Tuesday, September 1, 2009

Extended Cognition and Personal Fission

This post deals with implications Ex-Cog may have for the philosophical topic of Personal Identity.

Just suppose that there is a kind of close causal coupling (between a cognizing system and its environment) which does issue forth in the constitution of a cognizing system encompassing the coupled environment.

Say cognizing system C1 engages in this sort of coupling with its environment, thereby causing a cognizing system C2 to exist.

C2 has parts, and one of those parts seems to be, simply, C1. The system C1, after all, doesn't cease to exist just because it has coupled with its environment to form C2. Rather, C1 persists, as part of C2.

The physical system C1 persists as part of C2. But suppose that, at least before the coupling, C1 is a person. Call her Carla. There are now two cognizing systems—C1 and C2—which seem to have a prima facie case to be the continuation of Carla. For example, should C2 cease to exist, this would not seem to be a case of Carla's destruction. Rather, Carla continues, in that case, as C1. This is a prima facie case that C1 is the proper continuation of Carla and not C2. Yet, who Carla is seems to have much to do with how Carla cognizes. And once C1 couples with its environment to form C2, it seems as though C2's cognitions have the better claim to be continuations of Carla's own cognitions. For what C2 is trying to cognize its way through is what Carla is trying to cognize her way through. She does this (per the hypothesis stated at the outset of this post) by causing C1 to couple with its environment to form the cognitive system C2.

(The case for C2's being Carla's continuation may seem to assume a Psychologistic account of Personal Identity. I prefer such an account, but I don't want to presume it. In fact, similar arguments for C2's status as Carla-continuer can be built on other accounts of Personal Identity. For the kind of close causal coupling that is supposed to support the formation of an extended mind would seem also to support the formation of an extended cognizing body. Insert difficult discussions about biofunction here, but I think it's clear how, given the Ex-Cog picture, the physical boundary between body and environment is just as indistinct as that between mind and environment. Indeed, the body/environmental blurriness is probably more obvious than the mind/environment blurriness.)

A lot has been skipped over here, but suppose I'm right that both C1 and C2 can lay an equally valid claim to be Carla's continuation. In that case, an interesting situation arises. We have here a case of personal fission—that strange sort of event usually found only in philosophical thought experiments and science fiction stories. Two objects lay equal claim to being the same person as some past personage.

I'm pointing this out because I find it interesting—and kind of exciting—to think that, if the extended mind hypothesis is true, it turns out personal fissions aren't a hypothetical logical possibility that can serve only to jog our rarified philosophical intuitions. Rather, it turns out personal fissions happen regularly, as a matter of course, here on the ground. Our data for working through problems in Personal Identity don't have to be intuitions about what we would or should say given the actualization of various improbable fantasies. Rather, we can work through Personal Identity issues by looking at what people actually do, in the real world, in their day to day persistence as cognitive agents.


  1. I love that you're applying the principles of extended cognition to the idea of personal identity. And I agree we can learn a lot about personal identity issues "by looking at what people actually do, in the real world, in their day to day persistence as cognitive agents".

    In fact, I actually think you can take that reasoning much further than you're doing here. You start off by supposing that at some point in time, "C1 engages in [close causal] coupling with its environment, thereby causing a cognizing system C2 to exist." I'd ask, when, and in what sense, does any "cognizing system" (such as Carla or any other person) exist on its own *without* a close causal coupling with its/his/her environment? You refer to a time "before the coupling." For me this paints a picture of C1/Carla as an utterly disembodied abstraction, floating in (and not even interacting with) the ether. If we suppose, instead, that any cognizing system is *always* closely coupled with a wider external (physical, social, linguistic, etc.) environment, we can start to question the whole idea of personal identity as something we can ever adequately define without reference to the wider context(s) in which that system is embedded.

    I'll be curious to see how this sounds to you.

  2. Thanks for the comments, Boxed Beetle. In my post I didn't mean to imply that there's ever a time when a cognitive agent isn't involved in some "close causal coupling" or other. Moreover, it's not implausible to me to think that a cognitive agent might be involved in several different couplings at the same time. (I'm not claiming either that we're always coupled to something, or that it is useful to think of us as involved in several couplings at once--I'm just saying I'm _not_ opposed to these possibilities. It will take a lot more thought--and empirical research--to determine whether either of these situations obtains.)

    In any case, you're right--the thoughts in my post naturally lead in the direction of a claim that there's no formulation of personal identity that doesn't refer to "the wider context(s) in which [a] system is embedded." This in itself isn't too remarkable, though. Narrative and Constitution views of identity, for example, are probably best construed as having taken this idea on board already. As for Somatic accounts, it's almost a truism that you can't describe a body without reference to its environment. You need to have some way of distinguishing it from other bodies, after all. Psychological accounts? Assuming _some_ degree of externalism, most people already agree that psychological states can't be described without reference to the environment they're embedded in.

    So in a way, the idea you're driving at seems unsurprising. But I wonder if I've sold your idea short.