I first heard about Francisco Varela because my college friend and sometimes roommate was a big fan of his early work on autopoesis. While JL was a symbolic systems major at the time, shortly after graduation his study shifted to the paradoxical field of military intelligence, so I'm not sure he ever got a chance to read Varela's collaboration with Thompson and Rosch; its initial publication would have been a recent and controversial addition to the field of cognitive science during the time when we were at The Tree. I'm also not sure that he would have understood or appreciated the direction that Varela's work took with The Embodied Mind; I know I certainly wouldn't have been capable of appreciating a book that was so far ahead of its time in so many ways. As Jon Kabat-Zinn points out in his foreword to the revised edition of 2016, applying concepts drawn from Buddhist meditation practices to the scientific study of the mind was a bit out there back in 1991. And if a synthesis of this sort is now more commonplace, apparently this particular book deserves a significant amount of credit for opening that floodgate.
Though it's significantly better written and argued, the overarching thesis of The Embodied Mind is similar to the other science-dharma crossover book we explored in the individuation project -- Joanna Macy's Mutual Causality in Buddhism and General Systems Theory. The subject of cognition gets embodied, and its world enacted, when both arises out of pure emptiness as a self-creating feedback loop. We've seen a number of metaphors that try to capture the insight that emptiness = dependent (co-)origination, that the individual is never given but created as both product of and input to an ongoing process of individuation. We've talked repeatedly about the vortex. Simondon preferred to think of it broadly as crystallization. Macy simply called it mutual causality. In every case, the core idea is the way subject and object arise simultaneously from the figure-ground splitting of a groundless non-dual emptiness. Varela et. al. will take embodied cognition as the paradigm for this process of emptiness forming itself. In other words, while they don't put it quite this way, they suggest that cognition is a form of ontogenesis. Cognition doesn't represent the world, it creates more of it. Or, given that they always discuss the operation of cognition mostly in the context of an existing organism, perhaps we should borrow from Simondon's terminology and say that cognition is a form of onto-transformation that we can study through "allagmatics".
One understands that the project of allagmatics, which is already formulated in ILFI and MEOT in passages where Simondon enters into a dialogue with cybernetics, brings the philosophical project in close connection with the idea of a science (see ILFI 561), even if this new philosophical science is by definition transversal and unifying; whereas each positive science is a science of generic structures, allagmatics is the science of genetic operations: 'the operation is that which makes a structure appear, or that which modifies a structure' (ILFI, 559). (Fifty Key Terms in the Works of GilbertSimondon, 204)
To their great credit, the authors take this notion of cognition as ontological production and transformation all the way to its conclusion, and apply it to their own theory. So while much of the book is devoted to a critique of representationalist cognitive science that relies on the metaphor of computation, and some of it is occupied with their non-representational theory of enaction, all this content is sandwiched between two extended discussions of how their own representation of a theory of cognition might be used to actually transform our experience. Because their ultimate goal is not to come up with a more accurate theory of cognition, but to use the act of theorizing about cognition to change how we live. Few thinkers make it to this deeply philosophical point that seems to me to be the heart of the matter. Why is it that we search for a science of the brain? Or why do we philosophize or meditate? Are any these activities really about knowing reality, representing it, "as it really is"? Or are they more fundamentally about expanding the possibilities of our experience? Why, after all, do we even think? Is it not to have an experience that provides us with a doorway to even more experience? Thinking should be an experience that transforms us, not just another game of theoretical Tetris meant to move us up the high scorer list.
So the book begins and ends with philosophical sections devoted to the ground and the groundless. Their departing ground is the way the philosophy of Merleau-Ponty emphasized the body as not merely one object among others but as a lived experience. As the book's title suggests, the most important thing to realize about minds is that they are embodied. Which is to say that thinking is a lived experience of interacting with a world. Thus, we reflect on the world not from a position outside it, but from within, as a part of the world that reflects on itself. Merleau-Ponty points to this fundamental circularity of thinking about thinking in a passage they quote approvingly:
The world is inseparable from the subject, but from a subject which is nothing but a project of the world, and the subject is inseparable from the world, but from a world which the subject itself projects. (TEM, 3)
As a result, any theory of cognition that attempts to take up a perspective outside the world, as if it could occupy the 'view from nowhere', is bound to fail. It may produce a theory of how the brain works, but it will not direct us towards the experience of contemplating the theory. And of course, the same could be said for any philosophical theory, even one as circular as Merleau-Ponty's phenomenology. While phenomenology famously hoped to go to "the things themselves", there's no way to get around the fact that philosophy is ultimately always a form of literature.
In Merleau-Ponty's view, both science and phenomenology explicated our concrete, embodied existence in a manner that was always after the fact. It attempted to grasp the immediacy of our unreflective experience and tried to give voice to it in conscious reflection. But precisely by being a theoretical activity after the fact, it could not recapture the richness of experience; it could be only a discourse about that experience. Merleau-Ponty admitted this in his own way by saying that his task was infinite. (TEM, 19)
It's in this context that the authors turn to Buddhism. Because while Buddhism certainly contains both religious and philosophical theories of experience, it also provides a set of practices designed to produce experiences. This of course is what Western philosophy has long lost -- that there is a practice of philosophy which seeks a know how rather than a know that. As Nietzsche was fond of pointing out, we have lost the sense that philosophy should consist of more than bloodless theorizing but is a discipline meant for living. So the authors would like to invoke Buddhism as a practice for investigating experience. As an open-ended practice, it employs concepts, but departs from and returns to the actual living experience of the body, even if it may transform this experience in the process (pg. 30). Thus, it appears to pragmatically deliver on the theoretical slogan: "to the things themselves".
I say "appears to" because this is the place to insert some caveats that aren't present in the first edition of the book. While the authors never assert this exactly, one could sometimes get the impression they are claiming that meditation provides an empirical method for a "first person science of mind". That is, one might believe that we meditate to "see things as they are", and that therefore the phenomena we experience when we meditate are more real, more true, and more foundational than what we find in philosophical or scientific investigation. And in fact, there are many folks (some with much more meditative experience than I have) that describe the practice in this way. Needless to say, I don't see things this way. Meditation regularly supplies all kinds of unique and interesting experiences, including the experience of feeling like you see a deep and important truth, the experience otherwise known as insight. But to accept any of these experiences as the final, real, ultimate truth of the way things are is to enter back into precisely the realm of theory that we were trying to use meditation to escape. What we want to say is simply that meditation shows us more reality, and that these experiences can transform our way of approaching life so that we suffer a lot less because we experience a lot more. That is, we want meditation to open experience, not close it off. Trying to prohibit or limit or ignore certain experience is the very essence of ignorance. Thankfully, we don't have to spend much time critiquing the places in the book where the authors slip towards this kind of thinking because one of the authors already beat us to it. Evan Thompson is not a Buddhist, and his introduction to the revised edition provides a lucid description of the ways the book occasionally falls prey to the problematic assumptions of what he calls "Buddhist modernism". I completely agree with his caveat that meditation cannot provide a new, firmer, higher resolution, empirical ground to replace science's simple empiricism. If it contributes to our knowledge, it can only be in Deleuze's sense of a transcendental empiricism, inseparable from pluralism.
I have always felt that I am an empiricist, that is, a pluralist. But what does this equivalence between empiricism and pluralism mean? It derives from the two characteristics by which Whitehead defined empiricism: the abstract does not explain, but must itself be explained; and the aim is not to rediscover the eternal or the universal, but to find the conditions under which something new is produced (creativeness). (Dialogues, preface)
While this caveat is important, in truth, the middle sections of the book usually deliver fairly well on the goal of injecting some pluralism into theories of cognition. While the authors critique both cognitivism and connectionism as representational theories of cognition, they clearly appreciate the way these research agendas have, in fact, contributed a lot to our understanding of how the brain works. And their own positive contributions, their theories of cognition as "enaction" and evolution as "natural drift", are both open-ended theories that don't explain how everything must, lawfully, be, so much as draw our attention to the organism's (or the biosphere's) inherent creative potential.
After all this preamble, let's begin to address the content with their discussion of cognitivism in section 2. Cognitivism is the idea that the brain is simply a computational device meant to represent the features of an external world so that we can manipulate it effectively. It's the type of thinking that led Stanford to call its department of mind studies, "Symbolic Systems". While from the author's perspective, this is clearly an inadequate dualistic view, what they what they point out here is that the cognitivist program seems to simultaneously posit and refute a central self. On the one hand, the whole idea that the brain represents things seems to depend on the notion that it represents features of an external world to an internal subject. Something only becomes a symbol or representation when it is of something for somebody. Representation contains coded information about the world and, as Simondon pointed out, the existence of information requires a distinct sender and receiver. Representation entails more than a more correspondence or causal connection. Perfectly causal process that don't mean something are not called information but noise. Of course, one person's noise is another's information, but this fact is precisely what makes us realize that information is not simply 'out there'. In short, it seems that any theory of cognition as representation requires the existence of a cognizing subject. On the the other hand, in practice the cognitivist program treats the embodiment of this representing subject as a pure computational machine. Representations are constructed through neural computations that extract features of the world. This information is then passed through the brain and combined with other information about internal states and so forth. And the final result of this complex computation is a pattern of motor neuron firing that actuates a response to the information. The subjective self who would 'read' this information and decide what it all means doesn't appear anywhere in the chain. All the computations used to instantiate a representational model are simply the rule based pushing around of meaningless bits. As a result, the cognitivist paradigm starts off assuming there must be a subject, and then demonstrating that there can't be, very much in the same way that Daniel Dennett sets out to explain consciousness, but effectively ends up explaining it away. Conscious experience is required for the whole scheme to operate, but then doesn't seem to have any particular functional role or point within it.
Interestingly though, in what I would call a prime example of their pluralism, the authors see in this contradiction not merely as a critique of the inadequacies of cognitivism, but the beginnings of an understanding that perhaps a centralized self is not required for experience after all. That is, cognitivists like Dennett think they have explained qualitative experience away, but looked at from a wider vantage point, all they have succeeded in explaining away is the need for a self to serve as the locus of this experience. In fact, the authors explicitly analogize the cognitivist deconstruction of cognition to the Buddhist demonstration that none of the five aggregates contain a self. With the Buddhists, however, emptying the aggregates of a self doesn't empty them of lived experience and prove it is a mere illusion. No-self is not equivalent to non-existence. So while the cognitivist struggles with the paradox of understanding how their own, merely computational, brain can come up with such a theory of itself, the Buddhist deconstruction of the self removes nothing essential to our experience or to its own theory, but instead invites us to begin exploring how the sense of self might arise in a fundamentally non-self world. If we begin with experience, with awareness itself, as the given, we sidestep the hard problem entirely. Within this experience we find an experience of our self and of the world, neither of which experiences guarantee anything about their veracity or usefulness beyond their phenomenal existence as experiences. It's only when we begin with the flawed assumption that there must be an objective material world represented by an individual subject that we later discover these two things never seem to reconnect. This materialism disproves itself by revealing a hidden spiritual dimension it implicitly needs to operate.. The paradoxes of the hard problem that threaten to force us to condemn either the objective world or our subjective self as an illusion are entirely an artifact of a poorly posed question. There is no hard problem. But it's a nice part of the authors open-ended and pluralist approach to critique other views of the mind and yet still learn important things from them.
In section 3, the authors move on to a discussion of the brand of cognitive science in ascendancy back in my day -- connectionism. Connectionism doesn't completely dispense with the idea that the brain is a computational device aimed at representing features of the world, but it does move away from comparing its architecture to the serial symbolic processing we see in our von Neumann machines. Instead, connectionism approaches cognition with the more biologically realistic model of the neural network. Here the idea is that the global behavior and 'cognition' of the system emerges bottom-up from simple, mindless interactions amongst its parts. These stable attractors can of course still be shaped by an environment, but they are more than mere arbitrarily pre-programmed symbols that represent this environment, since they have a self-organizing internal dynamic of their own. While it operates at a much higher than neural level of abstraction, the authors use Marvin Minsky's "society of mind" metaphor to illustrate the idea that our cognition consists of a complicated interaction of simple agents. They then explicitly compare this to the Buddhist idea of the 12 links of dependent origination, treating each link as if it were a sort of agent characterized by a particular pattern of action. The interaction among these links, the way they chain together or form feedback loops, serves as an explanation of why we experience a sense of self when all that's really happening is the casual interaction of simple parts. In Minsky's analogy though, we are forever condemned to experience what is really a society as a stable and enduring self. We can't help believing in the illusion. For the Buddhist, the hope is that deconstructing our experience as a product of the ongoing conditioning of the 12 links, we are eventually able to use our theory to change our very experience of the self. While I don't find the analogy between Minsky's sub-personal agents and and the 12 links particularly compelling or useful, I do sympathize with their observation that there must be something lacking in any theory of cognition that doesn't account for the fact that it can change. For Minsky, as we saw for Dennett, hypothesizing that our sense of a stable interior self is an illusion doesn't actually change how either of these thinkers act. Despite saying that it theoretically can't be done, they continue to talk and write as if they have the same self as always. This leads to a disconnect between theory and experience that threatens to make all cognitive theory completely meaningless in experiential terms. In short, it leads to a threat of nihilism. If deconstructing experience 'scientifically' does not change it, if knowing ourselves 'objectively' has no value for our life, then it seems perhaps it makes no difference at all what we think or believe or take ourselves to be. As Nietzsche would say, the highest value of truth devalues itself; if the truth doesn't do anything, what's the point of it? So while connectionism is an improvement on cognitivism in scientific terms, and illustrates the same lack of self we saw before, it still leaves us with a gap between a representational theory and a lived experience.
It's really only at this point (section 4) that the authors come to the heart of the matter I pointed out earlier. If cognition is not representational, if it's goal is not to mirror certain pre-given features of the world either through explicit cognitivist symbols or through implicit connectionist attractors, then it must be "enactive". Instead of mirroring the world, it creates it at the same time that it creates itself. The subject gets a body at the same time that the world gets an action -- embodiment and enaction are really two sides of the same non-dual coin.
This shift requires that we move away from the idea of the world as independent and extrinsic to the idea of a world as inseparable from the structure of these processes of self-modification. This change in stance does not express a mere philosophical preference; it reflects the necessity of understanding cognitive systems not on the basis of their input and output relationships but by their operational closure. A system that has operational closure is one in which the results of its processes are those processes themselves. The notion of operational closure is thus a way of specifying classes of processes that, in their very operation, tum back upon themselves to form autonomous networks. Such networks do not fall into the class of systems defined by external mechanisms of control (heteronomy) but rather into the class of systems defined by internal mechanisms of self organization (autonomy). The key point is that such systems do not operate by representation. Instead of representing an independent world, they enact a world as a domain of distinctions that is inseparable from the structure embodied by the cognitive system. (TEM, 139)
Now, of course, we're really sliding into the philosophical deep end where both self and world disappear. It's one thing for science to examine the brain and fail to find John Malkovich lurking inside. Especially when this doesn't result in any change in our sense of self or in how we live. Because science and most of analytic philosophy implicitly and subconsciously bracket the inquirer as categorically separate from the object of inquiry, deconstructing the self (by hypothesis) isn't any different from deconstructing anything else. But once you begin to question the substantial reality of the subjective world, the objective world also begins to fall apart, and this throws science into crisis. After all, if there's not a substantial subject to know the immutable laws of physics, then how can we assume such a law governed world even exists? And if there's no objective world, then who are we scientists studying it, amirite? We have a sense that it's dangerous to see the solid ground slip from our feet in this way, one the authors diagnose as a "Cartesian anxiety". We worry that if we let go of the world, and then the self, and then the world again, we'll come unmoored and be cast adrift on the sea of nihilism. Which is of course why we tend to avoid this type of thinking at all costs. We really want to have some stable reference point we can take for granted. For the authors, this amounts to the core teaching of the Buddha -- grasping after the ground, regardless of whether we affirm or deny it, leads to suffering. In this case the problem seems to stem directly from the idea that our thinking might actually require us to change something about how we live. If the cognitive scientist really believed their own theory about the self, they would be forced to act differently. And we know quite well that change is the end of the world.
They use a detailed discussion of color vision to illustrate the notion that the experience of color is something that happens between the subject and the world. It is neither an objective property that can be reduced to the wavelengths of reflected light, nor a purely subjective property that happens in each individual's interior mental space. Instead, the perception of color depends on the possibilities of interaction between the two. While I won't go into the details, it seems to me they make a strong case that the known facts about the experience of color only make sense if we see it as constituted by the feedback loop between perception and action. Color is not 'out there' -- we perceive its reality only because it affords some possibility for action in the world. At the same time, color is not simply 'in here' as an arbitrary category -- we are able to act on the basis of it only because there is indeed something there to perceive. The fact that this creates a circle of feedback or chicken and egg problem is the whole point.
We can now give a preliminary formulation of what we mean by enaction. In a nutshell, the enactive approach consists of two points: (1) perception consists in perceptually guided action and (2) cognitive structures emerge from the recurrent sensorimotor patterns that enable action to be perceptually guided. (TEM, 173)
In short, what the world is depends on how we look at it and what we want to do with it; and how we look at it and what we hope to do with it depends on what the world is. Cognition is not the representation of a pre-given world by a pre-given subject. It is the process by which the world and the self come to mutually specify one another. It is the inevitably historical process of feedback by which these two structure one another.
But what, ultimately, do we really want to do with the world? The authors take their theory of enaction a step further by considering a neo-Darwinian objection to it. Sure, we might say, each organism perceives a different world based on their structures and capacities for acting in it, and these capacities are defined on the basis of past perceptions and their associated cycles of action. But the whole point of all this perceiving and acting is survival. And the way to survive is to be 'the fittest' -- to optimally fit your perceptions to the world so as to maximize your survival chances. It's easy to see that the motivation behind this objection is the evolutionist notion that the real point of life is to survive and its one real action is to reproduce. In other words, in spite of how evolutionary theorists are constantly pointing out that evolution has no goal, they surreptitiously convert the ideal of survival into a goal. With this finality, it promises to restore the ground under our feet -- we are all just machines 'designed' for maximizing reproduction, and most of our features, including our cognition, are simple consequences of this fact. So while we all cognize differently, the goal of all cognition is the same -- optimal manipulation of the world as judged by reproductive success.
I think it's also easy to see what's wrong with this objection at a high level. It's completely dualistic. It assumes that there are separate and solid selves out there to be reproduced through some sort of optimal adaptation to a pre-defined environment. But these are precisely the reference points we've been calling into question. If there's no inherent self, whose goal is it to reproduce? And if the world isn't inherently and necessarily any particular way at all, what is there to adapt to? Finally, if the unit of reproduction isn't clear, and the raw material from which it draws is not uniquely defined, then what does optimal fit even mean in this context? If the 'goal' of life were for genes to maximally reproduce themselves, the earth would just be one big DNA crystal by now. This would appear to be an 'optimal fit' in the deepest sense that all of the available material world would be converted into self. The fact that life looks absolutely nothing like this should tip us off that something is wrong with the logic here.
The authors of course produce a much more detailed criticism of neo-Darwinian adaptationism than my series of rhetorical questions and thought experiments. But their overarching point is that the theory of natural selection doesn't really explain what happens, it only explains what doesn't. It's true, systems that cannot survive, don't. But this tells us almost nothing about the diversity of systems that can survive. Survival or adaptation, in other words, is not a strong constraint but a quite weak one. It doesn't winnow life down to a single optimal solution, it simply weeds out the non-viable solutions to leave us with what remains an incredible richness of possible ways of making a living. Thus instead of natural selection, the authors talk about evolution as "natural drift". They don't mean the drift of genetic mutation that people discuss in evolutionary biology, and which is in some sense 'unnatural' because it comes from the chance invasion of life's carefully controlled mechanism. They mean the immanent natural ability of biological matter to spontaneously form new patterns. Kind of like an evolutionary version of Uexküll, these patterns give body to an organism at the same time as they enact a world as a particular slice of nature's chaos. This deepens or doubles their theory of enactive cognition because it explains how the specific biological structures that support an open-ended enaction which transforms the world are themselves constructed through what amounts to nature's own enactive cognition. It's empty turtles all the way down.
As I mentioned earlier, the final section of the book returns to purely philosophical considerations. If the circularity of Merleau-Ponty's phenomenology was the departing ground, with the nihil-ism of Keiji Nishitani we arrive at the groundless. I'm not familiar with Nishitani, but even their short discussion of his merger of Western metaphysics (he was a student of Heidegger's) with the Zen understanding of emptiness was enough to get him on my future reading list. We've covered all the basic ideas already, so I won't repeat the story here. But be on the lookout for the forthcoming JPiPE blog!
No comments:
Post a Comment