Thinking more about the question I posed earlier with respect to Spinoza and Mandel’s gorgeous follow-up I find myself returning to themes I discussed a month or so ago: the evolution of freedom. In discussions of empiricism Deleuze repeatedly emphasizes that the central tenant of empiricism is that abstractions do not explain, but must be explained. Deleuze is a transcendental philosopher in the tradition of Kant, yet he differs from Kant insofar as he holds that things such as categories and forms of time and space must be given a genetic account. We cannot, Deleuze contends, treat these as transcendental givens, but must instead provide a genetic or developmental account of how these things come to be. This is the secret of the theme of aesthetics and learning that pervades Difference and Repetition.
When Deleuze refers to aesthetics, he is not referring to art or the appreciation of beauty per se (though that too), but rather aesthetics in the sense of Kant in the first Critique, where there are transcendental, a priori givens of sensibility. Time and space, for Kant, are a priori forms of sensibility in the dual sense that 1) all phenomena presuppose space and time, 2) time and space are not sensed through the five senses (they aren’t like tastes, smells, touches, sights, and sounds), and 3) they originate from mind, not world. For Kant, mind structures phenomena in terms of time and space, rather than time and space being properties of things-in-themselves. We thus here get a form of transcendental intuition. Where empirical sensations such as sounds or sights must be received from the world, space and time, according to Kant, originate from mind and structure the sensations of experience. In this way Kant hopes to account for why sciences like geometry and arithmetic are possible a priori or independent of experience. Insofar as mind structures world in this way we don’t have to await sensations of time and space in the way we must await the taste of a cherry doughnut to know what it will taste like.
Deleuze accepts all this. The major difference is that Deleuze requires a genesis of the a priori form of sensibility that accounts for how these structures of experience come into being. This is the origin of the theme of learning in Deleuze. Sensibility is a structure that develops and evolves, generating new a priori possibility, not a static and fixed form that is already there as in the case of Kant. Here Deleuze straddles both developmental psychology and Darwin. From the standpoint of developmental psychology our minds develop at the level of sensibility in such a way that new a priori forms of inference become possible that have, as it were, their own “mathematical” a priori structure. This is what Deleuze is exploring when he explores the work of great artists such as Bacon, Cezanne, Proust, or Kafka. What Deleuze is exploring is the new aesthetic a prioris these artists have invented or introduced to the world. He is doing a transcendental aesthetic in the sense that Kant proposes a transcendental aesthetics in the Critique of Pure Reason, but at the level of artistic creativity. Such a move was already suggested by Kant’s Critique of Judgment where Kant describes aesthetic judgment as intuition in search of a concept (i.e., art creates a new spatio-temporal schematization of experience).
On the other hand, Deleuze’s transcendental empiricism is thoroughly post-Darwinian in the sense that the evolution of a species is not merely the evolution of a new form of life, but also the evolution of a new form of sensibility. The bat, for example, evolves a transcendental form of intuition structured around sonar that has its own logic and organization. Acoustic signatures such as those found in sonar are not a feature of the in-itself, but rather are a way in which the bat minimizes the complexity of its environment such that it selectively relates to that environment. This selective relation to the environment, so brilliantly explored by von Uexkull, has its own transcendental spatio-temporality, its own organizational logic of transcendental signs, that marks the particular way in which bats are open to their Umwelt. In this regard, there is no contradiction between the transcendental empiricism proposed by Joe Hughes and the transcendental empiricism that I propose in Difference and Givenness. Joe adopts the Husserlian-Kantian orientation in his reading of Deleuze, emphasizing epistemology, how an entity grasps or relates to its environments, and questions of synthesis and genesis. Hughes dispenses with categories altogether– the abstract doesn’t explain but is to be explained –instead exploring the manner in which genesis and synthesis produce forms of experience. In this way he is able to equally explain the world of a plant, a bat, a singular human being, etc., in terms of how they reduce the complexity of their environment and bring an Umwelt into being. My difference from Joe lies merely in the fact that I emphasize that the entities that result from these genetic processes are just that, entities. The bee does not merely grasp its Umwelt genetically in a particular way, but also is a particular entity or system. Here, I believe, we’re both on the same page.
Returning, then, to the issue of freedom, I contend that freedom as well must be given a genetic account. Freedom must not be conceived as something we have but as something we acquire, evolve, or develop. Contrary to Rousseau’s thesis where we are born free and everywhere we are in chains, this thesis would argue that we are born in chains and perhaps become free. The issue can be set up in terms of the opposition between heteronomy and autonomy. An autonomous being is a being that is self-directing such that it chooses its own action and is the ground of its own action. A heteronomous being is a being whose action is determined by something external to itself. My car, for example, is heteronomous in the sense that it requires me to steer it, to turn it on, to keep it running, etc. The car is not the origin of its own action. It does not determine itself.
The idea that I’m toying with is that autonomy is not something we’re born with, nor something that we have, but something that we develop in much the same way that the artist cultivates a particular system of percepts. So how would this work? Minimally, in order for a system to become autonomous, that system must have 1) memory, 2) be capable of formulating reasons or grounds for its action, and 3) be self-referential such that the reasons it provides for its actions can become causal grounds for subsequent causal actions. These are conditions for autonomy but do not guarantee autonomy. Rather, freedom or autonomy is something that must be acquired, produced, or created. What is the process by which this takes place?
Let’s agree with the thesis put forward by many cognitive scientists and neurologist where it is argued that we do not act based on reasons, that we have little to no accurate knowledge of why we act as we do, but rather that the reasons we give for our actions are produced retroactively and have nothing to do with the real processes that led us to this or that action. As the neurologists show– through the pathbreaking work of cyberneticists such as Ashby –all thoughts have a certain electric potential signature such that we can detect when a thought is taking place in the brain. The remarkable thing about these measurements of electric potential is that the neurologists are able to detect a rise in electric potential nanoseconds prior to the person becoming aware of the thought they are having. Phenomenologically we experience ourselves as the origin and author of these thoughts but this time lag– not unlike an echo in a cave –suggests that thoughts always occur behind our backs such that we are not the authors of our thought, but are instead authored by our thought.
This would seem to spell the ruin of autonomy insofar as it would entail that we are thought rather than thinkers. Put differently, it would entail that we are heteronomously determined by pre-personal neurological processes, rather than being autonomously self-determining and self-directing agents. Any reasons we give for our action would be illusory rationalizations of why we acted as we did. The real reason that we acted as we did would simply be mechanisms– highly non-linear to be sure –of our brain.
However, things are not as bleak as all that. The solution to the riddle, I suggest, already lies in the “echo-phenomena” of retroactively providing reasons for why we did what we did. The reasons I provide for what I did do not provide reasons for what I did, but can provide reasons for what I will later do. The mistake lies in believing that these reasons were grounds of my previous action (these reasons are, after all, retroactive). What is to be accounted for is why these retroactive reasons reasons are produced. Are these reasons “spandrels” of developmental processes that serve no neuro-cognitive function, or do they have a functional and adaptive purpose in neuro-cognitive development? That is the question. Why do systems such as ourselves, many animals, social systems, and some computers retroactively produce reasons for their actions if these reasons have nothing to do with why those actions were undertaken? That’s the question.
Here’s an example. Suppose I get angry at my friend Joe and strike him while we drinking at a conference together. The neurology tells me that in this moment of striking my dear friend Joe I was a pure mechanism. The reasons I provide for striking him are retroactive rationalizations of a sequence of events that were the result of an actualization of a set of electro-chemical potentiations that were evoked when I was perturbed in a particular way. Because the neurological events of my brain belong to the environment of my thoughts, I know nothing about this sequence of events. As a consequence, later, as I pace back and forth in my hotel room I am filled with remorse– after all, Joe is one of the kindest and friendliest people I’ve ever met and has been a great ally and source of encouragement for my work for years –and I begin to cast about for reasons as to why I struck Joe in this way. Having been trained in psychoanalysis I begin to free associate and conclude that Joe resembles a figure from my childhood that tortured me endlessly in elementary school, tripping me in the hallway, posting “kick me” signs to my back, taunting me mercilessly “Levi is Evil, Levi is Vile!”, etc., etc., etc. This childhood figure was even named “Joe”. I now conclude that somehow my unconscious thought process conflated Joe-1 (the elementary school figure) with Joe-2 (good’ole Joe Hughes), that Joe had said something that had released all these repressed representations and that I hit him accordingly.
Of course, the events that transpired that evening had nothing to do with this narrative. As a marionette, a series of gears and pulleys simply twisted and turned in me and I punched Joe (this scenario is purely fictional, by the way). So why, cognitively, do I produce all these reasons if they were never the actual ground of the event that took place? My suggestion is that this activity of producing reasons functions not to account for past actions that took place, but to develop schema for future action. In producing a set of reasons for why I did what I did, I do not account for why I did what I did– these reasons share no relation to the event that took place or its causal factors –but I do produce a set of cognitive causal factors that can become parameters of my future action. In other words, after developing this purely mythological account of the grounds of my action, the next time I encounter a stimuli from Joe that would set of this sequence of unfortunate events the myth that I’ve concocted sets in, preventing me from doing this again. I tell myself that it is not Joe-2 that I’m angry with but Joe-1 and that Joe-2 is not Joe-1. It matters little whether or not my action indeed had anything to do with Joe-1 and repressed memories. No, all that matters is that these reasons that I retroactively produced take on a causal power within my cognitive system. In this way I gradually become self-determining and free (note how nicely this gels with our intuitions about childhood development and our claims that children are not, in fact autonomous or self-directing). Gradually, through the retroactive production of reasons I create reasons that become grounds or causal factors in subsequent encounters such that I am no longer heteronomously determined but such that these reasons I’ve produced come to guide and determine me (such would also be the case for social systems that become over time self-directing through the formulation of missions, goals, aims, etc).
Let’s return to Spinoza and Mandel’s brilliant analysis. Pace Spinoza I am determined but this does not entail that I cannot form or generate ends over time. As I acquire knowledge of true causes, my beliefs about these causes can come to direct my action. I might, for example, suffer from profound depression. I believe this depression originates form how terrible my life is. I amuse myself my exploring all the ways my life is terrible. Meanwhile my buddy Joe looks askance at me, pointing out a number of ways my life is very nice. While causally reading an article I discover that the absence of omega-3 fatty acids in a diet can generate depression due to the way it affects brain chemistry. I notice that all I eat is barbeque (not really) and that I eat almost nothing in the way of fish. I therefore resolve to begin eating more fish only to witness my depression and its alleged reasons dissipate like a morning fog. Now what’s the point of this example? The point is not that eating fish re-adjusted my brain chemistry such that I was no longer tormented by phantasisms apparently coming from the external world telling me that life is terrible. No. The point is that knowledge or belief that omega-3 fatty acids in my diet could become a causal factor changing my behavior in a variety of ways. The reason is produced retroactively but once it is there it opens the possibility of a self-directing future. If we are self-referential systems with memory and that subsequently produce reasons for our action we can become self-directing. In other words, we can have our mechanical determinism and eat it too.