Instrumental Hypnosis: Putting Hypnosis to Work

Paper presented at the annual meeting of the Society for Clinical and Experimental Hypnosis, Berkeley, California, October 5, 2013. 

Thanks to the presenters for three excellent papers, which show clearly that the center for contemporary hypnosis research is now firmly located in Australia -- and in the best-looking building on the Macquarie University campus!

Long ago, Joseph Reyher (1961) offered a tripartite classification of hypnosis research:

• There is intrinsic research, directed toward understanding the nature of hypnosis itself.
• And applied research,directed toward the utilization of hypnosis in medicine and psychotherapy.
• And there is instrumental hypnosis, in which hypnosis is used as a technique to learn about other aspects of mind and behavior.

Reyher himself promoted instrumental hypnosis, which he used to study various processes discussed in psychoanalytic theory, such as repression.  But for most of its history, ever since the work of P.c. Young and C.L. Hull, scientific investigation of hypnosis has focused on intrinsic and applied research.

This symposium marks an important turning point in what Jack Hilgard called the "domestication" of hypnosis.  In a sense, since the 1920s and 30s, and into the "Golden Age" that began in the late 1950s, hypnosis has been proving itself -- that it's not a sham, that it really works, that we're beginning to understand it.  Instrumental research with hypnosis has been held back by the disconcerting feeling that it would amount to studying one unknown with another.

Apparently, that's no longer the case.  Almost 15 years ago now, for example, Pierre Rainville (1999) used hypnosis to separately manipulate the two components of pain, sensory pain and suffering, to study their differential neural correlates.  He didn't bother asking whether hypnosis was a state, or whether hypnotic subjects really felt no pain.  He just used hypnosis to study something else he was interested in.

Prof. Barnier's laboratory has been a leader in this area, pursuing several different lines of research.  Their work has been published in leading journals.  This testifies to the competence of the group in designing and executing their studies.  But it also testifies to the wider acceptance of hypnosis as an instrumental method.

The work on delusions described by Prof. Barnier is a beautiful example ("Toward a Good Likeness: Hypnotic Models of Delusions").  Delusions are one of the most dramatic symptoms of mental illness, and one of the Schneiderian "first rank" symptoms of schizophrenia.  But there's a real puzzle about where they come from.  They're not explained, for example, by the popular dopamine hypothesis of schizophrenia, because delusions have what the philosophers call intentionality, or what ordinary people call content, whereas neurotransmitters don't.  Delusions are beliefs about something, whereas neurotransmitters aren't about anything -- they just are. 

the best available theory of delusions, originally proposed by Karl Jaspers, and brought into psychology by Brendan Maher and others, is that they reflect the patient's attempt to explain his other anomalous experiences -- his other Schneiderian first-rank symptoms.  "I hear voices speaking to me, so I must be receiving broadcasts from Mars through my tooth fillings" (e.g., Kihlstrom & Hoyt, 1988). 

Max Coltheart has expanded on this idea with a two-factor theory:

1. Anomalous experiences seed the formation of delusions.
2. These aberrant beliefs persist in the face of logic and evidence.

Like Rainville, Prof. Barnier and her group use hypnosis to manipulate these two factors independently, proceeding one after another through a list of the most common delusions. 

Her example of mirror self-identification is of particular interest, on two counts.

1. The delusions itself strikes me as the reverse of the double hallucination popularized by Martin Orne.  In the double hallucination, the subject is asked to hallucinate a familiar person, and then is confronted with the real person as well.  If you're lucky, you get a classic double-take, as the subject looks back and forth between the hallucinated person and the real one.  In mirror self-misidentification, the subject is again confronted with two representations of the same person, but doesn't recognize them as identical.
2. The results of the experiment forced a revision of the theory.  It's not just the recognition failure that's critical, but also anosognosia.  This is interesting.  Coltheart didn't just dismiss the hypnotic findings as anomalous.  He took them so seriously that he revised his theory on the basis of them.

All of this is in line with the Australian tradition of hypnosis research, beginning with J.P. Sutcliffe's idea that the hypnotic subject is essentially deluded about reality -- he possesses an incorrect belief about himself or the world, firmly held despite evidence to the contrary.

As I said, Prof. Barnier's group has been studying all sorts of different delusions.  Dr. Cox has focused on delusions of alien control ("Disrupting Behavior and Self-Monitoring in a Hypnotic Analogue of Alien Control").  Here in America, we had a dramatic example of this delusion in the recent shootings at the Washington Naval Yard, where the perpetrator -- let's not call him just a "suspect" believed that he was being influenced by extra-low-frequency (ELF) radio waves.

While the research on mirror self-misidentification manipulated only the first of Coltheart's two factors, Dr. Cox's research attacks both of them:

1. suggesting control by another person
2. suggesting problems in reasoning to evaluate the belief.

She employs a very clever experimental manipulation of Factor 1.  Previous investigators used a very complex apparatus which I suspect really piles on the demand characteristics.  Dr. Cox's procedure is much more straightforward: a simple suggestion that "someone else" is controlling his arm and hand.

The manipulation of Factor 2, however, is I think a little more problematic: a simple suggestion that "You'll think that any explanation of [of the first effect] is plausible". This begs the question of why subjects should find the explanation plausible.  Elsewhere (Kihlstrom & Hoyt, 1988) I've suggested that the persistence of delusions might be explained by exaggerations of ordinary errors in reasoning -- in particular, an excessive reliance on the kinds of judgment heuristics described by Daniel Kahneman and Amos Tversky: representativeness, availability, simulation, and anchoring and adjustment.  So, for example, given the anomalous experience of the perceived loss of control, and the hypnotic context, it might occur to the subject that someone else might have control of his arm and hand (via availability or simulation); and then, when they think about it some more, they don't adjust their initial attribution in light of later evidence (via anchoring and adjustment).  If I'm right, an alternative tack would be to make suggestions that will directly exacerbate these sorts of errors.  Given the highly programmatic nature of this research, this might be a new line worth pursuing.

Dr. Polito's research -- and here we should give him a round of applause, for he just learned that he's passed his exams, and he's now officially  Dr. Polito -- addresses another theoretical question -- the nature of automatic, unconscious processing ("Using Hypnotic Models to Influence 'Automatic' Responses"). 

Since the 1970s, cognitive psychologists have distinguished between two sorts of processes:

1. Controlled processes, which are deliberate and voluntary, and consume attentional resources.
2. Automatic processes, which are executed in a reflex-like fashion, without consuming any attentional resources.

There also appear to be two sources of automaticity:

1. Some processes are innately automatic;
2. Others are automatized through extensive practice, such as the "10,000 Hour Rule" popularized by Malcolm Gladwell (but which was originally propounded by Anders Ericsson).

Either way, theory generally holds that once a process is automatized it stays that way.  But Amir Raz has reported that he can undo a cardinal example of automaticity, the Stroop interference effect, by means of hypnotic suggestions for agnosia.  If he's right, then it's possible to "unring the bell", and de-automatize an automatic process.  That would be a new empirical fact for automaticity theory to contend with.

Dr. Polito's work has focused on the sense of agency, or the experience of conscious control -- a topic that has garnered quite a bit of attention fro philosophers and other cognitive scientists lately.  Some theorists, like the late Dan Wegner, claim that conscious control is simply an illusion -- we don't have it, because everything we do we do automatically and unconsciously.  Wegner and many other advocates for this position base their arguments on the famous Libet experiment, which purported to show brain signature of an unconscious initiation of a response before the subject is consciously aware of it.  The Libet experiment has gotten a great deal of press, but I am here to tell you that it is entirely unwarranted.  Research by Jeff Miller, a New Zealand psychologist, shows that Libet's results were wholly an artifact of his method.  Maybe we don't have free will, but the Libet experiment doesn't prove it, not by a long shot. 

As is characteristic of the Barnier lab, Dr. Polito's work has employed a number of different paradigms.  One set of experiments employs Wegner's "Clever Hands" paradigm (Wegner loved cute titles, in this case a variant on "Clever Hans") in which subjects are asked to respond randomly to easy and difficult trivia questions.  Wegner found that subjects could respond randomly to difficult questions, but not easy ones, because the answers to easy questions occurred automatically to subjects.  Maybe so, but Dr. Polito is able to get subjects to respond randomly even to easy questions by means of hypnotic suggestion -- which means that he can de-automatize processing in that domain, too.

I'm particularly fond of Dr. Polito's work on self-tickling, which is kind of an oxymoron.  You can make yourself feel very good, but you can't tickle yourself (sort of the way a sandwich always tastes better when it's made by someone else -- but not really).  The traditional explanation invokes a kind of Gibsonian process, in which egomotion feedback from the skeletal musculature corrects tactile sensations arriving at somatosensory cortex, dampening the tickles that the subject would feel otherwise.  Again, most previous attempts to get people to tickle themselves have involved a complex apparatus that are, again, rife with demand characteristics and, perhaps because of that, and don't succeed very well.  Dr. Polito employs a much more direct approach, and voila!  People can tickle themselves after all, provided that they don't know they're doing it.

Even if we didn't have free will -- though of course we do! -- we'd still have this sense of agency, and Dr. Polito has taken his psychometric skills, honed by his hypnosis research, and put them to good use in developing his Sense of Agency Rating Scale (SOARS).  There's been a lot of interest in the experience of involuntariness in response to hypnotic suggestions, leading a number of us to develop scales for capturing this experience in hypnosis.  But Dr. Polito has produced something much more general -- an instrument that can tap the sense of agency in almost any domain of behavior.

Interestingly, Dr. Polito finds that the sense of agency is not unidimensional, but has two underlying factors -- involuntariness and effortlessness, that can be dissociated in hypnosis.  I'm pretty convinced, but I worry that you get out of factor analysis pretty much what you put into it.  Dr. Polito culled his scale items from philosophical discussions of the sense of agency, and this vocabulary might have been unduly influenced by the intuitive theories of the philosophers themselves.  He might get very different items, and very different factors, by taking a leaf from the work of Bertram Malle, on attribution theory.  Social psychologists have long embraced a distinction between personal and situational attributions, based largely on a misreading of Fritz Heider's work (or so Malle argues), leading to the "discovery" of such things as the "Fundamental Attribution Error" and the "Actor-Observer Difference" in causal attributions.  But when Malle performed a content analysis of ordinary people's causal attributions, he found that these "basic" attributional phenomena took on quite a different shape.  Maybe if we looked at the sense of agency as experienced by naive observers, without much by way of philosophical training, we'd get a quite different structure.  It's a hypothesis worth entertaining, anyway. 

By suggesting that hypnosis might serve as a neural model, by creating "virtual lesions" that impair the sense of agency, he takes us back to the very beginnings of hypnosis as a laboratory model.  These were at the Salpetriere, where Charcot, who after all was a neurologist by training, noticed the parallels between the symptoms of hysteria and those of certain neurological conditions, and between the symptoms of hysteria and the phenomena of hypnosis, and suggested that both hysteria and hypnosis involved "functional" rather than "organic" lesions affecting brain function.

Given the high level of theoretical interest in automaticity and conscious control, I suspect that one or another version of the SOARS will be picked up by researchers working in a wide variety of domains, not just hypnosis.  This is just another example of how hypnosis can contribute to other kinds of research, and offers another perspective on the instrumental use of hypnosis.

I thank the presenters for a collection of vary stimulating papers, and I thank the audience for your attention to them.