Hara, J. W. and Munger, M. P. (1962) The of sophistication-and to what extent merely as measurement of subjective experiences presumably related to hypnotic susceptibility. Scandinavian Journal changes to culturally more acceptable mis- of Psychology, 3: 47-64. nomers and disguised returns to old mistakes?' Aserinsky, E. and Kleitman, N. (1953) Regularly occurring (1972, p. 40). periods of eye motility, and concomitant phenomena, Shor did not know the answer then, and the during sleep. Science, 118: 273-274. answer is not clear even now. But it is clear what Bailly, J. S. (178412002) Secret report on mesmerism, or animal magnetism. International Journal of Clinical and we should do, which is abandon the stance of ExperimentalHypnosis, 50: 364-368. either-or and adopt a new stance of both-and. Baker, E. L. and Levitt, E. E. (1989) The hypnotic This 'third way' in hypnosis research construes relationship: an investigation of compliance and hypnosis simultaneously as both a state of resistance. International Journal of Clinical and (sometimes) profound cognitive change, involv- Experimental Hypnosis, 37: 145-153. Balthazard, C. G. and Woody, E. Z. (1985) The 'stuff'of ing basic mechanisms of cognition and con- hypnotic performance: a review of psychometric sciousness, and as a social interaction, in which approaches. Psychological Bulletin, 98: 28S296. hypnotist and subject come together for a spe- Balthazard, C. G. and Woody, E. Z. (1989) Bimodalitv, cific purpose within a wider socio-cultural con- dimensionality, and the notion of hypnotic types. text. To get beyond the misnomers and mistakes International Journal ofClinical and Experimental Hypnosis, 37: 70-89. of the past, hypnosis researchers must have Banyai, E. (1991) Toward a social-psychobiologicalmodel a vision as large as the phenomenon they seek of hypnosis. In S. J. Lynn and J. W. Rhue (ed.) Theories of to study. And the domain of hypnosis is very Hypnosis: Current Models and Perspectives, pp. 564-598. large indeed. Guilford Press, New York. 42 CHAPTER 2 The domain of hypnosis, revisited

Banyai, E. I. (1998) The interactive nature of hypnosis: Bernheim, H. (188611889). Suggestive Therapeutics research evidence for a social-psychobiologicalmodel. A Treatise on the Nature and Uses of Hypnotism. Contemporary Hypnosis, 15: 52-63. G. P. Putnam's Sons, New York Banyai, E. I. and Hilgard, E. R (1976) A comparison of Binet, A. and Fere, C. (1888) Animal Magnetism. Appleton- active-alert with traditional Century, New York relaxation induction. Journal ofAbnonna1 Psychology, Blum, G. S. (1967) Hypnosis in psychodynamic research. 85: 218-224. In J. E. Gordon (4.) Handbook of Clinicnl and Barber, T. X. (1964) 'Hypnosis' as a causal variable in Experimental Hypnosis, pp. 83-109. Macmillan, present-day psychology: a critical analysis. Psychological New York Reports, 14: 839-842. Blum, G. S. (1978) A conceptual model for hypnotic Barber, T. X (1965) Measuring 'hypnotic-like' suggestibility alterations of consciousness.Journal of Altered States of with and without 'hypnotic induction': psychometric Consciousness,4: 189-201. properties, norms, and variables influencing response to Blum, G. S. (1979) Hypnotic programming techniques in the Barber Suggestibility Scale (BSS). Psychological psychological experiments. In E. Fromand R. E. Shor Reports, 16 (Monograph Supplement 3): 809-844. (ed.) Hypnosis: Developments in Research and New Barber, T. X. (1969) Hypnosis: A Scientific Approach. Van Perspectives, pp. 457-481. Aldine, New York Nostrand Reinhold, New York Blum, G. S. and Porter, M. L. (1972) The capacity for rapid Barber, T. X (1972) Suggested ('hypnotic') behavior: the shifts in level of mental concentration. Quarterly trance paradigm versus an alternative paradigm. In Journal of Experimental Psychology, 24: 431-438. E. From and R. E. Shor (ed.) Hypnosis: Recent Boring, E. G. (1953) A history of introspectionism. Developments and Perspectives,pp. 115-182. Aldine- Psychological Bulletin, 50: 169-189. Atherton, Chicago. Bower, G. H. (1981) Mood and memory. American Barber, T. X (1999) A comprehensive three-dimensional Psychologist, 36: 129-148. theory of hypnosis. In K. Kirsch, A. Capafons, Bower, G. H. and Forgas, J. P. (2000) Affect, memory, E. Cardena and S. Amigo (ed.) Clinical Hypnosis and and social cognition. In E. Eich, J. F. Kihlstrom, Self-regulation: Cognitive-behavioral Perspech'ves, G. H. Bower, J. P. Forgas and P. M. Niedenthl (4.) pp. 21-48. American Psychological Association, Cognition and Emotion, pp. 87-168. Oxford Washington, DC. University Press, New York Barber, T. X. and Calverley, D. S. (1964) An experimental Bower, G. H., Monteiro, K. P. and Gagan, S. G. (1978) study of 'hypnotic' (auditory and visual) hallucinations. Emotional mood as a context for learning and recall. Journal of Abnormal and Social Psychology, 68: 13-20. Journal of Verbal Learning and Verbal Behavior, Barber, T. X., Spanos, N. P. and Chaves, J. F. (1974) 17: 573-585. Hypnosis, Imagination, and Human Potentialities. Bower, G. H., Gagan, S. G. and Monteiro, K. P. (1981) Pergamon, New York Selectivityof learning caused by affective states. Barnier, A. J. and McConkey, K. M. (2004) Defining and Journal of Experimental Psychology: General, 110: identifjmg the highly hypnotizable person. In M. Heap, 45 1-473. R. J. Brown and D. A. Oakley (ed.) The Highly Bowers, K. S. (1967) The effect for demands of honesty Hypnotizable Person: Theoretical, Experimental and upon reports of visual and auditory hallucinations. Clinical Issues, pp. 30-60. Routledge, London. International Journal of Clinical and Experimental Bates, B. L. (1992) The effect of demands for honesty on Hypnosis, 15: 31-36. the efficacy of the Carleton Skills Training Program. Bowers, K. S. (1981) Do the Stanford Scales tap the classic International Journal of Clinical and Experimental suggestion effect? International Journal of Clinical and Hypnosis, 40: 88-102. Expm'mental Hypnosis, 29: 42-53. Bates, B. L. and Brigham, T. A. (1990) Mowng Bowers, K. S. (1992) Imagination and dissociation in hypnotizability with the Carleton Skills Training hypnotic responding. Internatiunal Journal of Clinical Program: a partial replication and analysis of and Experimental Hypnosis, 40: 253-275. components. International Journal of Clinical and Bowers, K. S. (1993) The Waterlo&tanford Group C Experimental Hypnosis, 38: 183-195. (WSGC) Scale of Hypnotic Susceptibility: Bates, B. L. and Kraft, P.M. (1991) The nature of hypnotic normative and comparative data. International performance following administration of the Carleton Journal of Clinical and Experimental Hypnosis, SkiUs Training Program. International Journal of Clinical 41: 3546. and Experimental Hypnosis, 39: 227-242. Bowers, K. S. and Davidson, T. M. (1991) A neodissociative Bates, B. L., Miller, R J., Cross, H. J. and Brigham, T. A. critique of Spanos's social-psychological model of (1988) Modifjmg hypnotic suggestibilitywith the hypnosis. In S. J. Lynn and J. W. Rhue (ed.) Theories of Carleton Skills Training Program. Journal of Personality Hypnosis: Current Models and Perspectives,pp. 105-143. and Social Psychology, 55: 12lL127. Guilford Press, New York. Benham, G., Woody, E. Z., Whon, K. S. and Nash, M. R. Bowers, K. S. and Giore,J. B. (1969) Subjective report (2006) Expect the unexpected: ability, attitude, and and credibility: an inquiry involving hypnotic

responsivenessto hypnosis.. - Journal of Personality and hallucinations. Journal ofAbnormal Psychology, Social Psychology, 91: 342-350. 74: 443-451. References . 43

Bowers, P. (1982) The classic suggestion effect: Council, 1. R, Kirsch, I. and Hafner, L. P. (1986) Expectancy relationshipswith scales of hypnotizability, effortless versus absorption in the prediction of hypnotic experiencing, and imagery vividness. International responding. Journal of Personality and Social Psychology, Journal of Clinical and Experimental Hypnosis, 50: 182-189. 30: 270-279. Crawford, H. J. and Gulier, J. H. (1992) A midstream Bowers, P., Laurence, J.-R. and Hart, D. (1988) The view of the neuropsychophysiology of hypnosis: recent experience of hypnotic suggestions. International reserach and future directions. In E. Fromm and Journal of Clinical and Experimental Hypnosis, 36: M. R. Nash (ed.) Contemporary Hypnosis Research, 336-349. pp. 227-266. Guilford Press, New York. Braffman, W. and Kirsch, I. (1999) Imaginative Cronbach, L. J. and Meehl, P. E. (1955) Construct validity suggestibilityand hypnotizability: an empirical analysis. in psychological tests. Psychological Bulletin, 52: 281-302. Journal of Personality and Social Psychology, 77: 57LL587. Csikszentmihalyi,M. (1990) Flow: The Psychology of Braid, J. (1855) The Physiology of Fascination and the Critics Optimal 13pmenence.Harper and Row, New York. Criticized. Grant, Manchester. Damaser, E. C., Shor. R. E. and Orne. M. T. (1963) Brenneman, H. A. and Kihlstrom, J. F. (2006) Patterns of Physiological effects during hypnotically requested hypnotic abilities. Manuscript in preparation. emotions. Psychosomatic Medicine, 25: 334-343. Bryant, R. A. (2005) Hypnotic emotional numbing: a study Darley, J. M. and Fazio, R. H. (1980) Expectancy of implicit emotion. International Journal of Clinical and confirmation processes arising in the social interaction Expm'mental Hypnosis, 53: 26-36. sequence. American Psychologist, 35: 867-881. Bryant, R. A. and Mallard, D. (2002) Hynotically induced Deleuze, J. P.F, (1813) Histoire Critiquedu Magnetism emotional numbing: a real-simulating analysis. Journal Animal. Mame, Paris. of Abnormal Psychology, 111: 203-207. D'Eon, J. L., Mah, C. D., Pawlak, A. E. and Spanos, N. P. Campbell, D. T. and Fiske, D. W. (1959) Convergent and (1979) Effects of hypnotists' and subjects' sex on discriminant validation by the multitrait-multimethod hypnotic susceptibility.Perceptual and Motor Skills, 48: matrix. Psychological Bulletin, 56: 82-105. 1232-1234. Capafons,A., Cabanas, S., Espejo, B. and Cardena, E. Diamond, M. J. (1974) Modification of hypnotizability: a (2004) Confirmatoryfactor analysis of the Valencia scale review. Psychologicnl Bulletin, 81: 180-198. on attitudes and beliefs toward hypnosis: an Diamond, M. J. (1977a) Hypnotizability is modifiable: an international study. International Journal of Clinical and alternative approach. International Journal of Clinical Experimental Hypnosis, 52: 413-433. and Experimental Hypnosis, 25: 147-166. Carlson, R. (1984) What's social about social psychology? Diamond, M. J. (19776) Issues and methods for moddymg Where's the person in personality research? Journal of responsivity to hypnosis. Annals of the New York Personality and Social Psychology, 47: 1304-1309. Academy of Sciences, 296: 11S128. Chaves, J. F. (1968) Hypnosis reconceptualized: an Diamond, M. J. (1982) Modifying hypnotic experience by overview of Barber's theoretical and empirical work. means of indirect hypnosis and hypnotic skill training: Psychological Reports, 22: 587608. an update (1981). Research Communications in Coe, W. C. (1976) Effects of hypnotist susceptibility and Psychology, 7:233-239. sex on the administration of standard hypnotic Diamond, M. J. (1984) It takes two to tango: some susceptibility scales. International Journal of Clinical and thoughts on the neglected importance of the hypnotist Experimental Hypnosis, 24: 281-286. in an interactive hypnotherapeutic relationship. Coe, W. C. and Sarbin, T. R. (1971) An alternative American Journal of Clinical Hypnosis, 27: 3-13. interpretation to the multiple composition of hypnotic Dixon, M. and Laurence, J.-R. (1992) Hypnotic scales: a single mle-relevant skill. Journal of Personality susce~tibity-.and verbal automaticity: automatic and and Social Psychology, 18: 1-8. strategic processing differences in the Stroop color- Coe, W. C. and Sarbin, T. R. (1991) Role theory: hypnosis naming task. Journal of Abnormal Psychology, from a dramaturgical and narrational perspective. In 101: 344-347. S. J. Lynn and J. W. Rhue (ed.) Theories of Hypnosis: Dixon, M., Brunet,A. and Laurence, J.-R. (1990) Current Models and Perspectives, pp. 303-323. Guilford Hypnotizability and automaticity: toward a parallel Press, New York. distributed processing model of hypnotic responding. Coltheart, M. (2006a) Perhaps cognitive neuroimaging has Journal of Abnormal Psychology, 99: 336-343. - not told us anything about the mind (so far). Cortex, 42: Doris, J. (ed.). ( 1991) The Suggestibility of Children's 422-427. Recollections.American Psychological Association, Coltheart, M. (2006b) What has functional neuroimaging Washington, DC. told us about the mind (so far)? Cortex, 42: 323-331. Dumas, L. (1964) A subjective report of inadvertent Cooper, L. M. and London, P. (1971) The development of hypnosis. International Journal of Clinical and hypnotic susceptibility: a longitudinal (convergence) Experimental Hypnosis, 12: 78-80. study. Child Development, 42: 487-503. Dumas, R. A. (1977) EEG alpha-hypnotizability Council on ScientificAffairs (1985) Scientific status of correlations: a review. Psychophysiology, 14: 431-438. refreshing recollection by the use of hypnosis. Journal of Edmonston,W. E., rr (1977) Neutral hypnosis as relaxation. the American Medical Association, 253,1918-1923. American Journal of Clinicnl Hypnosis, 20: 6%75.