COMPLEMENTING DESCARTES

JS: "Your repeated "how/why" questions presuppose the very distinction which is in question here, namely that between feeling and functing.  Until this distinction is clarified, we will remain at an impasse."

How about the distinction between feeling and doing, then? Is that clear enough? (It's much the same distinction.) 

How and why the brain causes adaptive behavior is a tractable scientific question, a functional one, that will one day have a full, clear answer. 

Not so for how (and especially why) the brain causes feeling. (And that's the point, and the problem, and the gap.)

JS: "if feelings have no causal efficacy, they do not make a difference to anything, including the conclusions we draw in our discourse on feelings.  So why do we have words for them?"

(1) Feelings are there, being felt (when they are being felt).

(2) There is an (unexplained -- and I think causally inexplicable, though undoubtedly -- if not undoubtably -- causal) correlation between our feelings and our doings (hence between our feelings and our sayings), probably explained by the common functional cause that (explicably) causes the doings and (inexplicably) also causes the feelings.  

So there are feelings there, to speak of, and we do speak of them; and speaking certainly has causal consequences. But until and unless there can be a causal explanation of how and why we feel, the only available explanation of why we speak of feelings is that the same cause that (inexplicably) makes us feel and (explicably) act also (mysteriously) makes us speak of feeling; but the fact that we actually feel has no independent causal role, hence no causal explanation. It just dangles on the joint cause of the feeling (unexplained) and the speaking.

(I did speculate a bit -- on one of the earlier threads of this discussion: "WHY WOULD TURING-INDISTINGUISHABLE ZOMBIES TALK ABOUT FEELINGS (AND WHAT, IF ANYTHING, WOULD THEY MEAN)?" -- concerning why Turing-Test-scale robots, with behavioral capacities indistinguishable from our own -- if they were feelingless Zombies -- would speak of feelings at all. One possibility might be that the words would be used as metaphors for unobservable internal states -- unfelt states, but also states that are inaccessible to other agents with which the TT-passing robot must interact adaptively. So "you have hurt me" might be a short-hand for "you have caused damage to my internal functioning." That would make feeling-talk ("mind-reading") functional rather than a dangler, like feelings themselves. But I have not yet carried through the exercise so far as to try to construe what functional role "feeling" talk could play if the exchange between us [in this very email dialogue] were taking place between Zombies, and they were talking specifically about the difference between the functional role of talk about feelings between feelingless Zombies versus talk about feelings between feeling people. Maybe that's just further evidence that there could not be feelingless Zombies Turing-indistinguishable from us. But unfortunately that leaves completely unanswered, yet again, the [same old] question, this time in the form: how and why not! Same old explanatory gap... [Peter Carruthers has a recent target article on this in BBS, but I think he gets it somewhat backwards: it is feeling that is primary, not mind-reading, whether of the unobservable states of others or one's own...])

JS: "Your view makes all talk of feelings superfluous, including the claim that there is a feeling/functing distinction."

No. It just points out that how and why we feel is unexplained (and how and why I think it is also inexplicable: functional superfluousness; no telekinesis; causal inexplicability). 

JS: "The notion of 'what it is like to be a bachelor' does not pick out any particular feel or category."

"What it feels like to be a bachelor" picks out what every waking minute feels like (to a human male) from birth to the first minute one gets married -- at which point it is complemented (and one discovers how right or wrong one had been about "what it feels like to be a bachelor"). No such possibility for what it feels like to be awake, or alive...

JS: "there is nothing it is like to not have a third arm..."

That's largely true (except in contrast to what it feels like to have a third arm, as, say, siamese twins, spiders, or a surgically-altered-me might experience). 

But in general I do agree that arbitrary counterfactual complementations are of no more interest than "what it feels like to see something that is bigger than a breadbox" (which does happen to be complemented) or "what it feels like to have lived fewer than an infinite number of years" (which is not).

We only single out categories in cases where the complement is in some way salient (and where the invariant features of the category members -- relative to the complement members -- are used to resolve uncertainty about what is a member of the category and what is a member of its complement). It does make sense to say "I know what it feels like to be a bachelor," and I can even discover that I was wrong. 

In much the same way, it does make sense to say "I know what it feels like to be alive" or "I know what it feels like to be awake." And we probably do have a pretty good idea from our positive-only evidence. But the difference is that there is no way we can discover whether we were spot-on or not quite right; and perhaps we are not really justified in making all the inferences we tend to make from our uncertain grip on these problematic categories. 

(The standard kluge we use for "what it feels like to be alive" is to complement it with analogies, including an imaginary afterlife or rebirth; and for "what it feels like to be awake" we incoherently complement it with what it feels like to be asleep and dreaming -- which is of course not exactly a "nonawake" experience in the same way that delta [dreamless] sleep is -- but in delta sleep you're gone, so there is no one feeling what it's like...)

JS: "If we admitted all of these “what it is likes” into our experiential set, then each person would have to “sample” (to use your word) an infinite number of feels before they could know what it is like to feel anything at all."

No, not only do all those hypothetical complements never occur to us, but even when they do, they can easily be dismissed as arbitrary, inconsequential and uninformative. Not so for some of them, though, because we persist in thinking of and speaking of them as if the distinction were salient: "It feels good to be alive" or "Some of my brain functions are felt and others are not." Nor are the intended distinctions empty in those cases. They are merely uncomplemented, hence problematic.

(On arbitrary negative categories and their relation to our sense of similarity, see also Watanabe's "Ugly Duckling Theorem.") 

JS: "There is no "invariant feeling" running through all feelings."

The reason there is no functional invariant here is that it is normally the complement that determines what is and is not invariant in a category: The invariant is relative, based on contrasting what all members of the category share and what all members of its complement lack. (Please let's not get into family resemblances: invariants can be disjunctive and conditional too.) But with positive-only categories, we nevertheless have access to what all the positive instances have in common. After all, we do know we are feeling when we are feeling. We are never in doubt about that...

JS: "To complement the category of feeling something, we don’t need to know what it feels like to feel nothing at all.  Rather, we must simply have the category of not feeling anything.  And we have that category."

I'm afraid not. The positive category is "what it feels like to feel something" and hence the complement would have to be "what it feels like to feel nothing at all." And that category is empty, hence we have no idea (or only incoherent fantasies) of "what it feels like to feel nothing at all." 

(Your error is, I think, a bit like mixing up the categorical distinction between (1) what is alive versus what is non-alive with the categorical distinction between (2) "what it feels like to be alive" versus "what it feels like to be non-alive": We have no trouble distinguishing things that are alive from things that are dead [or have never been alive]; but we never even face the problem of distinguishing "what it feels like to feel something" from "what it feels like to feel nothing at all," because the latter is impossible, hence empty. The only reason you have that category in your repertoire at all is that you are going by the positive instances plus some provisional analogy-based imaginary complement -- as I would be doing, in imagining what it would feel like to be married, whilst I'm still a bachelor -- except that in the case of "what it feels like to feel something" it is certain the imaginary complement is impossible, hence empty.)

(I think you may also be missing the essentially relational nature of feeling: the feeling is always felt, hence it has an implicit feeler: this is taken up in the discussion of the cogito, later below.)

JS: "I can distinguish between something which feels and something which does not feel."

Of course you can, but that's like distinguishing between something that's alive and dead (as in (1) above). That's not the category we're talking about! (We are talking of (2), above.)

JS: "We have positive and negative categories for feelings.  Some feelings are categorizable as “not feeling boredom” and others as “not tasting mustard.”  

I've mentioned these before too. You are complementing the wrong category. What it feels like to feel this (versus that) is perfectly well-complemented. But that's no help if the category in question is "what it feels like to feel something (anything) at all" versus "what it feels like to feel nothing at all." 

(An analogy: If the only sense-modality were vision, and the only experience were to see shapes, and all shapes were colored -- counting black as a color -- then the subordinate category "red" would be complemented by anything non-red, but the superordinate category "colored" would be uncomplemented.)

JS: "I can thus form the categories of “not feeling this” and “not feeling that,” and I can further abstract and form the category, “not feeling anything”.  This is exactly what we do when we abstract from “feeling this” and “feeling that” to “feeling something.”  So why talk about uncomplemented categories here?"

You're simply repeating, I think, your conviction that in complementing subcategories of a category against other subcategories of a category, we are somehow also complementing the category as a whole, against its own complement. But we are not. You are making a category error...

JS: "Despite your assertion to the contrary, we do not know 'what it feels like to feel anything at all, be it headache or toothache.'  “Anything at all” does not pick out any particular experience.  There is nothing it is like to feel anything at all."

The category in question is "what it feels like to feel something," where the something is anything that can be felt. That's no different from saying that once a child has learnt the category "dog," he knows what a dog is, and can now (correctly) recognize any dog at all, not before seen, as a dog. The same is true for "feeling": We (correctly) recognize any feeling we feel at all as a feeling. The difference is that the child has learned the category "dog" from having sampled both dogs and non-dogs, and abstracting the invariants that reliably distinguish any dog at all from non-dogs. 

We have done only part of that with feelings: We can (correctly) recognize "what it feels like to feel something" on every occasion, but we really have no idea how to distinguish "what it feels like to feel something" from "what it feels like to feel nothing at all" (even though we think we have) because it is impossible to feel "what it feels like to feel nothing at all."

JS: "The abstract category of “feeling something” does not feel like something in general; rather, it feels like a particular concept."

The category in question is "what it feels like to feel something" -- not "what it feels like to have the "concept" of someone feeling something" (or of someone being alive, or of someone being awake).

JS: "Similarly, the category of “feeling nothing at all” does not feel like nothing at all."

It sure doesn't, for that would be a contradiction in terms. The category is as empty as a square circle. Only Meinong can manage such a feat...

JS: "We feel what it feels like to think about feeling something, of course, but we also feel what it feels like to think about feeling nothing.  

The uncomplemented category in question is not "what it feels like to think about feeling something," it is the category "what it feels like to feel something."

JS: [Re: The Sentio vs. The Cogito] "Descartes' explicit claim was that the cogito established to himself that he existed."

Actually, what was demonstrated (via the "method of doubt") was that it was not true that the necessary truths of mathematics were the only things one could be certain about. There was one other thing. The Cogito, which is that when I am thinking, I cannot doubt that there is indeed thinking going on.

A slight (strategic) mistake was to focus on "thinking" (a rather vague category) rather than feeling (something we all immediately know is happening, when it is happening). 

And a slightly bigger (exegetic) mistake was to infer that the indubitable truth of the cogito was not just that I cannot doubt that I'm feeling when I'm feeling, but that therefore "I" exist (for if the category "thinking" is vague, the category "I" is even vaguer: not empty, just vague) -- rather than just that feeling exists. 

Ergo Sentitur suffices, without overstating the case... It already shows that one does not have to be uncertain about everything other than the necessary truths of mathematics (e.g., the reality of the physical world, the existence of other minds, the truth of scientific laws). One can also be certain that feeling exists.

Feeling! That one certainty among all the other undoubtedly true yet doubtably uncertain truths, such as the physical world, scientific laws, induction, causality, "functionalism." And that one certainty amidst all that less-than-certain functing -- turns out to be a causal dangler, giving rise to an unbridgeable explanatory gap!

JS: "His soul was res cogitans; his body was res extensa.  And from there, he went on to prove the goodness of God and, only then, the trustworthiness of mathematics."

I'm neither a philosopher nor a historian, but I'll bet Descartes did not believe most of that voodoo (which is certainly not what he is rightly famous for). He just said it to avoid the ire of the Inquisition. (I believe he at certain points even stated explicitly that not every cartesian claim he was making was true, hence we would need to read attentively between the lines.)

Certainty about the truth of "not (p and not-p)" is based on necessity (pain of contradiction), not on the benignity of deities; and the certainty of sentio ergo sentitur is based on the (inexplicable, but indubitable) reality of feelings. "Dualism" was just a sop for the metaphysical bean-counters of the day. The force of the Cogito is epistemic, not ontic. The sceptics were not denying the reality of the physical world, just its certainty. By the same token, it was not news that feelings existed; the news was that that was a truth that -- unlike the existence of the physical world -- we could be certain about: as certain as of the necessary truths of mathematics. 

(But a consequence of that same, certain truth, happens to be that there is no way to account for feeling physically [i.e., functionally].)

JS: "You claim that, because the cogito is merely a tautology, it needs to be reformulated so that we can better understand its significance"

The cogito has to be reformulated in terms of feeling rather than thinking, and its conclusion is the fact that it cannot be denied, when feeling is going on, that feeling is going on. But that is not a tautology, even though it sounds like one! I don't know which one of Kant's baroque categories is the right name for it, but the cogito is either a "synthetic a-priori" or an "analytic a-posteriori": it certainly isn't an analytic a-priori (i.e., a tautology).

It cannot be denied that when flying is going on, then flying is going on: That is a tautology. A universal, non-existential statement, necessarily true "in all possible worlds."

But the fact that "it cannot be denied, when feeling is going on, that feeling is going on, hence it is certain that feeling exists" depends essentially on what each of us has actually felt, namely feeling. It is an existential statement that follows from the direct experience of each and every (sentient) one of us.

JS: "The cogito is not a tautology, but an inference following modus ponens.  (If I am thinking, then I exist. I am thinking, therefore I exist.)  You misrepresent it as 'I am thinking, therefore I am thinking'." 

I agree that the cogito is not a tautology (I never said it was). But the right way to put it is that if I am feeling, then feeling exists. (The "I" is a fuzzier, theory-laden notion, not further licensed as "certain" by the cogito. At best, we can say that "it feels like an 'I' exists": but, by the same token, it feels like a physical world exists too, and that's not certain either!) 

I will say this much more, though: Feeling is essentially a "two-part relation": Whenever there are feelings, the feelings are being felt. So it is intrinsic to a feeling that there is both feeling and "feeler." I'm not talking about a fancy self-concept. Just the fact that although there is such a thing as "free-floating depression" in the sense of a depression without a perceived external cause, there is no such thing as a free-floating depression -- or any feeling -- that is unfelt. An unfelt feeling is a contradiction in terms. To that extent, a feeler is intrinsic to feeling, so the existence of a feeling to that extent entails the existence of a feeler. Maybe that's what Descartes meant by the "ego" in the "sum." But that fleeting frame for any feeling is far from what most of us mean by an ego or self, let alone the reality of an immaterial, immortal soul!  

JS: "...your "I feel, therefore feeling is felt" is not a valid inference, because there is no feeling of feeling."

I would say quite the opposite: There is no unfelt feeling. A feeler/felt relation is intrinsic to feeling. And if that's what Descartes meant by "I exist" then he was right again. But that "I" is simply an intrinsic part of the nature of feeling itself. So the existential claim of the cogito (sentitur) is still only that feeling exists. The feeling/felt relation just comes with the territory. (One cannot be certain, for example, that the feeler of the feeling is the same feeler as an instant ago: that does not sound like a sound basis for an enduring ego, let alone an eternal soul...)  

JS: "I reject the claim that [the cogito] indicates or establishes a special kind of knowledge which you call 'Cartesian certainty'.” 

Call it what you like; it's the only truth other than the necessary truths of mathematics about which we can be dead-certain.

JS: "And I reject Descartes’ views that it establishes mind/body dualism and provides a foundation for all our knowledge."

(1) "Mind/body dualism" is a figure of speech; it means next to nothing. What the certain existence of feelings establishes is the certainty of the existence of something that cannot be explained in the same functional way that the rest of what exists (truly, but without the added boost of certainty) can be explained. Reformulated as the "feeling/function" problem, it becomes obvious that the problem is one of explanation -- explaining how and why there is feeling rather than just functing.

(2) Without feeling, there would be no "knowledge," only functing. (I never said or invoked a single word about "foundations of knowledge.") "Knowledge" in books and computers and (insentient) robots is not knowledge; it is just data and dynamical states. The only knowing is felt knowing. Ditto for meaning.

JS: "Wittgenstein’s point is that there is no gapless foundation to be revealed."

Wittgenstein seems to have spent half his life trying to build foundations and the other half tearing them down. That's fine, but it has next to nothing to do with the rather straightforward, non-foundational question at issue here: Why and how do we feel rather than just funct? And I don't know about other explanatory gaps, but the one at issue here is that one. Generalities about multiplicities of foundational gaps, all over the map, don't answer the rather straightforward question of how and why we feel rather than just funct (any more than specific foundational quantum gaps do).

JS: "The cogito only serves as a reminder that the sentence "I do not exist" is not a valid proposition in our language.  It is a reminder of the rules of our grammar, and not a foundation for knowledge."

To repeat, I said nothing about grammar, nothing about foundations of knowledge, nothing particular about language, and nothing even about whether or not I exist. I just said feelings exist, for sure: And then I asked "how and why?"

JS: "how could one doubt that one had a body?"

Same way you can doubt there's a world, causality, reliable induction, other minds. You'd be wrong to conclude they do not exist, because they're all real enough; but there's certainly room for doubt wherever there are no guarantors for certainty. Descartes pointed out the two exceptions. One (necessary truth on pain of contradiction) was no big surprise; but the certainty of feeling (surely the nether pole of the platonic-personal or objective-subjective spectrum!) was a bit of a jolt. And the upshot was the explanatory gap.

JS: "What could such “doubt” consist in...?"

I have no trouble at all distinguishing the (foolish) sceptics who claimed that the world was an illusion, from the wise ones who simply pointed out that there were some truths one could know with certainty and some truths one could only know with probability. Without Descartes, we might wrongly have thought that the mathematical truths were the only ones we could know with certainty.

JS: "Those are empty words, no different than, 'I don’t have a mind . . . I am just a body'...Repeating them does not constitute doubt, because these words have no discernible consequences.  They are insignificant.  It would make as much sense to say, “all logic is invalid . . . there are no valid inferences,” or perhaps, “there are no thoughts, only words; no feelings, only functions.”  Such mantras are not to be taken seriously."

I'm afraid it sounds to me more as if it is you who are repeating mantras without reflecting on the meaning or grounds for what you are saying: Doubting I feel is self-contradictory (if/when I do feel, and I do). Doubting I have a body is not self-contradictory, just false. Doubting things that are provably true on pain of contradiction reduces both affirmation and denial to empty gibberish. 

JS: "...your argument resembles some unconvincing theistic arguments [such as] God’s existence is self-evident by the very fact of knowledge.  Therefore, a person who claims that God does not exist is begging the question against theism and is denying their own knowledge... Do you find the argument compelling?

No more compelling than that "the Great Pumpkin's existence is... etc." It's just arbitrary gibberish. Please see the discussion of Pascal's Wager. The existence of feelings is anchored in our undeniable experience: gods and goblins are arbitrary inventions of feverish imaginations or charlatans.

JS: "Your argument for a functing/feeling dichotomy is similar.  You claim that the unique status of feelings (be it epistemic or ontological or both) is self-evident, and that it is self-evident by the very fact of feeling.  You defend this notion by accusing those who reject it of begging the question and denying their knowledge of feelings. How is your argument different from those theistic arguments?"

Let me ask you, instead, what plays the demonstrative role of the cogito in the case of hobgoblins?

-- SH