Tuesday, 26 February 2013

Embodiment and 'The Hard Problem'

First off, I apologise for the corny picture. I could not resist. Consciousness is a good topic for these things. It’s mysterious still, an undiscovered country, theoretically up for grabs and thus ripe for colonisation by idiosyncratic, metaphysically consoling worldviews lending themselves to expression in gauche pictures. Why not celebrate that a little? Now, on to the post.

Chalmers (1996) offers a particularly simple critique against all current theories purporting to explain either the function or cause of conscious experience. This is the ‘in the dark’ problem – essentially, there is nothing about consciousness that makes it necessary. The logical possibility exists of a world identical to our own but populated by experience-free zombies. At the level of current theories, this is devastating. For example, Baars (1988) suggests consciousness is a global workspace, a system for the communication of information between separate agencies within the brain. But why should this intra-brain communication produce the phenomenon of conscious experience? Can’t intra-brain communication go on in experiential darkness? This criticism can be rinsed and repeated for every mainstream theory of consciousness I can think of. If we momentarily limit the scope of the critique to the orthodox position of the mind as an enclosed, information-processing system, there seems to be no necessary connection between the processing of information and the genesis of experience. What, then, about embodied theories? Does embodiment necessitate conscious experience in a way that information-processing does not, and can it thereby approach ‘the Hard Problem’ of explaining why we have experience in a more fruitful manner?

In this sympathetic commentary on Evan Thompson’s Mind in Life (2007), philosopher Charles Siewert lays out the basics of the embodied response to Chalmers-style Zombie arguments. Drawing on Husserl and Merleau-Ponty, Thompson claims that because objects achieve their perceptual unity through our bodily experience of them from a variety of perspectives, a functionally equivalent Zombie world could not exist. We understand the world and thus successfully interact with it precisely because we have individuated objects within it by experiencing them from a multiplicity of different points in space and time. Embodiment produces experience and experience is necessary for action. But still, why do we need experience for this exactly, rather than just detection? Would sensory detection in the absence of personal phenomenal experience not suffice?

Andy Clark (of extended cognition fame), in this 2002 overview of theories of visual consciousness, initially focuses on such phenomena as change blindness to illustrate the inaccurate nature of our perception of the world. This leads him to contrast a number of theories, among them one representationalist and one embodied. He dubs the representationalist theory ‘The Grand Illusion’, the idea that we have a rich but flawed internal representation of the world. The embodied theory, namely Skill Theory (Noe and O’Regan), proposes that we have no such rich inner representation. ‘The world is its own best model’, and rather than constructing an internal representation, the world ‘out there’ forms a consistent, external source of memory which we consult through ocular saccades to the relevant point in the visual field. Consciousness bleeds out into the world and both constitutes and is constituted by it. Interestingly, Clark, usually a committed post-cognitivist, adopts what looks like a compromise position here because he feels that positions such as Noe and O’Regan’s disregard the Hard Problem. If Noe were correct, then current generation robots capable of making rudimentary environmental discriminations would by necessity have a basic form of conscious experience. Clark feels this stance arises due to ‘sensorimotor chauvinism’ and exhorts us to take note of the fact that neurologists have identified two visual pathways, the dorsal and ventral stream. Clark seems to imply that consciousness should be associated with the latter, and is a process of projecting our ‘interests and concerns’ (these terms remain somewhat vague) onto the world for the purpose of complex direction of future action. Visual consciousness is the construction of a world which we can think with and should not be over-associated with direct sensorimotor causes. Here, Clark seems to be aligning with Higher Order Theories of consciousness as put forth by David Rosenthal; fine-grained concepts produce fine-grained conscious experiences. This seems a very representationalist stance for a post-cognitivist. Does the Hard Problem force a certain degree of back-tracking on embodied theories, as Clark suggests, or has he just been hanging round with Chalmers too much and thus formed an overly daunting impression of the explanatory gap?

In this paper, Jesse Prinz characterises positions such as Noe and O’Regan’s (and presumably Thompson’s as well) as maintaining that in the absence of bodily interaction with the world, consciousness could not exist. Brains in vats would thus have no conscious experience. Prinz provides a variety of objections to this strong stance (such as that people can hallucinate without physical interaction with the world, or have elaborate conscious experiences while inside sensory deprivation chambers, a kind of voluntarily-endured short-term human equivalent of Harry Harlow’s Pit of Despair). He ends up endorsing a moderate view of embodied consciousness, suggesting that such higher-order, secondary phenomena as self-consciousness and the perception of unity and ownership of action may rely on embodiment, but consciousness could exist without it. In this way, he seems to come to conclusions which are comparable to Clark’s. Consciousness is a top-down, higher order product as opposed to something that arises out of our sensorimotor system. A very different view is put forward briefly in this blog post, which outlines the position of Ed Slingerland, someone who sees an embodied approach as providing a way between the rival pratfalls of objective realism and postmodernism as we attempt to explain human experience.

Unsurprisingly, I don’t have the answer to the Hard Problem myself. I am unsure as to whether cognitivist or strongly embodied theories provide the better approach towards either answering or successfully dissolving it.


  1. Excellent foot for thought there Hugh!

  2. Nice post, I've the links loaded and ready to read. I certainly agree with Chalmers' emphasis on the hard problem and his dissatisfaction that various theorists set out to solve the 'problem of consciousness' but inevitably end up talking about something else.

    Where I disagree with Chalmers is his rather polite approach of allowing everything else but consciousness to be explained and then trying to latch on the consciousness that is left. The only reason we know of anything physical is via consciousness.

    That's not to say that things don't exist when we're not aware of them but simply that we can't speak of them in any meaningful way, and certainly not empirically, without recognising them as aspects of consciousness.

  3. Thanks John. I'm not sure I subscribe to your idea that things 'are aspects of consciousness'. I'd say our experiences of things (or the fact that we experience some of the sensory flux as particular 'things', if you prefer) are aspects of consciousness, but this is a more trivial claim.

    So I grant that the world presents itself to us through consciousness and we can have no experience of it except through the structure of our consciousness. However, I think there must be some relationship between our conscious experience of things and things as they are in themselves, though it is no doubt highly constructed. I don't think we experientially constitute a world without any anchor in reality outside of ourselves; this would just be far too unadaptive. What is useful is very often also what is true independently of us. I suppose I feel that the attempt to say the world is an aspect of consciousness is anthropocentric, even solipsistic. We can dissolve the boundary between mind and world to a degree, or make it shiftable according to context, but I don't know if we can let one consume the other. That way lies eliminative materialism or solipsism, depending on the direction.

    I certainly agree that a scientific treatment of consciousness would be wonderful (were such a thing possible), but I'm not sure how we could go about ultimately ground scientific enquiry in phenomenology.

    1. I do, however, completely agree that Chalmers' speculative notions of what consciousness is are spurious.

  4. "without recognising them as aspects of consciousness."

    My wording in the above sentence may have been a little more categorical than I intended it: I mean simply to emphasise that speaking empirically of something, absent of conscious experience, will never be possible by definition as any such thing will always have the imprint of consciousness upon it, i.e. the experiencing and the experience will have an inseparability.

    "I suppose I feel that the attempt to say the world is an aspect of consciousness is anthropocentric, even solipsistic."

    I think I'd see 'the world' and 'consciousness' as simply being inseparable from each other: neither better or worse but just parts of the same whole that can be split for analytic purpose but ultimately co-occur.

    So in that sense consciousness certainly isn't better than or more important than that which we experience, but I do take issue with the materialistic approach of posing an objective world that exists outside of conscious experience.

    I think there is a risk of anthropocentrism if we assume that everything works in the way we do, however it is equally dangerous, and probably covertly anthropocentric, to forget the role of the observer.

    Probably the best human enquiry can yield is a form of constructive objectivity, where we put together venn diagrams of what we each subjectively know, however not losing sight of our own limitations in doing so.

  5. I see what your saying. I will try and lay out my position; perhaps they really aren't so different. I think that positing that we have an objective view of the world is wrong, but positing that there is a world is not.

    I don't advocate disregarding the role of the observer, and I don't intend on reifying some particular analytic split, but I do think that if I were the last conscious being in the world and I died, the world would still be there. That is not to say that it would be as I experienced it; clearly that makes no sense. But it would be there.

    I would also think that, while alive, I experienced the world the way I did because I had to survive in it. This does not necessitate that all things with experience experience the world the same way; the way the world is experienced by them will depend on how it is advantageous to them that it should appear. However, based on this, I also make the assumption that it will generally be most advantageous to have experience with some kind of truthful relationship to the world, even if it is not, and can never be, some sort of absolute accurate reflection of reality.

    I also think any effort to overcome an analytic split between subject and object or self and world can only begin when we aknowledge our experience is actually of something. Otherwise how do we know we're not the old brain in the vat?

  6. "I think that positing that we have an objective view of the world is wrong, but positing that there is a world is not."

    I like that distinction. I also agree that there is more going on than we are conscious of at a given moment; and calling that the world is fine though as you say from a non-human perspective it may be perceived as a very different type of thing than what we know as the world.

    Emphasising the role of the observer in science shouldn't, I feel in practice, lead to the need for changing things that much- it's more just how theories are framed. What I think is the wrong way to proceed is to try hard to explain everything but consciousness, leaving it to the end of the list because it is the hardest, and then try to explain it as if we are starting from scratch. Chalmers, though I admire his emphasis on consciousness, does this in his article.

    Everything that we will have just said while explaining the 'physical' stuff will have necessarily involved a description of consciousness too, as we won't just have described the 'world' but our phenomenal experience of the world.