Hypnosis and Automatic Behaviours

Vince Polito is is Postdoctoral Research Fellow at Macquarie University, and member of the Belief Formation Program at the ARC Centre of Excellence in Cognition and its Disorders. His work investigates alterations of agency and body representation associated with hypnosis, virtual reality, flow, meditation, and psychoactive drugs. You can find him on twitter here.

Hypnosis is used clinically as a treatment for conditions such as chronic pain, and is also becoming more commonly used as a research tool in cognitive science. Despite growing levels of interest in hypnosis, the mechanisms that underlie hypnotic effects are still not agreed upon. A common view amongst researchers is that hypnosis can profoundly influence the way that individuals monitor and evaluate their experiences but that it is not able to influence behaviours that are normally outside of conscious control.

A recent study we completed provides evidence that, in certain contexts, hypnosis may actually be able to inhibit typically automatic responses.

Our study adapted an intriguing experimental paradigm developed by Daniel Wegner: the Clever Hands task.

In this task participants are given a series of trivia quiz questions. These are all binary choice, yes/no questions. Most of the questions are incredibly easy (e.g., ‘Is the sky blue?’) but a small number of questions are extremely difficult (e.g., ‘Are there 7107 islands in the Philippines?’ There are not. The correct answer is 7641, in case you were interested). But there is a catch: participants are instructed to answer all the questions randomly. It turns out people are terrible at this task.

In Wegner’s original study he found that participants answered 82% of the easy questions correctly. Importantly, participants estimated their responses as very close to chance. So people believed they had responded randomly in this task when actually the knowledge of the correct answers dramatically influenced their responses. Wegner tried to encourage random responding by imposing time limits on answers and by offering financial rewards for random sequences of responses. These manipulations usually are effective ways of increasing response rates for controllable behaviours, but in this case neither changed the pattern of results. Wegner concluded that giving correct responses in this task was an automatic behaviour outside of conscious control.

In our adaptation, we wanted to see whether hypnotic suggestions might lead to a different pattern of performance in the Clever Hands task. Our lab has been using hypnosis to model the features of clinical delusions and we have found that suggestions based on specific clinical phenomena can lead to marked changes in agency and belief. In this task we used two suggestions based on clinical thought insertion and clinical alien control, and also a control suggestion that just instructed participants to respond randomly.

As expected, the participants asked to answer randomly actually didn’t. They answered most questions correctly but believed that they had followed the instructions (i.e., the same pattern of results as Wegner’s original study). The same was true for the ‘thought insertion’ condition, although these participants were slightly better at estimating how many questions they answered correctly. The real surprise came from the ‘alien control’ condition; these participants actually responded in a more random way.

This indicates that for these people, hypnosis inhibited the usual tendency to answer easy questions correctly – something that was not possible by standard behaviour manipulations outside of hypnosis. This is a curious finding that indicates that hypnotic suggestion may be able overcome responses that are usually thought to be automatic.