Mixing Memory

One of the top ten coolest experiments ever has to be Botvinick and Cohen’s “rubber hand” experiment1. I’m going to let them describe the manipulation:

Each of ten subjects was seated with their left arm resting upon a small table. A standing screen was positioned beside the arm to hide it from the subject’s view and a life-sized rubber model of a left hand and arm was placed on the table directly in front of the subject. The subject sat with eyes fixed on the artificial hand while we used two small paintbrushes to stroke the rubber hand and the subject’s hidden hand, synchronising the timing of the brushing as closely as possible. (p. 756)

This lasted ten minutes, after which the participants (only in Nature can you call them “subjects” anymore) were asked several questions designed to get them to describe their experience. Each participant indicated that after a short time, they experienced feeling the brush rubbing against the rubber hand, instead of their own hand. In order to confirm this experience, Botvinick and Cohen ran the experiment again. This time, both before and after the brushing on the rubber and real left hands, participants slid the index finger of their right hand across the bottom of the table until they believed they’d reached the position of their hidden left hand. After watching the brush strokes on the rubber hand, participants estimates of the location of their left hand were shifted to the right, toward the position of the rubber hand. Subsequent studies have confirmed this illusion, demonstrating, for example, that if the brush strokes are asynchronous, perceived shift from one’s real hand to the rubber hand does not take place2. It seems that in addition to believing, seeing is also feeling.

That’s just really cool, right? But this post isn’t about rubber hands, it’s about mirror neurons. In case you haven’t heard, mirror neurons are cells, first discovered in the premotor cortex of monkeys, that respond both to the movements of an individual and to similar movements that the individual observes in others. Hence “mirror” neurons. There is now a large body of evidence that a similar “mirror neuron” system exists in humans, in the areas around the right superior temporal sulcus. The locations of the mirror neuron systems in monkeys (A) and humans (B) can be seen in this figure3:

i-cb51df9b3cacdb73346cc60a4b53f313-mirrorneuronsmonkeyhuman.jpg

Now, all sorts of wild claims have been made about what the existence of a mirror neuron system in humans means. Overzealous researchers have claimed that the system underlies everything from empathy to the evolution of language, with one even saying that “mirror neurons will do for psychology what DNA did for biology”4. To date, however, the rhetoric surrounding the mirror neuron system has far outpaced the actual research, and when experiments are actually conducted, the conclusions researchers draw from them often go well beyond their data. C’est la neurologie cognitive.

One of the interesting things about mirror neurons is that they’re supposed to encode both our own actions and those of others. The question that immediately arises, then, is how do we distinguish between our own actions represented by mirror neurons, and the actions of others represented by those same neurons? This is important because if a set of neurons fires both when we perform an action and we observe someone else doing so, we need to make sure that they don’t cause us to perform the action when we’re observing the action. Then we’d just look like silly mimes. An important research question, then, concerns what differs in the brain’s response to performed vs. observed actions.

You probably see where this is going by now. A problem in testing the differences between self and other processing in the mirror neuron system is that when we perform actions, we have a whole set of visual and other perceptual (non-motor, and thus non-mirror system) cues that tell us the actions are ours. Since the rubber hand illusion presents an interesting example of a failure to detect a difference between self and other, it provides an interesting way to look at differences in brain activity when processing the actions of self vs. other without those visual cues getting in the way. In a clever experiment published in last week’s issue of Current Biology, Schütz-Bosbach et al.5 used it for just this purpose. However, instead of using a rubber hand, they used the real hand of an experimenter, whose body was hidden behind a partition (see the figure below , from p. 1831).

i-aacc4a9c49a6ffac5813bccb817a5a3c-Mirrorneuronsrubberhand.JPG

In one condition, the synchronous condition, the rubber hand illusion was induced in the same way as in the original, with a small paint brush used to stroke both the experimenter’s visible hand and the participant’s unseen hand synchronously, thus creating the rubber hand illusion. In the other condition, the asynchronous condition, the brush strokes on the experimenter’s and participant’s hand were not synchronized, meaning that the participant would not “feel” the brush strokes on the experimenter’s hand. After the initial stroking, the experimenter began to move his index finger randomly, at random intervals, with either synchronous or asynchronous (depending on the condition) brush strokes to the experimenter’s and participant’s hands in between movements. The idea is that if, in the synchronous condition, the participant experiences the brush strokes in the experimenter’s hand, then he or she may also experience the actions performed by that hands as being self-performed. In the asynchronous condition, the participant would not feel the strokes on the experimenter’s hand, and thus would interpret the actions of that hand as being performed by the experimenter, and not his or herself. In order to cause evoked potentials, after watching the movements (in both conditions), the experimenters applied transcranial magnetic stimulation to the motor cortex (in the areas thought to be associated with the mirror system). Transcranial magnetic stimulation allows you to measure activation in particular areas of the brain, sort of like applying an electrode directly to that area of the brain, but without having to cut open the skull and deal with all that blood and goo. The experimenters then measured the electrical response in the muscle that controls the finger associated with the finger that the experimenter used to perform the observed actions (the right first-dorsal interosseus).

The important contrast, then, is in activation of the nerves in that muscle when observing an action felt to be performed by the self vs. an action felt to be performed by another. They found that when observing an action performed by another (asynchronous condition), activity increased in the muscle that controls the index finger (the finger that the experimenter moved), and a decrease in electrical activity in that muscle when the action was thought to be performed by the self (synchronous condition). Thus, the motor system clearly distinguishes between actions performed by the self and by others. Here’s how the authors interpret this result:

This suggests that the neural mechanisms underlying action observation are intrinsically social. These mechanisms map the actions of others to corresponding actions on one’s own body but do not simply represent the other agent as a derivative of or even an equal to, the self. (p. 1832)

The results thus show that the motor system, and perhaps then the mirror neuron system, can distinguish between actions of the self and other. It’s not exactly clear why viewing the actions of another increases activity in the motor system, while viewing one’s own actions (or actions believed to be one’s own) suppresses it. The authors don’t offer much in the way of an explanation for the particular way in which the motor system represented self and other in their study, though I suspect it has something to do with the fact that when one performs an action, one doesn’t need to use the motor system to subsequently represent it (it’s already been represented there, when it was performed, unless the brain is tricked as it is in the rubber hand illusion). These results also show, as the authors argue, that the motor system is important in social interactions, as mirror neuron proponents have argued. So, for once, mirror neuron research has caught up to a piece of mirror neuron rhetoric.


1Botvinick, M., & Cohen, J. (1998). Rubber hands ‘feel’ touch that the eye sees. Nature, 391, 756.
2Armel, K.C., & Ramachandran, V.S. (2003). Projecting sensations to external objects: Evidence from skin conductance response. Proceedings of the Royal Society of London, B: Biological Sciences, 270(1523), 1499-1506; Tsakiris, M., & Haggard, P. (2005). The rubber hand illusion revisited: Visuotactile integration and self-attribution. Journal of Experimental Psychology: Human Perception and Performance, 31, 80-91.
3From Rizzolatti, G., & Arbib, M.A. (1998). Language within our grasp. Trends in Neurosciences, 21(5), 188-194.
4V.S. Ramachandran, quoted in Hurford, J.R. (2004). Language beyond our grasp: what mirror neurons can, and cannot, do for language evolution. In D.K. Oller & U. Griebel (eds.), Evolution of Communication Systems: A Comparative Approach, pp.297-313. MIT Press: Cambridge, MA.
5Schütz-Bosbach, S., Mancini, B., Aglioti, S.M., & Haggard, P. (2006). Self and other in the human motor system. Current Biology, 16, 1830-1834.

Comments

  1. #1 Tanasije Gjorgoski
    October 3, 2006

    Hi Chris,
    . They found that when observing an action performed by another (asynchronous condition), activity increased in the muscle that controls the index finger (the finger that the experimenter moved), and a decrease in electrical activity in that muscle when the action was thought to be performed by the self (synchronous condition).
    I’m wondering about “the action was thought to be performed by the self” formulation. Seems to me that in the experiment there is no reason to assume that in the synchronous condition the action is thought to be performed by the self. I would expect for example that it might “break” the illusion that the experimenter’s hand is subject’s.
    Am I missing something?

  2. #2 Chris
    October 3, 2006

    A couple responses. One thing I left out of the post, just for brevity’s sake (I know, that’s funny given how wordy I am) is that they did the “pointing” test, as in the original rubber hand study, to demonstrate that the illusion was taking place. That means people really were “feeling” the brush strokes in the rubber hand. Second, the rubber hand illusion is well established, using several different objective meausres. Of course, we can’t be entirely sure that they experienced the action as their own, but I assume future research will seek to replicate these results with other experimental paradigms.

  3. #3 Tanasije Gjorgoski
    October 3, 2006

    Chris,
    Thanks for the explanation. I think it was pretty clear from what you wrote that the subject is experiencing the hand as his own.
    I’m sorry I wasn’t clear enough – what I was wondering is how the lack of agency (i.e. the lack of decision to move the finger), might affect the whole experiment, and the results. For example, the subject didn’t decide to move the finger, and yet he sees “his finger moving”. I can imagine the subject thinking “why does my finger move?”. But, I guess even in this case, the subject is wondering about something happening with his body, so the experiment tells us something about distinction between our own and action of other.

  4. #4 Chris
    October 3, 2006

    ah, I just misread the question. The answer is probably yes, the lack of agency could affect it. That’s one of the questions that future replications will have to look at . However, work by people like Dan Wegner has shown us that the experience of agency is often post hoc, and it’s not unreasonable to think that people may even have had a feeling of intentionally moving the finger, if only for a moment. In fact, it might be interesting to use one of Wegner’s paradigms to try to replicate these results.

The site is undergoing maintenance presently. Commenting has been disabled. Please check back later!