From the Journal of Neuro Aesthetics

 


CHARLES WOLFE on PROPRIOCEPTION
Charles Wolfe

Proprioception: from "proprius-ception, ‘one's own’ ception (...) the 'body' itself as, by movement of its own tissues, giving the data of, depth"
(Charles Olson, "Proprioception")

Up until now, it has not proved common in philosophy for readers to experience the breathless rush through sensory landscapes one finds in novels by William Gibson, or better, Jeff Noon (Vurt, Pollen). Philosophers of most stripes like to recapitulate; to reconstruct a context; to take you back a few steps in time and show you the emergence of an idea. Granted, there have been a few books in recent years coming out of the phenomenological tradition which emphasize the most physical (embodied, carnal, corporeal, fleshly, perceptual, etc.) elements of that tradition. The relevant names here are Merleau-Ponty, and before him Husserl and Kurt Goldstein, whose fascinating book The Organism has recently been reissued by Zone Books. (Thinkers who seek to emphasize the materiality or concreteness of social relations, that is, the primacy of the social world in structuring our consciousness of self, will instead invoke Heidegger, who has been translated into the American idiom, or should one say Californian, by people like J. Searle over the last 15-20 years: no part of human action can be understood in isolation from the "Background." But here we are concerned with the body and its environment, rather than society and the self.) But nonetheless, ever since Kant, in Dreams of a Spirit-Seer, reproached Swedenborg for trying to build a philosophy on private experience, whether mystical or hallucinatory, philosophers have shied away from the 'kaleidoscope' approach. Well, Andy Clark's Being There (subtitled Putting Brain, Body and World Back Together) puts the aforementioned novels to shame. We move at lightning speed through Artificial Life robotics, Gibsonian perception studies, Scott Kelso's dynamic systems models, the slime mold ... all of this glazed over or dusted with a fine sheen of dernier cri phenomenological utterances. The title, Being There, would hopefully not be allowed outside of MIT's cognitive science series, and its cheerful cult of bad taste. What I'll be trying to do here is to talk a leaf out of Clark's book and argue for the priority of bodily reality in understanding brain and mind.

Why proprioception? Isn't it a mere physical feature of the working of our organism, like metabolism? Well, if we are interested in thinking through the body, as a critic once put it, we must care about the ways in which the body is radically non-Cartesian, that is, the way in which it can talk to itself without having to retreat into a state prior to any body or world. Knowledge in the Cartesian view only begins after we have eliminated all physical determinants through a process of doubt. The view I am associating with "proprioception" has been taken by a variety of thinkers, including Maurice Merleau-Ponty, J.J. Gibson, and most recently Andy Clark: "the body is a great system of reasons, a war and a peace ..." (Nietzsche, Thus Spoke Zarathustra). The word itself was first used in this illustrative sense, rather than as the sole concern of physical therapists, by the poets Ezra Pound and Charles Olson. Olson was dean of Black Mountain College, author of Call me Ishmael, The Maximus Poems, and many other pieces of "projective verse," as well as the prose collection entitled Human Universe, which includes "The Resistance," in which our theme of the primacy of the body is sounded with ethical overtones -- quite reminiscent of what Foucault expressed a few years later: the last threshold of resistance I have is my body. Ethico-political norms can only be evaluated in accordance with what is tolerable or intolerable for my body. But to return to proprioception, here is how Olson defines it:

The data of depth sensibility / The 'body' of us as object which spontaneously (...) produces experience of, "Depth," viz. SENSIBILITY WITHIN THE ORGANISM BY MOVEMENT OF ITS OWN TISSUES.


He speaks of the body as a cavity, an "interior empty place filled with 'organs'? for 'functions'?", and explains, in very Germanic accents, that it "removes the false opposition of 'consciousness'" (p. 182). Andy Clark, neurophilosopher from Washington University, gives a shorter definition of proprioception as "the inner sense that tells you how your body is located in space" (Being There, p. 22).

Here, the word 'proprioception' will serve as a short-hand designation for the priority of dynamic embodied activity over isolated 'mental' and 'physical' regions. Rather than asking "How can a brain accomplish reasoning?", the question becomes "How can a brain have experiences?", that is, "What is it like for a brain to be embodied?".

Rodney Brooks' Artificial Life movement at MIT has produced odd mobile robots called Creatures. We can learn from these creatures and the cockroach robots of Beer and Chiel described in Clark's Being There that cognitivist architectures moved too far away from biological inspirations, including the capacity of biological cognition to adjust to new environments. As the brain does not produce output in the way traditional machines do ("the principal activities of brains are making changes in themselves"), it is self-modifying, unlike the standard idea of an organ which exists to represent external states. Rather, the world is itself present in processes of self-modification. For New Roboticists, this means that representation is the wrong unit of abstraction in building the bulkiest parts of intelligent systems, as Rodney Brooks puts it. For us, this will become a crucial part of the standpoint called "proprioception." Even a classic neuroanatomical paper such as Lettvin et al.'s "What the Frog's Eye Tells the Frog's Brain" already states that "the eye speaks to the brain in a language already highly organized and interpreted, instead of transmitting some more or less accurate copy of the distribution of light on the receptors" (p. 1950). The information circulating in the brain is not reducible to some raw features of the external world which are then turned into faithful copies or representations. The brain is always already a part of the physical (bodily) and social world. Let's turn for a moment to the brain's skillful behavior in the world.

Biological agents always know very little at a time; they
infer a great deal (famous examples include reaching for a cup, or assuming what the trajectory of a ball coming towards us will be, rather than making the actual calculation in our head). J.J. Gibson's notion of "affordances" implies that an object contains an infinite amount of information, since it is different for each perceiving individual; Walter Freeman speaks of humans as finite-state devices, and often emphasizes our incapacity to comprehend infinity. Conscious mental states possess intentionality, the faculty of being 'about' something, and thus the substrate of these states, the brain, is more than an organ for rapidly initiating the next move in a real-world situation, as Clark would put it, just as the lived time and historical nature of an organism is more than the sum of its adaptive responses. Skill is a revealing case of how we deal with challenges from the environment, mobilizing our resources in real time. It escapes some difficulties of the model of thinking as the manipulation of discrete quantities of information by means of symbols, both because "the world is its own best representation" (Rodney Brooks, quoted by Clark, p. 46), i.e., it bypasses a system of representations for the sake of speed, and because our brain has the talent for making use of the environment, "'piggy-backing on reliable environmental properties" (ibid., p. 45), for the sake of economy. "Scaffolding" is one of the vehicles humans employ, so that language, culture and institutions empower cognitions (Clark, pp. 21, 87). However, the human brain does integrate information. Proprioception itself, e.g. one's hand reaching for an object without having to calculate all variables of the trajectory (an example used both by McClelland et al. and Clark, p. 22), is the manipulation, re-shaping, and taking in of information. The brain is not the "central planner" but possesses a "scaffolding" which is inseparable from the external world.
II.


If proprioception is a code word for the way in which our brain is inextricably part of our body, which in turn is a dynamic sensory system interacting with the world and producing a 'customized' environment (von Uexk¸ll's
Umwelt, the specific environment which belongs uniquely to each species: the world of the fly, of the bat, of the human ...), what about consciousness? It is quite possible to maintain that consciousness is a biological property of the brain (in humans and certain other animals) without believing it has to be naturalized, forced into categories we would use strictly for the natural world (physical, spatio-temporal, causally closed). Precisely, there is no need to naturalize it, Searle would say: it already is natural. This would be like "seeking to electrify electricity." But notice the problem: if consciousness is indeed such a natural entity, then it can be studied like any other biological process. Yet for an organism to possess mental functions, it needs to be able to distinguish representational states from the objects stated, doesn't it? Isn't there a level of performance which is distinct from the strictly biological one? Or is this a way of falling back into a representational understanding of the way the mind works? The physical substrate of consciousness is entirely amenable to (neuro)biological study, let's not deny it; but is consciousness itself also thereby amenable to the same study? "Wait!", you'll say: "weren't you denying naÔve idealism and its claims to an autonomy of consciousness, as a kind of mysterious substance that hovers above and around the physical, causal world? Are you saying that thought or behavior can occur independently of physico-chemical changes in the nervous system?" No. The point is to emphasize that the brain is context-bound by the physical (bodily) and social worlds within which it finds itself. "Further problem!", you say. "If we claim that knowledge is socially constructed, and that the world as made by humans is actually the only world to which we have access as humans (barring possibilities of an artistic 'investigation' of sub-rational, perceptual zones, such as in Op Art), do we then toss biological explanations (explanations appealing to the features of our organic life) out the window?" No: symbolic activity is a constitutive feature of human embodied minds: "we may be even more symbolic animals than we are visual animals," says the psychologist Michael Posner.

If we say:
the mind is embodied, perception is also proprioception, we are saying that the texture of the physical world is an irreducible component of brain-mind activity. It cannot be "reduced" in the sense of chemical purification to a pure consciousness; and it cannot be "reduced" to a more physical level in which action itself is not present. (Speaking philosophically, at least, in actual neuroscience, a specialist of perception like William Newsome can be agnostic about the existence of Barlowian 'grandmother' neurons, i.e. the possibility of identifying isolated neurons as "critical signaling units [which] govern performance," emphasize distributed cognition and redundancy, but nonetheless maintain that cognition is fundamentally "computations executed in parallel and in sequence within real neural pathways," such that cognition can be explained by studying these computations. But in a remarkable turnaround, after having given us as good a version of computational 'reductionism' as we can find, Newsome says that the monkey's internal experience is crucial for understanding the nervous system functions. It is not good enough to provide an external account relying on causes and effects in central neural pathways.)

Let's say this another way: it is extremely valuable to prove that a computer can fulfill various key tasks which we previously thought were the exclusive prerogative of our minds. Our minds are, among other things, calculators and chess-playing machines. To demonstrate that computers
can do what we can do in this sense is an admirable project of demystification -- one might even say, of secularization ! -- but it completely abstracts from the biological embodiment of these algorithmic processes. I enjoy declaring that "a cruise missile has intentionality," but in fact the truly exciting statement is that the mind is embodied, part of our organic functioning just as much as eating, digesting, and other such functions. To declare that consciousness is an organic function like any other does, however, run into certain problems when it comes to ontology. Consciousness is always consciousness of something; if we say that hearts, livers, and other organs are also possessed of 'information about X' then we would have to grant them consciousness (as David Chalmers does in The Conscious Mind (Oxford, 1996), a book ably dispatched by Searle). To justify the way in which we relate to this 'aboutness' of consciousness we can't just provide biological explanations. However, this does not affect the position of 'embodied cognition' or 'incarnation' (here termed "proprioception"), since the debate on the existence of consciousness is an ontological one, unlike our descriptive or methodological stance. When the issue is ontology, we can understand why Dennett finds it necessary to explain consciousness by means of properties which are not themselves 'conscious': "has liquidity been explained away by the physicists because (...) they don't attribute liquidity to anything at the atomic level? The physicists have left out the wetness, and I've left out the qualia." (Qualia being all that is subjective about my experience, and yet somehow 'real': classic examples are the experience of color, and pain.)

For Searle, there still
is consciousness, and he would find it silly to deny its existence. For Dennett, what is true is what can be scientifically investigated, and the latter investigations can only take place from the third-person perspective. What is only real to me cannot be real for science. Searle's response to this amounts to a digest of phenomenological investigations into the reality of consciousness, a.k.a. "intentionality," for the past century. In the realm of the mental, he will say, appearances are what is real. (Not just any appearances, but my appearances.) We seem far from proprioception now, and indeed Searle is at best a fellow-traveller in our inquiry: we agree as to the priority of the biological, and we agree that human reality cannot be divorced from the natural world, but for the topic of how 'incarnation' provides a more workable model of understanding mind, brain and body in interaction, Searle has little to say. The most influential book here, which still repays some study, is Hubert Dreyfus' What Computers Can't Do.


II
.

If perception includes sensory-motor acts in such a way as to include our body itself, conversely, we must think of the body so as not to exclude its subjectivity, what Merleau-Ponty, in the Phenomenology of Perception, called one's "corps propre," the sense that one's body is one's own. My lived history as an active individual, as an organism with a certain intake of meat and nicotine and alcohol within specific cultural boundaries, is not merely the development of the set of frozen boundary conditions which comprise my individuality physically, from electrons and protons to atoms to molecules to my genetic make-up. It also includes what Searle calls a Background, a "scaffolding" which is partly tied to my own perceptual make-up (Gibson's "affordances," more or less). My experience has accumulated to form part of my cognitive horizon. Clark adds one twist on this: "the constraints of evolutionary holism" (p. 89), i.e., the way in which I evolve as a series of responses to environmental challenges. One can push this a final step further and say that biologically speaking, an organism is historical inasmuch as it cannot be fully understood in abstraction from its 'prehistory', including its heredity. The history of my organism is not just the sum of the situations it faced and the solutions it found through adaptation (or skill). This is still a strictly spatial understanding, and the organism is truly a time-bound entity, with what the cognitive scientists call "time-critical behavior." As Simondon puts it, each organic act presents a temporal structure which is adequate to a given situation.

Time to conclude. The body which I experience is always my body, regardless of prostheses, hallucinations, etc.; experience is always my experience. In proprioception, I have an inner sense of how the various parts of my body 'communicate' among one other (cf. neurological interest in "body images," the well-known phenomenon of 'phantom limbs', etc.), but also a projection towards the physical world around me, all at once. To understand what it means to think, to compute information, to have a trillion neurons firing in my brain, to have consciousness either at a 40 Hz frequency (F. Crick) or at the quantum level (Penrose), one must really understand this: it's all about bodily sensation. Earlier, I had praised the computational model for having demystified the 'sanctum' of brain and cognition; now I can add Feuerbach's immortal words: "The sense of touch is atheist from birth": our faculty of sensing, of which touch is a particular case, constructs the world for itself. The world is not pre-given by a divinity, or by a rational re-construction, a 'set of instructions' in the current vocabulary. "Vision and touch are not reason's 'raw material', its substrate, but rather the essence of the fully realized human relation to reality." Call it what you will: touch, sensation, embodiment, incarnation; the brain in its bodily context "creates chaotic activity to make sense of the world" (in Freeman and Skarda's phrase), and thus it makes the world its own.