Steven Jay Lynn, Irving Kirsch and Michael N. Hallquist
5.3. Genesis of our own point of view
5.3.3. Spanos’s multifactorial model
Nicholas P. Spanos (1986) extended Sarbin’s role theory and Barber’s cognitive-behavioral theory into one of the most influential contemporary theoretical approaches to the understanding of hypnotic behavior. Spanos and his colleagues (Spanos, 1986, 1991; Spanos and Chaves, 1989) focused on the importance of social psychologi-cal processes and the importance of goal-directed activities and strategic responding. Like his mentor T. X. Barber, Spanos advanced a mul-tifactorial model of hypnotic suggestibility that acknowledged the role of attitudes, beliefs, imag-inings, attributions and expectancies in shaping hypnotic phenomena. Extending Sarbin’s role theory, and White’s observations about the goal-directed nature of hypnosis, Spanos (1991) used the construct of strategic role enactment to explain how individuals transform imaginings, thoughts and feelings into experiences and behaviors that are consistent with their ideas of how a good hypnotic subject should respond to the overall hypnotic context and specific suggestions in particular. How subjects construe the hypnotic role is thus a key determinant of hypnotic responsiveness.
Spanos chose the term ‘enactment’ to distin-guish individuals’ attempts to fulfill the perceived requirements of the hypnotic role from pretence or faking. The term ‘strategic’ captures the consis-tent emphasis in Spanos’s research program on identifying cognitive strategies that individuals employ to create suggested effects. For example, in order to reduce pain in response to an analge-sia suggestion, subjects may use a variety of atten-tional strategies (e.g. distraction) to attenuate their subjective responses to painful stimuli.
Relatedly, suggestions often contain strategies that assist individuals in responding appropri-ately (Barber et al., 1974; Spanos and Barber, 1974; Spanos et al., 1985; see also Wagstaff, 1991, 1998). Raising the hand following sugges-tions for the hand to lift occurs as the subject imagines along with suggestions worded to imply that the hand will lift involuntarily (e.g.
‘Your hand is getting lighter and lighter, it will rise by itself ’). Spanos (1971) hypothesized that subjects tend to define their overt response to suggestion as involuntary when they become absorbed in goal-directed fantasies (GDFs).
GDFs are defined as ‘imagined situations which, if they were to occur, would be expected to lead to the involuntary occurrence of the motor response called for by the suggestion’ (Spanos et al., 1977, p. 211). For instance, persons administered a hand levitation suggestion would exhibit a GDFr (i.e. goal-directed fantasy report) if they report such events as imagining a helium balloon lifting their hand, or a basketball being inflated under their hand.
Studies have indicated that GDFrs are related to the tendency to define overt response to sug-gestion as involuntary occurrences. However, GDFrs are not necessarily correlated with the number of suggestions that subjects pass on a suggestibility scale (for a review, see Lynn and Sivec, 1992). In fact, several studies indicate that such fantasies may even diminish responding to suggestions (Hargadon et al., 1995; Comey and Kirsch, 1999). Why is this the case? Spanos argued that certain individuals can be fully absorbed in GDFrs yet passively wait for a sug-gested event, such as the lifting of an arm in response to a hand levitation suggestion, to
‘happen.’ Adopting this response set virtually guarantees failure. In contrast, individuals who understand that it is important to lift their arm will succeed in passing the suggestion. In short, how suggestions are interpreted is an influential determinant of whether or not they are accompa-nied by behavioral responses to suggested events.
One of Spanos’s most important contribu-tions has been to highlight the fact that hypnotic responsiveness is malleable and can be substan-tially modified and enhanced. According to Spanos, hypnotic responsiveness is not an immutable trait or propensity locked in at birth.
Rather, it can be substantially modified by teaching subjects to do the following: (1) physi-cally enact responses, as opposed to waiting pas-sively for the suggested effects to happen to them (Spanos, 1986); and (2) use goal-directed imagery or fantasies
In more than 15 studies (see Gorassini and Spanos, 1999), Spanos showed that between 50 and 80 percent of initially low hypno-tizable subjects who underwent a multifac-eted cognitive skill training program (termed the Carleton Skills Training Program; CSTP) scored as high hypnotizables at post-testing.
The training program includes the following
components: (1) information to instil positive motivation, beliefs, attitudes and expectations about responding to hypnotic suggestions; (2) instructions regarding how to use imaginal strategies to promote successful responding; (3) instructions regarding how to interpret sugges-tions (e.g. one must actually lift the hand while imagining that it is rising by itself); (4) exposure to a videotaped model who enacts successful responses to suggestions and verbalizes imagery-based strategies to facilitate subjective response;
and (5) practice in responding to test sugges-tions. Spanos (March 6, 1985, personal commu-nication) also noted that the training program encourages rapport with the trainer/hypnotist.
Not only have treatment effects been of large magnitude, but they also persist for an average of two and a half years after training (Spanos et al., 1988a), and generalize to novel, demand-ing test suggestions (e.g. Spanos, 1986). Research in Spain (Cangas and Perez, 1998), Poland (Niedzwienska, 2000) and the UK (Fellows and Ragg, 1992) provided cross-cultural support for the effectiveness of the CSTP. Finally, treatment-related gains are maintained even when tested in an entirely novel test context in which subjects fail to connect their earlier training and subse-quent hypnotizability testing. Research that addressed this latter issue (Zivney and Lynn, 1996) provided strong evidence that large mag-nitude treatment gains cannot be attributable to simple compliance effects. Combined, these findings constitute strong support for Spanos’s model and challenge the received wisdom that hypnotic responsivity is a trait-like, stable attrib-ute that can be modified only within narrow limits (for more discussion of the CTSP and modifiability of hypnotizability, see Laurence et al., Chapter 9, this volume).
Spanos and his colleagues devoted substantial effort to testing hypotheses derived from Hilgard’s neodissociation theory (Hilgard, 1977). The hidden observer phenomenon is integral to Hilgard’s (1977) neodissociation the-ory, according to which all hypnotic experiences and responses are instances of dissociation. That is, they reflect a division of executive cognitive functioning into two parts, divided by an amnesic barrier. One part of consciousness (the hidden part) directs the person’s behavior and experiences stimuli normally, but the other is Genesis of our own point of view · 119
unaware of self-agency and experiences stimuli in a manner that is consistent with suggestions.
Hilgard (1973) serendipitously discovered the hidden observer phenomenon in a class demon-stration of hypnotic deafness with a blind per-son. Hilgard’s curiosity was aroused regarding whether a person who received a suggestion for complete deafness might be still aware of what was going on in his surroundings, despite his lack of startle response to loud noises. He told the person that, ‘Although you are hypnotically deaf, perhaps there is some part of you that is hearing my voice and processing the informa-tion. If there is I should like the index finger to rise as a sign that this is the case’ (Hilgard, 1994, p. 34). The person’s finger rose as requested, despite his denial of hearing anything. Post-hypnotic inquiry revealed that the ‘separate part’ had been aware all along of everything that had been taking place.
This demonstration inspired many studies of the hidden observer. Hidden observers were obtained in association with hypnotic sugges-tions for analgesia to experimental pain, deaf-ness, negative hallucinations (i.e. not seeing something that is there) and ideomotor responses (reviewed in Kirsch and Lynn, 1998;
also see Green et al., 2006). The basic paradigm of these studies is to select individuals who have demonstrated high levels of hypnotic sug-gestibility, tell them that there is (or may be) a hidden part of them that is more aware than their hypnotized part, establish a cue for com-municating with the hypnotized part, administer a suggestion, and assess its effects with and with-out the hidden observer cue. Typically, hidden observers report higher levels of pain, less deaf-ness and less blinddeaf-ness when the hidden observer cue is given. These hidden observer reports are interpreted as responses from the
‘normal (i.e. not hypnotized) part’ of the execu-tive ego, the part that is fully experiencing stim-uli and remembering information that is temporarily unavailable to the ‘hypnotized part’
of consciousness.
For the hidden observer to constitute empiri-cal support for neodissociation theory one must assume that it reveals a hidden part of con-sciousness that is present during hypnosis, independently of the subject’s expectations and the instructions used to communicate with it.
However, Spanos and his colleagues reported data in which hidden observer reports were affected by the nature of instructions given to subjects, leading these authors to the conclusion that the hidden observer is an experimental creation.
In the first of these studies, Spanos and Hewitt (1980) provided two groups of highly suggestible subjects with very different information about the hidden observer. One group was provided with information modeled on Hilgard et al. (1975) that conveyed the expectation that the hidden part would give accurate pain reports. Other subjects were informed that the hidden part was ‘so deeply hidden’ that it would experience even less pain than the hypnotized part. These two differ-ent sets of instructions elicited hidden observers with opposite characteristics (i.e. high sensitivity versus greatly reduced sensitivity to pain). Spanos and his colleagues (Spanos et al., 1983) later repli-cated this effect using a within-subject manipula-tion. As instructions shifted to convey different expectations regarding the characteristics of the hidden observer, the same subjects reported either less hidden pain than overt pain or more hidden pain than overt pain.
Spanos et al. (1984a) produced two hidden observers in each of eight, highly suggestible, research subjects. The subjects learned a list of concrete and abstract words and then were given a suggestion to forget them. Half of the subjects were told that abstract words were stored by their right hemisphere and that con-crete words were stored by the left. The remain-ing subjects received the opposite information concerning where information was stored. They were also told that there was a hidden part of them connected to each cerebral hemisphere.
When the hypnotist contacted the ‘hidden observer’ associated with the right and left hemispheres, the subjects were able to recall the information that they had been told was stored there. That is, when the hypnotist contacted the right hemisphere ‘hidden part’, the subject recalled all of the words purportedly stored in the right hemisphere (e.g. concrete words), but none of the words stored in the left hemisphere (e.g. abstract words), and vice versa.
Spanos et al. (1988b) demonstrated a simi-larly flexible hidden observer in conjunction with suggested negative hallucinations. Previously,
Zamansky and Bartis (1985) had reported a study in which they had administered negative hallucination suggestions for subjects to see a blank page when the page actually had a clearly visible number imprinted on it. The authors found that all of the subjects who had initially denied seeing the number on the page claimed they had seen the number when the experimenter contacted their ‘hidden observer’.
Spanos et al. (1988) replicated this effect, using the number 18 as the stimulus that was not to be seen. However, half of the subjects were told that the hidden part reversed everything it saw. The results were clear-cut. Subjects’ hidden observers reported what they had been told about the hidden observer: those who received typical hidden observer suggestions reported that the number was 18, whereas those told that the hidden observer reversed what it saw reported that they saw the number 81.
Most recently, Green and his colleagues (Green et al., 2006) tested high and low sug-gestible role-playing (simulating) subjects for the amount of effort they experienced in responding to an arm suspension (i.e. ideomo-tor) task across baseline, hypnosis, hidden observer and post-session trials. Subjects were told that their hidden observer would be more aware of the effort required to complete the task; less aware; or they received no specific instruction concerning how their hidden observer should respond. The authors found that both real and simulating subjects’ hidden observer reports of effort were influenced by the wording of the instructions they received.
Support for a social cognitive account of the hidden observer also stems from findings that the frequency of hidden or covert reports varies with the explicitness of the suggestions used to elicit the phenomenon. Studies in which highly suggestible subjects receive very explicit suggestions to report a hidden observer, or are given practice at performing the ‘hidden tasks’, have produced hidden observer response rates ranging from 82 to 94 percent (Knox et al., 1974; Spanos and Hewitt, 1980; Mare et al., 1994). This high rate of response contrasts sharply with hidden observer response rates ranging from 25 to 42 percent in studies of hyp-notic analgesia and deafness in which less explicit instructions were used (Hilgard et al.,
1975; Crawford et al., 1979; Laurence and Perry, 1981).
Spanos et al. (1983) tested the hypothesis that hidden observer response rates depend on the explicitness of the instructions by varying the saliency of the behavioral cues within the hidden observer suggestion. In research on overt and covert reports of pain, they found that only 14 percent of high hypnotizable subjects displayed hidden observers (i.e. generated dis-crepant overt and covert reports) under the low cue condition. In contrast, the high cue condition resulted in hidden observers in 58 per-cent of subjects. This suggests that the hidden observer may be a product of perceived situa-tional cues.
Hidden observer studies indicate that hidden observer reports are exquisitely sensitive to situ-ational demands produced by instructional cues. Depending on the instructions with which they are created, hidden observers report more or less pain than otherwise reported, they either do or do not reverse figures, and they report more or less effort associated with an ideomotor response. Accordingly, reports of ‘hidden enti-ties’ appear to be byproducts of suggestions and contextual demands, rather than spontaneously occurring dissociated aspects of the personality (see Laurence et al., 1983; Nogrady et al., 1983;
Kihlstrom, 1998, 2003, for a defense of the ‘hid-den observer’, and a rebuttal by Spanos, 1991).
Such findings led Kirsch and Lynn (1998) to dub the hidden observer a ‘flexible observer’. Instead of being the basis for all suggested responses, the hidden observer appears to be a suggested response that can be considered part of the domain of hypnosis (for a different view of the implications of the hidden observer, see Woody and Sadler, Chapter 4, this volume).
5.3.4.