January 19, 2006
How the body shapes the mind - An interview with Shaun Gallagher
SCR Feature, discussion & interview — thomasr
An Interview with Shaun Gallagher
- Professor Gallagher, in your book called “How the body shapes the mind” you start by pointing out and discussing the crucial distinction between body image and body schema. In what way is this distinction meaningful?
- In the book I cite a number of studies that suggested that these concepts are problematic, and they certainly are in the way that they have been used in the literature - in psychology, medicine, neurology, psychiatry, psychoanalysis, and even philosophy. I argue, however, that these concepts can be useful as long as we make a clear distinction between body image and body schema. I suggest a distinction along the lines of the difference between perceiving one’s own body (or having a concept of it, or having an emotional attitude toward it - in all these cases we take the body as an intentional object) and being able to control bodily movement, which doesn’t depend on perceptual monitoring or taking the body as an object. The former helps to constitute the body image; the latter aspect of motor control is related to the body schema. I try to be careful and point out that this is a conceptual distinction and that in our everyday behavior these things are not so unambiguous. But to justify the conceptual distinction I point to clear dissociations between body image and body schema, in cases of unilateral neglect, for example, where there is a problem with body image, but, in many cases, We need to make a clear distinction between body image and body schema no problem with motor control even on the neglected side. On the other hand we can point to cases of deafferentation, as in the case of IW who has lost proprioception and the sense of touch from the neck down. Jonathan Cole and I have interpreted this as a loss (or at least a partial loss) of body schema, which IW compensates for abnormally (in terms of motor control) by his enhanced body image (Gallagher and Cole 1995). So we have something of a double dissociation as the empirical evidence for the distinction.
- Let me add that recently Gallese (2005) has suggested that this distinction is less clear on the neurological level because some of the neurological processes responsible for the body image are also activated in body schematic (motor control) processes. I can’t argue with the neurological account, and in fact this may help to explain how IW is capable of using a body image to control his movement. As I indicated, I think that in normal behavior it is not easy to see these processes distinguished, and some shared neurological processes make sense of this. Still, I think it is possible to find cases in which one of these aspects of embodiment is disrupted but the other left intact. And conceptually it makes just as much sense to distinguish between body image and body schema as it does to distinguish between seeing someone act, or imagining an action, and actually doing the action, despite the fact that there are some common neuronal processes involved in those phenomena.
- Gallese has definitely been one of the most prominent researchers on so-called ‘mirror neurons’, sometimes referred to as the “single most important ‘unreported’ story of the decade” (Ramachandran at). Mirror neurons are described as neurons which fire both when performing an action and when observing the same action performed by another (possibly conspecific) creature. How would you relate your distinction between body schema and image to the mirror neuron findings? Would there be a difference between the body image of seeing others’ actions and seeing oneself?
- I agree about the importance of the mirror neuron discovery. The big question is how to interpret this. In regard to the body image and body schema distinction, again I want to say that this is a conceptual distinction and that on behavioural and neurological levels these The simulating subject uses her own mind as a model for understanding the other person’s mind things are intertwined. The most interesting thing about mirror neurons, or more generally resonance systems, is that they are not just about one body but about what Merleau-Ponty called intercorporality - how one body relates to another. It’s the neuroscience of intercorporality that is helping us explain how our intersubjective relations are possible, without resorting to amorphous concepts like ‘universal spirituality’ or ’shared human nature’. But even with this good neuroscience we need to be careful about finding the right theoretical framework for developing an account of how we understand others. Gallese has joined forces with simulation theorists, and I’m not sure this is the right way to go. I’ve argued, in some recent papers (Gallagher, in press), that the concept of simulation, as it is developed in simulation theory, signifies an activity on the part of the subject. As it is usually described, the simulating subject uses her own mind as a model for understanding the other person’s mind. She introduces “pretend” beliefs or desires into the model, and in the end makes an inference about what the other person must be experiencing. Gallese is not defending this explicit, and often introspective version of simulation theory (Goldman is someone who does defend this view). Rather he is proposing that the simulation process is subpersonal and carried out by the mirror system. But in that case, my question is, What does simulation mean? All descriptions of simulation that I have seen suggest that it is something in which the subject actively engages. But the activation of the mirror system is not something that we actively engage in; it happens automatically. It’s not something that I do; it’s something that happens to me. In a sense, it is something that the other person elicits in me. So it strikes me as strange to call this a simulation. I think it’s better to view the activation of resonance systems as part of a perceptual process that gives us access to the other person’s intentions. And I think this is clearly consistent with the idea of enactive perception.
- So, the notion of ’simulation’ implicates a conscious agent, while the description put forth by Gallese actually points out that these processes occur at a subconscious level. From this, it seems that a distinction between conscious and unconscious processes can help both to clarify concepts like ‘mirror neurons’, as well as to put forth specific questions relating to the subserving mechanisms? Could such a distinction be helpful in distinguishing between processes at a neural level?
- I think the issue is not about conscious versus non-conscious - although there is certainly a question about how much is conscious and how much is not. The real issue has to do with how one defines simulation. Whether simulation is conscious or not, it is always defined as some kind of proactive process - I do something, or my brain does something or uses some kind of model in a controlled way to accomplish something. If you look at the way simulation is described, it is always this kind of proactive engagement. Goldman’s (2002, 2005) descriptions of a mindreader trying to predict or “retrodict” someone else’s mental states pictures the subject as engaged in stepwise activity. The idea is that I, as subject, first deliberately create some pretend beliefs. I then, as Goldman describes it, “feed” these beliefs into some kind of routine mechanism, and then I assign the result to the other person. Pierre Jacob (2002), for another example, characterizes simulation as engaging in an “activity,” a “heuristic,” or “methodology” in order to compare and predict mental states. If this is what we mean by simulation, what happens to the concept when we attribute simulation to subpersonal brain processes? We still get descriptions of proactive processes. Gallese, for example, still considers it a process of modelling in a stepwise fashion in which the second step is simulative modelling of the other’s intentional actions. What we know is that when I see another person’s action there is, among other things, sensory activation in the visual system followed by activation in the pre-motor cortex. Moreover, it’s agreed that this is something that takes place automatically. But to call it simulation is to suggest a controlled, proactive process. It would be better to call it a perceptual elicitation since the resonance is automatically generated in my system by the action of the other. If you insist that this kind of perceptual elicitation is a simulation, then, at best, you are changing the definition of simulation, and at worst, distorting our understanding of what is going on.
- Your question about conscious and non-conscious processes also points to another problem that I often see even in some of the best thinkers. The problem is that when they find a set of distinguishable neuronal processes occurring, they think that the subpersonal distinctions demonstrate distinctions on the personal or conscious level. So if there is a stepwise process at the subpersonal level, that signals a stepwise process in conscious experience. A good Phenomenologically, intentions in almost all cases come already clothed in agency example of this can be found in an otherwise excellent article by Marc Jeannerod and Elizabeth Pacherie (2004). They say lots of things that support the idea that our understanding of others is based on a direct perception of their intentions, but they nonetheless defend a version of the simulation theory. In their version they find a stepwise process on the neuronal level. They can distinguish (1) sensory input, (2) the activation of shared representations, and (3) the activation of what Georgieff and Jeannerod (1998) call the “Who system.” The claim is that activation of shared representations is neutral with respect to first- or third-person determinations. That is, the resonance processes don’t tell us who is acting; specification of agency involves some further step in neuronal processing. On this basis they suggest that we are aware of “naked intentions” to which we then have to add authorship. Here is what they say.
We can be aware of an intention, without by the same token being aware of whose intention it is. … Something more than the sole awareness of a naked intention is needed to determine its author. … When the naked intention one is aware of yields an overt action, the extra information needed to establish authorship may be found in the outside world. The question ‘Is this intention mine?’ would then be answered by answering the question: ‘Is this my body performing the corresponding action?’ (Jeannerod and Pacherie 2004:140).
- I think there is an obvious problem with this. Phenomenologically, intentions in almost all cases come already clothed in agency - the ‘who’ question hardly ever comes up at the level of experience. The neural systems have already decided the issue - one way or the other - even if I’m wrong about who is acting, I am still attributing agency. The mistake is to think that there is a necessary isomorphism between the phenomenological level and the neuronal level. If the neuronal processes can be defined as involving three steps, this does not mean that those three steps need to show up in consciousness. The wonderful thing about the “Who system” is that it is neurological - and the results of its activation are hardly ever experientially manifested as “making a decision about who did the action.” Rather, the results of its activation are experientially manifested as “X’s action” where X is either you or me. Of course experiments and pathologies may generate or reveal ‘who’ problems, but in normal ecological behavior it is generally clear whose intention/action it is, and as a result, the identification question - “Someone is intending to pick up the apple, is it me?” - just doesn’t come up.
- You suggest that in cases of psychopathology such as schizophrenia, agency and ownership are disentangled and made visible. How is this so? Has this distinction been recognised at a neuronal level?
- Well we can start thinking about this distinction in the simple example of involuntary movement. If someone unexpectedly pushes me, I have a sense that my body is moving - that I am moving - and this is what we can call the sense of ownership. But since I am not the cause of my movement, I don’t have a sense of agency for it. So this is a distinction that seems straightforward - and it’s purely a phenomenological distinction at this point.
- I think these experiences of ownership and agency are manifested at the level of Experiences of ownership and agency are manifested at the level of first-order, pre-reflective, phenomenal consciousness first-order, pre-reflective, phenomenal consciousness. That is, I don’t need to reflect on what I’m doing to generate these experiences. Rather, they are part of and implicit in what my movement feels like. Of course this does not preclude reflection on my movement, or the possibility of explicitly attributing ownership and agency to myself. Graham and Stephens (2000), for example, make a similar distinction, but place it on a reflective, higher-order level of consciousness. They talk about the “attribution of agency” and the “attribution of subjectivity” or ownership, and they conceive of these things as the result of a subject attributing agency to himself in the same way as they might attribute agency to someone else.
- In the case of schizophrenic delusions of control and symptoms of thought insertion, the subject’s sense of agency for his movement or for his thoughts seems to go missing. The subject complains that these things are happening to him, so he has a sense of ownership for the movement, or the idea that seems inserted is inserted precisely in his stream of consciousness, but he does not have a sense of agency for the movement or thought. Again, in contrast to Graham and Stephens, who suggest that this is the result of something going wrong on the level of the schizophrenic’s reflective or introspective consciousness, I’ve suggested that the schizophrenic actually experiences this alienation at the level of first-order consciousness. It’s not just a case of mis-attribution; the subject really experiences it. And I think the neuroscience backs this up.
- Actually, there has been a flurry of neuroscience experimentation based on this distinction - or at least a good number of recent brain imaging studies have cited the distinction which I made in a TICS article in 2000 (Gallagher 2000; see e.g., Chaminade and Decety 2002; Farrer and Frith 2002; Farrer et al. 2003; Ruby and Decety 2001). These experimental studies try to sort out the neural basis for a sense of agency. Farrer et al. (2003), for example, showed that the sense of self-agency correlated with greater activity in right posterior insula. In contrast, activity in the right inferior parietal lobe seems to signal that agency belongs to someone else. Following the contrastive logic of involuntary versus voluntary movement, I had suggested that the sense of agency was likely based on efferent signals involved in movement preparation (something present in voluntary action), while the sense of ownership was associated with proprioceptive and sensory feedback (which we have in both the case of involuntary movement and voluntary movement). Tsakiris and Haggard (2005a, 2005b) have recently run some fascinating experiments on this showing that while efference provides an intrinsic sense of agency, the sense of ownership is more complex and depends on multisensory comparisons between efferent and afferent signals. The more integrated these signals, the clearer the sense of ownership for bodily movement.
- How would you couple this distinction between agency and ownership to the distinction between body image and body schema? As you noted in the beginning of this interview, a body image is the perception of one’s own body, while a body schema is related to the ability to control movement. As you also noted the neuroscientific evidence points to less clear boundaries between the image and schema. But would you say that the sense of agency is (mostly) related to a body schema process, and that the sense of ownership is (mostly) related to the body image?
- I think it is good to be cautious here. Lining up these distinctions in this fashion is very tempting, but I think that each of these concepts is complicated, and the underlying factors of proprioception and motor efference are also complex. One could think of sensory feedback for action as contributing to the body image, Body image is also influenced by higher-order reflective cognition especially if one understands proprioceptive and visual feedback as conscious. But much of that kind of sensory processing remains non-conscious and closely connected with motor control - and I would associate that with body schema. One could conceive of body schematic or motor control aspects as generating first-order conscious senses of ownership and agency, and these experiences as contributing to the constitution of the pre-reflective aspects of body image. Body image is also influenced by higher-order reflective cognition, as when we actually think about our bodies. Drawing some basic conceptual lines between these different phenomena is important in trying to come up with explanatory models, but we always need to remind ourselves that at the behavioural and phenomenological levels all of these phenomena are interconnected. Explicit body awareness (which contributes to body image) sometimes contributes to motor control, and to the senses of ownership and agency; non-conscious proprioceptive processes that are primarily about motor control sometimes generate first-order experiences that contribute to body perception. The conceptual distinction between body image and body schema can be useful in some theoretical and experimental contexts, but it can also be an oversimplification and it is sometimes the case that we need to dig deeper into the sensory and motor processes in order to get an adequate framework for explanation. This is what makes embodiment so interesting. There are important distinctions between body and environment, action and perception, body schema and body image, proprioception and other sensory modalities - but no one of them carves up the conceptual landscape in a way that adequately captures the variations in normal and pathological behaviour and experience. So there are other concepts like first-order and higher-order cognition, enactive perception, situated cognition, self and degrees of self-consciousness, and intersubjectivity - and we have to try to integrate these concepts to draw a more complete picture of how embodiment generates and shapes human experience.
- Chaminade, T. and Decety, J. 2002. Leader or follower? Involvement of the inferior parietal lobule in agency. Neuroreport, (In press).
- Farrer, C. and Frith, C.D. 2002. Experiencing oneself vs another person as being the cause of action: The neural correlates of the experience of agency. NeuroImage 15: 596-603.
- Farrer, C., Franck, N. Georgieff, N. Frith, C.D. Decety, J. and Jeannerod, M. 2003. Modulating the experience of agency: a positron emission tomography study. NeuroImage 18 (2003) 324-333
- Gallagher, S. 2000. Philosophical conceptions of the self: implications for cognitive science. Trends in Cognitive Sciences 4 (1): 14-21.
- Gallagher, S. 2005. How the Body Shapes the Mind. Oxford: Oxford University Press/Clarendon Press.
- Gallagher, S. (in press). Logical and phenomenological arguments against simulation theory. In D. Hutto and M. Ratcliffe (eds.), Minding our Practice: Folk Psychology Re-assessed. Springer Publishers.
- Gallagher, S. and J. Cole. 1995. Body schema and body image in a deafferented subject. Journal of Mind and Behavior 16: 369-390.
- Gallese, V. 2005. Embodied simulation: From neurons to phenomenal experience. Phenomenology and the Cognitive Sciences 4 (1): 23-48.
- Georgieff, N. & Marc Jeannerod. 1998. Beyond consciousness of external events: A Who system for consciousness of action and self-consciousness. Consciousness and Cognition, 7, 465-77.
- Goldman, A. I. 2002. Simulation theory and mental concepts. In J. Dokic and J. Proust (eds.), Simulation and Knowledge of Action (1-19). Amsterdam: John Benjamins.
- Goldman, A. 2005. Imitation, mind reading, and simulation. In Hurley and Chater (eds.) Perspectives on Imitation II (80-81). Cambridge, MA: MIT Press.
- Jacob, P. 2002. The scope and limits of mental simulation. In J. Dokic and J. Proust (eds.), Simulation and Knowledge of Action (87-109). Amsterdam: John Benjamins.
- Jeannerod, M. and Pacherie, E. 2004. Agency, simulation, and self-identification. Mind and Language 19 (2) 113-46.
- Ruby, P. and Decety, J. 2001. Effect of subjective perspective taking during simulation of action: A PET investigation of agency. Nature Neuroscience, 4 (5): 546-550.
- Stephens, G. L. and Graham, G. 2000. When Self-Consciousness Breaks: Alien Voices and Inserted Thoughts. Cambridge, MA: MIT Press.
- Tsakiris M & Haggard P (2005a). The rubber hand illusion revisited: visuotactile integration and self-attribution. Journal of Experimental Psychology: Human Perception and Performance, 31(1):80-91.
- Tsakiris M & Haggard P (2005b). Experimenting with the acting self. Cognitive Neuropsychology, 22(3/4), 387-407.