Monday, January 18, 2010

Supersizing Introspection

I've always enjoyed Andy Clark's work (hence my desire to emulate his drink preferences), but I hadn't ('til now) got around to reading his latest book, Supersizing the Mind. Clark is one the the leading advocates of the view that cognitive processes extend beyond the boundaries of the brain to include aspects of the body and environment. The boundary of skull and skin is no privileged border, such that human cognition can only take place within it. If mental images of Scrabble letters are part of your cognitive process when thinking about your next play, then so also are the actual physical tiles when you manipulate and use them in an analogous way. When you work to create an environment that helps you remember, your knowledge is partly distributed into that environment. The mind is not just in the skull; it is "supersized".

I've been struggling lately to develop a general account of what introspection is. I characterize my view as "pluralist" -- I think a variety of mechanisms drive what are rightly thought of as introspective judgments. It now suddenly dawns on me that what I'm really doing is "supersizing" introspection. Introspective processes -- what are sometimes thought of as the most "inward" things there are -- often include the body and world, and broader aspects of the mind than is generally supposed.

How do I know what emotion I'm in? Do I turn on the inner emotion-scanner mechanism, which then produces the judgment that I'm (say) envious? How do I know my preferences? My imagery? My sensory experience? Philosophical opinion basically divides into two camps: First (probably the mainstream) are those who advocate "detection-after" accounts, according to which I have the experience (or other mental process in question) and once that completes (and maybe also while it continues) a separate scanning process of some sort detects the presence or absence of that state.

Second are those who advocate one or another of a variety of non-detection processes. One example is Alex Byrne, who holds that figuring out whether I believe that P (e.g., whether it will rain tomorrow) involves figuring out whether P is true (that is, whether it really will rain tomorrow -- a fact about the outside world) and then applying a belief formation rule according to which when P is true it is permissible to form the belief that you believe that P. On such a view, we know our beliefs not by introspection, but rather by "extrospection" of the outside world, plus the application of some simple inference rule. Similarly, we might learn about our visual experience by attending not to the visual experience itself but rather to the outward objects that we are seeing. We might learn about our emotions just by attending, proprioceptively, to states of our body. There is no turning in, no self-scanning of the mind, in introspection.

It has always seemed to me that both types of view are partly right: Contra the detection-after views, it seems to me unlikely that introspection is the operation of a simple subpersonal scanning module wholly distinct from the cognitive process that is the target of introspection, and outward-looking processes must be part of the story. Contra the outward-looking views, however, it seems to me that outward-looking processes, too, are only part of the story.

Okay, so how do I know that I'm feeling envious? Partly, I look outward: I notice that I am in the type of situation that is apt to promote envy. Someone has something valuable that I don't have. Maybe I look more carefully at or think more carefully about that thing itself. Partly, perhaps, I notice, proprioceptively, my own physical state -- an arousal of a certain sort. Maybe I notice that I have a visual image of the person suffering a painful death. (How do I know what imagery I have? Well maybe that's by introspection, too, and there will be a pluralist story to tell there also.) Maybe I try turning my thoughts toward what is enviable and not enviable about this person and notice whether my bodily arousal crests and falls. Maybe, in the very labeling of myself as "envious" I partly make it true; I was feeling more diffusedly negative before, and the label crystallizes it. On top of such processes, I see no reason to reject the possibility, and I see several reasons to accept the possibility, that there are subpersonal causal processes (not, necessarily, the operation of dedicated modules) that show some sort of sensitivity directly to the emotional state itself, i.e., that work directly to increase the likelihood of my reaching the judgment that I'm envious, given that I am indeed envious.

I doubt we can usefully carve out some subpart of this multifarious mash-up and say that it, alone, is the "introspective process". I think that introspection, like much of cognition according to Clark, is multi-faceted, partly in short connections in the head, partly in broad interactions in the head, and partly spread out into the body and environment.

This is, of course, not independent of my view that we often get our introspective judgments badly wrong.


John Miedema said...

I really must stop reading your blog. I always walk away with another book that I absolutely must read now. Cheers.

Gary Williams said...


Do you think it's possible that a "mental behavior" like introspection, which involves turning our "Mind's Eye" towards an internal mindspace or conceptual landscape, is really just bootstrapped from our normal behavior of looking at external landscapes? That is, mental acts like introspection are functional analogs of behavioral acts such as perception. This would explain the findings of Lakoff and Johnson, who look at the metaphors of mindedness in terms of being analogs for embodied experience. If language itself bestows us, as Clark puts it, the "cognitive scaffold" of internal epistemic manipulation, then it doesn't seem too far off to suppose that introspection itself is bootstrapped by a functional analog of normal perception and behavior. Thus, the "Mind's Eye" viewing an internal workspace is simply a functional operation generated by our experience of turning our real eyes towards external workspaces.

If you can't already guess, such thoughts stem from Julian Jaynes' views on introspection.

Eric Schwitzgebel said...

@ John: Delighted to contribute to your delinquency!

@ Gary: I'm a little reluctant to push too hard on the introspection / perception analogy, especially if the relevant perceptual sense is vision. There are parallels, of course, but also big differences. For example, judging introspectively that such-and-such is the case can help make it true that such-and-such is the case; not so for vision. The causal loops in introspection, I think, are short as well as long and the feedback and interactivity are richer. I do agree with you (and Jaynes, and Vygotsky, and Clark, and others) that the scaffolding of language and culture adds a dimension of richness to the whole procedure that can transform it fundamentally.

Gary Williams said...

"For example, judging introspectively that such-and-such is the case can help make it true that such-and-such is the case; not so for vision."

This is an insightful point, but I would be curious as to whether such judging capacities are bootstrapped from the more primordial phenomenon of "getting a better look" in terms of perceptual scanning. We might see something in the corner of our eyes, or in a low light area, and then "double check" the veracity of our original perception by looking harder, focusing or rubbing our eyes, etc.

Anibal Monasterio Astobiza said...

The detection-after account of introspections fits well with the so-called "global workspace model" of consciousness where the prefrontal cortex is the region in charge of selectively filtering what enters the stream of conciousness from a widely distributed inputs arising in the rest of the mind/brain.

The prefrontal cortex would be the "scaninng mechanism".

Regarding the second account of introspection as far as my knowledge reaches i can´t see any analogy with current models.

But if the embodied or enactive conceptions of the mind are expanding the boundaries of our way to understand cognition far beyond our skins and skulls i wonder if we can also expand and even train our limited instrospection.

Outward-related cognitive activities can be trained or "supersize" (to see the world i use glasses) why not expect the same with inward-related cognitive activities.