Thursday, September 06, 2018

Inflate and Explode

Here's a way to deny the existence of things of Type X. Assume that things of Type X must have Property A, and then argue that nothing has Property A.

If that assumption is wrong -- if things of Type X needn't necessarily have Property A -- then you've given what I'll pejoratively call an inflate-and-explode argument. This is what I think is going on in eliminativism and "illusionism" about (phenomenal) consciousness. The eliminativist or illusionist wrongly treats one or another dubious property as essential to "consciousness" (or "qualia" or "what-it's-like-ness" or...), argues perhaps rightly that nothing in fact has that dubious property, and then falsely concludes that consciousness does not exist or is an illusion.

I am motivated to write this post in part due to influential recent work by Keith Frankish and Jay Garfield, who I think make this mistake.

-----------------------------------------

Some earlier examples of the inflate-and-explode strategy include:

Paul Feyerabend (1965) denies that mental processes of any sort exist. He does so on the grounds that "mental processes", understood in the ordinary sense, are necessarily nonmaterial, and only material things exist.

Patricia Churchland (1983) argues that the concept of consciousness may "fall apart" or be rendered obsolete (or at least require "transmutation") because the idea of consciousness is deeply, perhaps inseparably, connected with false empirical views about the transparency of our mental lives and the centrality of linguistic expression.

Daniel Dennett (1991) argues that "qualia" do not exist, on the grounds that qualia are supposed by their nature to be ineffable and irreducible to scientifically discoverable mental mechanisms.

Unfortunately, philosophical enthusiasts for the importance of conscious experience tend to set themselves up for the inflate-and-explode move, making Feyerabend's, Churchland's, and Dennett's critisms understandable.

The problem on the enthusiasts' side, as I see it, is that they tend to want to do two things simultaneously:

(1.) They want to use the word "consciousness" or "phenomenology" or "qualia" or whatever to refer to that undeniable stream of experience that we all have.

(2.) In characterizing that stream, or for the sake of some other philosophical project, they typically make some dubious assertions about its nature. They might claim that we know it infallibly well, or that it forms the basis of our understanding of the outside world, or that's irreducible to merely functional for physical processes, or....

Now if the additional claims that the enthusiasts made in (2) were correct, the double purpose would be approximately harmless. However, I'm inclined to think that these types of claims are generally not correct, or at least are quite legitimately disputable. Thus, the enthusiasts unfortunately invite inflate-and-explode. They invite critics to think that those dubious claims are essential to the existence of consciousness in the intended sense, such that if those dubious claims prove false, that's sufficient to show that consciousness doesn't exist.

The reason I think that Feyerabend, Churchland, and Dennett are inflating the target, rather than just correctly interpreting the target, is that I believe the enthusiasts would much more readily abandon the dubious claims, if required to do so by force of argument, than they would deny the existence of consciousness. Those claims aren't really ineliminably, foundationally important to their concept of consciousness. It's not like the relation between magical powers and witches on some medieval European conceptions of witches, such that if magical powers were shown not to exist, the right conclusion would be that witches don't exist. Even if we must jettison thoughts of infallibility or immateriality, consciousness in our communally shared sense of the term still exists. The core conception of phenomenal consciousness in philosophy of mind is, I think or suspect or at least hope, the conception of the stream of experience that it is almost impossible to deny the existence of -- not that stream-of-experience-plus-such-and-such-a-dubious-property.

-----------------------------------------

Frankish's and Garfield's more recent illusionist arguments, as I see them, employ the same mistaken inflate-and-explode strategy. Keith Frankish (2016) argues that phenomenal consciousness is an "illusion" because there are no phenomenal properties that are "private", ineffable, or irreducible to physical or functional processes. Jay Garfield (2015) denies the existence of phenomenal consciousness on the broadly Buddhist grounds that there is no "subject" of experience of the sort required and that we don't have the kind of infallibility about experience that friends of phenomenal consciousness assume.

Now it is true that many recent philosophers think that consciousness involves privacy, ineffability, irreducibility, infallibility, or a subject of experience of the sort not countenanced by (some) Buddhists; and maybe they are wrong to think so. On these matters, Frankish's and Garfield's (and Feyerabend's and Churchland's and Dennett's) criticisms have substantial merit. But it does not follow that consciousness is a mere illusion or does not exist. We can, and I think normally do, conceptualize consciousness more innocently. We need not commit to such dubious theses; our shared conception can survive without them.

To avoid commitment to dubious theses, we can and do define consciousness primarily by example. We gesture, so to speak, toward our sense experiences, our imagery experiences, our vividly felt emotions, our inner speech. We notice that there is something extremely obvious that all of these examples vividly share. Consciousness is that obviously shared thing. Maybe it's reducible; maybe not. Maybe there's a "subject" in a Cartesian sense; maybe not. Why commit on such matters, right out of the gate? Keep it theoretically innocent! Consciousness, in this innocent sense, is almost undeniably real. (I say "almost" because the clever philosopher can find a way to deny anything.)

Now admittedly, this sort of theoretically innocent definition by example is not quite as simple as I've just portrayed it. For a more careful attempt see Schwitzgebel 2016.

-----------------------------------------

I've tried this argument on both Frankish and Garfield, in critical commentaries (contra Frankish; contra Garfield). They remain unconvinced. (Well, this is philosophy!) Let me summarize their replies and share my reaction.

Frankish says that he agrees that consciousness, defined innocently by example as I have done, does indeed exist. He graciously allows that I have executed the important task of identifying a "neutral explanandum" for theories of consciousness that both realists and illusionists can accept (p. 227). However, Frankish also asserts that my definition is "not substantive" "in the substantive sense created by the phenomenality language game" (ibid.), and thus he feels licensed to continue to embrace illusionism about phenomenal consciousness.

I remain unsure why my definition by example is insufficiently substantive. Surely some definitions by example are substantive, or substantive enough. For instance, I might define "furniture" by reference to a diversity of positive and negative examples. That seems to pick out a substantive target of things that exist, and done well, it's good enough to let us start counting pieces of furniture (maybe with some disputable cases), evaluating the quality and function of different types of furniture, etc. Why wouldn't example-by-definition of consciousness work similarly? What is missing?

Garfield responds differently, doubling down, as I see it, on the inflation move:

I argue that if by 'qualitative states' we mean states that are the objects of immediate awareness, the foundation of our empirical knowledge, inner states that we introspect, with qualitative properties that are properties of those states and not of the objects we perceive, there are no such states (Garfield 2018).

Whoa! I don't think I meant all that! My whole aim in definition by example is to avoid such commitments.

Maybe Garfield takes himself to be denying the existence only of properties that most 21st century Anglophone philosophers don't actually endorse? No, I don't think so. It is clear from context that in denying the existence of qualitative properties, Garfield takes himself to be in conflict with the mainstream view in philosophy of mind, the view of people like me who accept the existence of phenomenal consciousness. But I don't see why Nagel, Block, Searle, Chalmers, Strawson, Carruthers, Kriegel, Siegel, Siewert, Thompson, etc. need to be committed to the dubious package of views Garfield lists in the blockquote above, simply by virtue of accepting the existence of consciousness. Of course they may also make other, further claims about consciousness, besides merely asserting that it exists, and those further claims might commit some of them to the dubious theses that Garfield wisely rejects.

-----------------------------------------

[image source]

20 comments:

Luke said...

FWIW, I prefer (and use) a Schwitzgebelian definition of phenomenal consciousness (e.g. here), but I am also a Frankish-esque "strong illusionist" (see here).

That said, I prefer a different vocabulary than Frankish, in part because my primary audience is not professional philosophers who have spent decades arguing about very specific notions of qualia, consciousness, etc. I find it clearest to describe my view as: "Phenomenal consciousness (defined roughly as Schwitzgebel does) exists and is morally important, but it doesn't have many/most of the properties that are often ascribed to it, including those which are supposed to make it particularly special and unique (e.g. intrinsicness), and (per Frankish 2012) I can't see what properties "diet qualia" are supposed to have that "zero qualia" don't. (Or when I can tell, I think we have good reason to believe qualia don't have those properties.)

If someone denies the existence of qualia because they insist on a "classic qualia" definition (see Frankish 2012), I think it's arguable that they're doing something like "inflate and explode." But if someone merely argues (as Frankish does) that there isn't anything more to "diet qualia" than "zero qualia," is that really "inflate and explode"? Is it still "inflate and explode" if someone (e.g. me) roughly agrees with the substantive points (e.g. about classic vs. diet vs. zero qualia) of Frankish's illusionism, but prefers my language for talking about consciousness over statements like "phenomenal consciousness is an illusion"? How much are you objecting to Frankish's substantive claims and argumentative strategy vs. his rhetoric?

Eric Schwitzgebel said...

Thanks for the thoughtful comment, Luke!

I am inclined to agree that phenomenal consciousness doesn't have many of the properties that are often ascribed to it -- especially special epistemic privilege, as I've argued extensively (I'm willing to tolerate some others, e.g., immateriality, as live epistemic possibilities deserving a small but non-trivial credence).

One difference between Frankish's zero qualia, if I recall, and his strong illusionist position, and my position is his emphasis on dispositions to report... wait -- more later!

Eric Schwitzgebel said...

Completing the thought above: Frankish's offers a judgment- or report-centered view of what "quasi-phenomenality" or "zero qualia" amount to. As I articulate in my 2016 criticism, and as I still believe (I don't think he adequately addresses it in his reply to me), that sets up false dichotomy in which consciousness must either be some inflated thing or is *only* something like a disposition to make certain sorts of judgments. I believe there is substantial space between inflated consciousness and dispositions to judge, and that a substantive notion of consciousness, defined by example, can pick out consciousness in a way that is neutral between the inflated thing, consciousness-as-reducible-to-dispositions-to-judge (if it is so reducible), and consciousness-reducible-to-something-else (e.g., integrated information, higher-order representations, or whatever).

Lee Roetcisoender said...

Eric,

FWIW, "illusion" is an overused, inherently useless and divisive idiom in the consciousness debate. The preferred term to replace illusion would be "condition"; like herpes is a condition of biology. Since our phenomenal realm is the appearance of reality, the first person objective experience of consciousness would be a condition on the possibility of real experience. The idiom "condition" expresses the first person objective experience of consciousness much better than the term illusion, plus it is inherently less divisive. So, by dropping the term illusion, maybe we can all agree on something?? This vocabulary also cuts the legs out from under the materialistic view by eliminating their exclusive claim to the moral high ground of "illusion". And yes, consciousness is a first person objective experience, not a subjective one; so now one possesses a vocabulary that inflates and explodes the Buddhist position as well, a position that requires a subject...

Eric Schwitzgebel said...

Thanks, Lee, I agree that "illusion" is a somewhat confusing term in this context. One reason I think Frankish likes the term, though, is that it contains the implication of error, which I don't think "condition" does in the same way. I think Frankish and Garfield do want that implication of error, and since they take himself to disagree with the mainstream view, they aim to highlight their differences, even if it results in being divisive.

Keith Frankish said...

Hi Eric,

Nice post. A quick response. I don’t see my strategy as “inflate and explode”. It’s more a challenge to people who don’t want to inflate: deflate or explain.

Let’s put it in terms of properties of experience -- feels, phenomenal properties, qualia, whatever. You can have an inflationary concept of these -- as intrinsic, ineffable, genuinely “feely” properties (“classic qualia”), but that means giving up on the physicalist explanatory scheme. Or you can have a deflationary concept of them -- as properties that dispose us to judge that the experiences that possess them have ineffable, intrinsic, feely properties (“zero qualia”). This is the illusionist’s concept (genuinely feely properties are illusory).

Now, my challenge is to people who don’t want to adopt the inflationary concept: either adopt the deflationary concept or explain what your concept is and how it differs from the other two. I think any attempt to do the latter will end up either inflating the concept into a version of classic qualia or deflating it into a version of zero qualia. But feel free to prove me wrong on that!

As for a neutral explanandum for theories of consciousness, it needs to be defined in a way that doesn’t beg the question between these different concepts of phenomenality. That is, it mustn’t use a concept of feel, what it-is-likeness, subjective character, etc. It could be something like “states of the kind we’re disposed to describe as having a qualitative feel” (that create the illusion of phenomenality), or “states of the kind that occur when we perceptually attend to something” -- where these are compatible with a range of concepts of phenomenality from the inflationary to the deflationary.

I thought your attempt to define a minimal conception of consciousness in your JCS reply to me did a pretty good job of meeting this requirement. (Though I’m not sure about the word “steam” you use in the post -- what’s flowing and what or who is it flowing past?)

Anyway, that’s the moral. Not inflate and explode but deflate or explain!

Josh Weisberg said...

Hi Eric,

Nice post!

A few thoughts:

1. You might want to add Richard Rorty to your list of inflaters and exploders: see his stuff on privacy and the mental.

2. I get that qualophiles reject the inflated definition, but it seems to me that at the next step in the dialectic, they tend to appeal to those sorts of properties, either in rejecting functionalists (physicalists, etc.) views for failing to account for those properties, or by using the epistemic authority conveyed by those properties to defend their views. This suggests that without those properties, they can’t mount a decent rejection of deflationary (or eliminativist) views. Does that make the properties essential to consciousness for them? Maybe not, but it makes them essential to a defense of qualia, the hard problem, etc. Or so it seems to me.

3. The (alleged) conceivability of zombies is supposed to show that the “intension” of our “ordinary concept” of consciousness has more to it than functional/physical stuff. Put another way, does your minimal concept allow for (explain?) the “easy conceivability” of zombies? I would imagine that Dave Chalmers might protest if it doesn’t.

4. A general worry in this direction. If the minimal conception is right, why do we think (some of us) that there’s a problem of consciousness? (This, I take it, is something like Dave’s “meta-problem.”)

5. A Quinean thought: maybe what would happen when qualophiles are confronted with inflate/explode is that they change concepts. That is, when faced with the explosion, they move to defend a new and different concept: SCHENOMENAL CONSCIOUSNESS, though they don’t change its English name (“phenomenal consciousness”). So Keith and Jay are right, and elimination occurs. So your evidence from linguistic behavior doesn’t make the point. But my Quinean worry is how could you tell if that’s what folk did, rather than revising the original concept to drop some inessential stuff? Is there really such a clear line between elimination and revise and reduce here?

Cool stuff!

(Gonna crosspost this on facebook to get it out there.)

Lee Roetcisoender said...

Keith wrote:

"It could be something like “states of the kind we’re disposed to describe as having a qualitative feel” (that create the illusion of phenomenality)..."

I do not find this type of vocabulary useful in the consciousness debate, it may give the materialist the moral high ground over the idealist, but it does not help those "people who don’t want to adopt the inflationary concept: either adopt the deflationary concept or explain what your concept is and how it differs from the other two." Now, if one would preference the idiom condition over the suppressive term illusion, the same statement slightly modified would add clarity to our understanding, for example: "It could be something like “states of the kind we’re disposed to describe as having a qualitative feel” (that create the 'condition' of phenomenality)..."

Taking this concept one incremental step further in its natural progression, one could then state that consciousness is the "condition of phenomenality", and if consciousness is the condition of phenomenality, then that "condition" is a fundamental qualitative property of consciousness. Consequently, this fundamental qualitative property can be reducible all the way down the hierarchy of form to include mass, spin and charge, particle, waves, and the infiniteness of both inner and outer space.

This simple example is just one of the many paths which lead one down the rabbit hole of the mystery of consciousness. As unconventional as this approach may appear, one has to keep in mind that consciousness defies all of our conventional rationales, so maybe it's time to scrap those models of thought and abandon the conventional paradigm??

Eric Schwitzgebel said...

Keith and Josh — thanks for those terrific, thoughtful replies! More soon....

Christopher Hitchcock said...

Hi Eric,

Nice post. I've long held a pet theory that every concept of philosophical interest goes through at least one episode of what you call "inflate and explode" (I've sometimes described it as being analyzed out of existence). In many cases, but not always, this is a needed precursor to the emergence of a more apt deflationary view. Some examples:

Time: Inflated and exploded by McTaggart. In order for something to be time, it must have A-theoretic properties, but these are contradictory. The needed deflation was the B-theory, which McTaggart himself formulated, but thought to be inadequate.

Knowledge: Had been inflated for millennia; exploded by Locke. Locke didn't fully explode knowledge, but argued that most of what we would call empirical knowledge was impossible. Hume provided the needed deflation. In the 20th C, Unger was an inflater & exploder. Fallibilism is the needed deflation (which was already in existence).

Causation: Inflated and exploded in the 17th C by, e.g. the occasionalists, Berkeley, Leibniz. Not fully exploded, but the standard for causation was so demanding that only God could do it. Again, Hume provided the needed deflation. Perhaps there was another inflation and explosion around the beginning of the 20th C, e.g. in Russell, and Karl Pearson, among others.

Free will: Pereboom and Galen Strawson are inflaters and exploders. In order to have free will, something like libertarianism would have to be true. But libertarianism is incoherent or at least wildly implausible. In this case, the deflationary view -- compatibilism -- had been around for a long time. Perhaps Spinoza is an earlier inflater and exploder.

I think there is a slightly different phenomenon, where sensible precautions get exaggerated to the point of explosion. Sensible precaution: correlation is not causation. Exaggeration: causation is woolly metaphysics, to be avoided at all costs. Sensible precaution: scientists have the same prejudices as other people. Exaggeration: science is by its very nature racist and sexist.

Eric Schwitzgebel said...

Josh, thanks for the great comment!

1. Yes, that sounds right, but it has been a long time since I read Rorty on this.

2. Agreed. I am not a qualiaphile in that strong sense. I hope one doesn't have to be one in order to think that consciousness exists in the standard, phenomenal sense of "consciousness" shared among most philosophers interested in the study of consciousness.

3. I think zombies are conceivable with a low bar on conceivability. I also think it's low-bar conceivable that x^2 - x + 1 crosses the x axis. My opinions about high-bar conceivability are a bit woollier, but I'd hope that my attempt at an innocent definition by example doesn't commit on high-bar conceivability.

4. I guess I regard that as a separable question. As you point out, David Chalmers has an interesting paper on this. I am broadly attracted to the idea that we learn about consciousness differently than we learn about physical/functional facts and (if naturalistic materialism is true) that is part of the explanation of an explanatory gap.

5. I agree that there's a possible gray area here.

[crossposted on FB]

Eric Schwitzgebel said...

Very cool, Chris! Those are very cool examples. I'm inclined to agree that it is a common move for philosophical eliminationists of all kinds -- in the intended pejorative sense, of course, only when elimination is unjustified (e.g., not among atheists arguing against the existence of the orthodox monotheistic God).

Eric Schwitzgebel said...

Keith, thanks again for the thoughtful and generous reply!

I'm inclined to ask, when you allow the existence of phenomenal properties in the deflationary sense of "properties that dispose us to judge that the experiences that possess them have ineffable, intrinsic, feely properties" are you thinking of this characterization of zero-qualia deflationary phenomenal properties as a de dicto or a de re characterization? If it's de dicto, then I think you're still posing a false dilemma. If it's de re, then maybe our views are closer than I had been thinking.

To spell this out a little more. On the de re reading I'd be happier with there is some natural kind, or at least good ontological kind, that as a matter of fact tends to cause us to attribute these dubious properties to our mentality, and zero-qualia-phenomenal-properties are just whatever property in the world happens to be doing that for us -- and which might also exist, for example, in mice even though mice have no tendency to attribute dubious properties to their mentality. (Compare: I might define "water" as whatever happens to be in this cup here, though of course it can exist outside of this cup.)

On the stronger de dicto reading, there isn't that looseness between zero-qualia-phenomenal-properties and the disposition to report. Ceteris paribus (even if not perfectly) to have qualia in this deflationary sense just *is* to be disposed to report (possible caveat: if you are a creature with linguistic capacities, but then my mouse example gets complicated).

Keith Frankish said...

Thanks for your reply Eric. I don't have a simple answer to your question. I think an illusionist can take the de re view (it's a broad church), but my own position is more qualified. I expect other mammals do share some of the properties centrally involved in disposing us to judge that we have qualia, but I suspect that language is also necessary to creating the illusion (though I don't think that actually making phenomenal judgements is).

There may be other differences between us too. For example, I doubt that the properties involved are introspectable in a perceptual sense, and I suspect that some of them at least are properties of events downstream from experience rather of experiences themselves.

By the way, I really liked Josh's comments, especially his point 2. Qualiaphiles need an inflated conception in order to make a case against deflationism/illusionism.

Garret Merriam said...

Perhaps the eliminativist can make a different kind of move. Rather than committing to any particular essentialism about 'consciousness' (or 'qualia' or whatever), they might say that ANY substantive theory of these things is bound to tie us up with unnecessary philosophical baggage (including your ostensive, case-driven approach.) In Laktatosian terms, we might call such theories 'degenerating.' Meanwhile, we have a completely separate vocabulary (e.g.--a neuroscientific one, or whathaveyou) that is progressive (again, in the Lakatosian sense.) So why not make the semantic migration towards the vocabulary that seems like it will actually DO more for us, and away from the thorny, muddled mess of the 'old program.' To use your example, we should eliminate 'witch' from our vocabulary, not because witches are inherently magical and there is no such thing as magic, but because 'witch' is vague, ambiguous, open to too wide an array of interpretation, laden with conceptual baggage, and not terribly informative.

To wit--We should eliminate 'consciousness' language simply because it's too philosophically CONFUSING and doesn't buy us anything for that confusion.

Eric Schwitzgebel said...

Thanks for those comments, Keith and Garrett! Just a quick reply for now:

Keith: I think both of those issues will be difficult to untangle in blog comments. I look forward to chatting with you more about it soon. (I’m not a straight perceptualist about introspection, but I do think that there are often perceptual elements.)

Garret: Yes, I can see that argument. I’d pull apart the questions about irreducibility and such from the questions about the types of structural properties of conscious experience that we are trying to make progress on in scientific research, such as the frequency and vividness of imagery experience, the experiential character of peripheral vision, etc.

Unknown said...

"inflate-and-explode" also suggest intentionality is unaffected in argument...
...if intention is toward one's own consciousness...

The idea there or here is, if I can be conscious then everyone can be conscious...
...that material has and is purpose...

Anonymous said...

What is your view on the relation between consciousness and the brain? Is it possible something survives death or is that completely ruled out?

Callan S. said...

That doesn't seem fair

Person A : "I think mental processes and qualia and ineffable."

Person B : "Mental processes and qualia don't exist because only material things exist"

Person C : "Person B, you're just say they don't exist"

That's not fair of person C - person B is referring to persons A's belief/claim on the matter. Treating it like person B is not referring to someone else's beliefs/claims and is instead just claiming it themselves, that's really not fair.

Callan S. said...

*'Person B, you're just saying they don't exist'

To clarify further it's like B is saying both things themselves - that qualia are ineffable and that ineffable things don't exist because only material things exist.

Person B is not setting up the terms of qualia then knocking down their own terms. They are trying to work with someone else's claims in some kind of reasoned manner. They are working with a balloon someone else blew air into, they aren't doing the inflation. That it pops in B's hands doesn't mean B just set it up to pop. It was just a badly made balloon.