Wednesday, September 12, 2007

Three Reasons to Mistrust Reports about Ongoing Conscious Experience

I'm off to Oxford tomorrow for the annual meeting of the Consciousness & Experiential Society Section of the British Psychological Society. Russ Hurlburt and I will be keynote speakers. We'll be talking about our forthcoming book, but for most of our 2 1/2 hours we'll be "beeping" the audience. That is, we'll set a random beeper to sound while we're talking. When the beep occurs, audience members will reflect on their "last undisturbed moment of inner experience" immediately before the beep. Then we'll interview people about their sampled experiences, right there on the spot, Russ as a long-time practitioner of experience sampling methods and I from a more skeptical perspective. (Hopefully, no one will say "I was thinking about how boring and awful your talk is and how ugly you are!")

For the more conventional part of the presentation (so people will have something to think about while waiting for the random beeps to surprise them), I've worked up a fifteen-minute essay on why I'm inclined to mistrust even confident reports about currently ongoing conscious experience. I have three main reasons:

(1.) Historically, even "expert" introspectors have tended to make radically different claims about the ordinary stream of conscious experience. Some of them must be pretty badly wrong, despite their evident expertise and care.

(2.) We don't have much practice thinking or talking about our stream of experience. Our vocabulary and concepts are built for making judgments about the world around us (or non-introspective judgments about ourselves). Yet the stream of experience is plausibly complex and fast-moving.

(3.) Although people often convey confidence in their introspective judgments, that confidence can be undermined and the confident judgments reversed under certain styles of questioning, suggesting that the confidence may not be well founded.

I've posted the full text of the talk in The Underblog.


Anonymous said...

Have a safe trip, and hope your talk goes well!

Eric Schwitzgebel said...


Anonymous said...


Reason (1) seems more appropriate for casting doubts on generalizations about experiences instead of reports.

Eric Schwitzgebel said...

Hi Pete! Is your thought, then, something like this? When William James introspected his emotional experience, he did so accurately, finding only bodily phenomenology, but then he leapt too quickly to the conclusion that emotional experience in general is like that? (If you like James's phenomenology of emotion, alter the example.)

If so, is the further thought that James just had very different emotional experience in general from, say, Titchener -- so that their phenomenological differences matched their theories? I agree that's a possibility, though one reason I find it unattractive is that it poorly handles cases in which theories change (did the person's whole phenomenological structure change right along with the theory?).

Or is the idea that maybe James had the same sort of emotional experience as Titchener, overall, but leapt to conclusions based on too small an introspective sample size? That seems a little far-fetched. He must have talked and thought about and reflected on his emotional phenomenology quite a bit, given the prominence of his view about it in his professional life.


Genius said...

I'm interested in reasons why they are wrong rather than evidence that they are wrong (2 and 3). 1 has the former sown up.

I'm inclined to think WJ was mostly right and that everyone else just doesn't want to believe it. Of course I'm a bit of a fan....

Anonymous said...

Hi Eric,

I'm happy with both of the first two choices that you offer as interpretations of what I meant to suggest. So, in other words, it's possible that people are ok at reporting their current experience, but poor at intrasubjective generalizations over multiple times as well as intersubjective generalizations.

I don't feel the force of the problem you try to raise about theory change. What's wrong with the suggestion that theory changes are either causes or effects (or both) of that which they are theories of when it comes to phenomenology? Or how about the suggestion that one's own phenomenology is identical to one's current application of an autophenomenological theory?

Of course, an analogous hyupothesis regarding theory changes about stuff in the external world would be weird indeed, but the external world is something we have multiple independent lines of evidence about. Why not suppose that one's phenomenology and one's theory of one's own phenomenology jump around in concert?

Eric Schwitzgebel said...

Sorry about the delay in responding, Pete. I wasn't able to keep up with the blog in Oxford.

Different people have different intuition about the plausibility of these sorts of radical phenomenal differences between people, I suppose. My co-author, Russ Hurlburt, for example, thinks people really are radically different inside.

Consider the case of phenomenal reporting about the visual field. Some people report 100 degrees of simultaneously clarity in a stable, largely unchanging field. Others report (as I would) that visual experience is of a small region of clarity moving rapidly around a fairly indistinct background. Do some people just have radically more precise vision in this regard than others? Although one can't read phenomenology *straight off* behavior, I'd suspect that the 100-degree folks wouldn't perform any better on tests of peripheral visual acuity. I'd also suspect that physiologically, our visual systems are very similar.

Furthermore, some of these 100-degree people change their report when I ask them to notice that they can attend to parts of the visual field away from the region of visual focus. Their estimates of the range of clarity in their visual field shrink.

There's a plausible explanation for why people might be mistaken; there's a pretty strong physiological and behavioral argument for phenomenological similarity in this respect; and people will change their reports when coached against a certain sort of potential error.

In this case, at least, it seems a strain to me to suggest that our phenomenology differs so as exactly to correspond to the differences in our report. Do you disagree? Or do you think this particular case is atypical? ...

Anonymous said...

Hi Eric,
I’m enjoying this exchange. Here are some further thoughts.

I don’t know if I would reject the visual-field case as atypical, but I do have some worries about how it’s being used. It is, of course, quite tricky just how one should go from objective facts to phenomenology. Just because there is no pink elephant in the room doesn’t mean that it can’t seem to someone that there is. Likewise, just because non-foveal visual field doesn’t have a high resolution doesn’t mean that it can’t seem to someone that it does.

Eric Schwitzgebel said...

I'm enjoying the exchange, too, Pete.

I agree that we cannot leap straightforwardly from physical facts about the fovea to phenomenological facts about visual experiment. Thus, it's key to my argument that people change their views as a result of introspecting in the right (or at least a different) way. Many of the people I've interviewed about such matters go from saying the visual field is very broadly clear to saying that they were originally mistaken in their views about how broad the region of clarity is.

To have them be right in their original claim of broad clarity and in their later claim of narrow clarity two things need to be the case: (1.) their visual experience needs to change substantially from time a to time b, and (2.) they must be mistaken about previously having been wrong. (1) seems physiologically unlikely and (2) seems an odd combination of introspective infallibility in a moment with profound ignorance of change over time.

Anonymous said...

Hi Eric,

I think that perhaps you and I disagree on what count as the phenomenological facts ("phacts"?) here.

I would not, for instance, count as phenomenological facts
facts concerning whether people can reliabliy identify determinate colors and shapes in their peripheral visual field. Similarly, while I regard it as a fact that the visual field is not "broadly clear," I would not regard it as a phenomenological fact. It's no more a phenomenological fact, in my book, than that it's a phenomenological fact that I'm experiencing red for the 506th time since noon last wednesday.

I'd offer as a candidate for a phenomenological fact in the visual field case the fact concerning whether the visual field seems broady clear as opposed to whether it is broadly clear.

Eric Schwitzgebel said...

Ah, well, Pete -- I happen to think the word "seems" in these contexts is very problematic! (I think we've discussed this?) So I'm not quite sure how to understand your last sentence.

I would say that facts about being able to reliably identify peripheral colors and shapes are behavioral or cognitive facts rather than "phacts" (cute word!).

But what exactly are the phacts, then, if they're not behavioral competences and they're not epistemic "seemings"? Well, I can't give a reductive definition! (Why should I be able to?) But I think we all know what facts I'm talking about -- don't we? What makes it the case that there's "something it's like" to be you, your consciousness, your qualitative experience, your inner stream of experience.... If such phacts happen to correspond with facts about what our judgments are or what our cognitive processes are, that's a *substantive* truth and not something true simply by the definition of "consciousness".

I suspect you'll disagree with me about all or some of this! Is this really what our dispute about the reliability of introspective judgments comes down to, then? A disagreement about whether phenomenal facts are really just epistemic "seemings" of some sort?

Anonymous said...

Hi Eric,

I think the suspicions you express in the last paragraph of your comment are largely correct. I've got more to say on this and am working it up into something to post on Brain hammer. Stay tuned.

Eric Schwitzgebel said...

I check the Hammer at least once a week. It's nice to think that we've got to the root of our disagreement here, and it turns out to be a really fundamental issue.