Author: Stevan Harnad

Inadvertently sending bombs near to babies is tragic, indescribably tragic; intentionally sending bombs from near babies is treacherous, psychopathic and unforgivable.

Is this causal chain too complex for the world to apprehend, in its well-meaning but simplistic calculus of “proportionate response”?

If we don’t read the sinister, cynical handwriting on today’s terror-tech wall, no one’s babies will be safe.

SH: “If you mean that my brain is doing an awful lot of work for me.. for which “I” take the credit, I agree. But you are just what you feel; the rest is as opaque as your respiration in deep sleep.

JE (Judith Economos): I don’t think so any more. I embrace a lot more as “me” — all the behavioral stuff my brain does is me. Some things my brain does — secrete stuff, for example, may not be “me” except in the sense in which my whole body is “me” (as in “you are hurting me”, when what you are doing is hurting my arm, or “They kept me in a cell”, etc.).

When it comes to moral responsibilty, for example for things I do when I am not conscious, that becomes another problem, another special sense of “me”. Consider the various acts you perform while driving (bicycling, walking). They are generally not conscious, but you are responsible for them, and if you hit someone, it is no excuse to say “I was absent-minded and it wasn’t I who did that — it was my brain.” Whereas if you have some palsy, say, and your arm flies up and whaps someone, then you really are not to be blamed, even though it was your arm and your brain that caused the offense.

I agree with all that, for the ordinary business of attributing legal or moral responsibility, but not for explaining (how/why) my brain does all it does and can do. In other words, causation is not really decided in a court of law but in a lab. And if the lab data (if/when we ever get it!) were admissible in a court of law, I suppose everyone would get off, whether they did the crime wide awake or while sleep-walking. In fact, probably the Big Bang would be “responsible” for everything.

Is this inconsistent? No. The appearance of cause clearly plays a causal role. To put it bluntly, courts that acquitted everyone blame-free would increase the crime rate as surely as not having and or not enforcing laws would. The prediction can be tested empirically, many times.

Besides, few people are found guilty for doing something that they did not feel like doing at the time. So verdicts even meet the criterion (I’m not sure it’s “cartesian,” really: I think Decartes would probably plead “nolo contendere” on both the metaphysics and the pragmatics of moral culpability) of distinguishing the intentional deed from the somnambulistic (zombic) one.

JE: But long ago when I was a Cartesian and a Rationalist I accepted only my consious awareness as “me”, so I understand your stance.

It is just that that excludes so much, that when you find yourself doing something you never thought you’d do, you are sandbagged. Too much of what holds together and makes sense of my fragmentary consciousness — what makes me me, lies in the not always concious or even never-conscious parts of my mind. Brain. Whatever. —

I think this doesn’t pertain to the courtroom question of whether I pulled the trigger because I felt like it, but to the lab question of whether the feeling came first and was causal, or was caused by something else…

That comes up in the lab already when I press a button, but not in life, or in court.

JE: Why do you think you are the same person when you wake up in the morning? What if you woke up in a quite different place? Even a different time? You knew nobody; nobody knew you; what makes you the same person you used to be? What, for that matter, makes you the same person from moment to moment?

Because it feels like it — in every case. Whether my feelings are correct is partly a court-room question (do I really have amnesia, or am I faking it? did I really commit or receive child abuse long ago, or is it a false-memory syndrome induced by my psychotherapist?) and partly a clinical neurological question (is this person really X, or does he just imagine he is X?). Separate from both is the lab question: how/why is there this feeling of continuity of identity and what, if anything, is its causal role? (Provisional answer: it has no causal role. Don’t raise the inner counterpart of the “appearances” I invoked above in court: it’s a non-starter in the lab.)

JE: So there are a lot of senses of ‘me’. If you render me permanently unconscious, yet quite healthy otherwise, have you murdered me?

You’ve reduced my body to an inert zombie; if you could reduce it to a dynamic zombie, doing all I do without feeling a thing, that would be a more interesting question. I’d say a brain-dead person (if truly brain-dead: we never know) is gone. For that matter, I’d say someone in a deep but temporary coma — maybe even during deep sleep — is gone too, but comes back!

JE: If you (as the sci-fi folks imagine) copy my working awareness onto a conveniently receptive zombie brain, and destroy my consious body; or use a transporter, which scans my living body as information, sends the information somewhere else, reconstructs an identical living body out of available atoms, and zaps the copy standing here, have you murdered me?

I’ve avoided the sci-fi counterfactuals and hypotheticals, partly because I think they are so uncertain and arbitrary as to be indeterminate (beyond their own premises, about which we can’t even know whether they are “implementable).

(But I’d say the real challenge to the continuity of the felt self is not the one you describe, but the one where the sci-fi exercise simply duplicates it: Then it’s not murder but which one’s the real me? The bodily continuity criterion might staisfy the courts, but you, I think, would be pretty miffed to wake up and be told you weren’t you, but just a clone. You might accept it after awhile, because we are compliant and gullible, even malleable, but it would go against the felt grain mightily, and it’s that felt grain that keeps our mental lives rolling.)

JE: I think dumping the Cartesian self as any sort of whole story of me-dom has a lot to recommend it. I don’t trust the bugger.

But trust itself is a feeling, a perfectly cartesian one. Besides, Descartes is not to blame, indeed not even at issue, in these juridical, clinical and sci-fi scenarios. because he never enfranchised feelings against doubts for anything except doubts that feelings were indeed being felt (if/when they were). What was being felt — e.g., that “I” have a headache, or am on the moon, or that I am you — is certainly not sanctioned by Descartes; only that I am feeling (whatever I am feeling) is, and that’s as incorrigible as it is indubitable (as long as I am compos mentis: a precondition for exercises in cartesian rationality, by the way). And that feeling — that res cogitans, whatever it might be — is the only thing I am entitled, beyond any reach of doubt, to call “me”. For the rest, for whatever in particular it might feel-like I am, all bets are off (no cartesian certainty there!).

JE: I understand Descartes pretty well — not that this is anything to brag about: the cogito is easy to understand. What I am insisting upon is that the res cogitans is at best a momentary thing. Nothing gives it coherence or continuity, and it is not any use at all for creating the world, as D hoped it would be. He needs a truthful God for that, who would not give him a clear and distinct idea that was not true.

And I’m not sure I know what Descartes actually meant (since there was so much nonsense in there too, about god, and that sophistical “ontological” proof), but pretty confident of what he ought to have meant, on the strength of his own premises and reasoning, which is nothing more nor less than that, suprisingly, there is not one but two things we can be sure about: unsuprisingly, the provably necessary truths of maths, plus , very surprisingly, the fact that we feel: Not what we feel, but that we feel. That’s all. It feels coherent and continuous, but that’s not certain, any more than it’s certain that there’s an outside world and that tomorrow will be like today. What we feel can be wrong, but that we feel cannot.

“Clear and distinct” is just rhetorical flim-flam: Maths are indubitable because a contradiction’s a contradiction, and a proof short enough so I can grasp it shows something to be true on pain of contradiction. Feeling’s something else, but equally compelling. One might say it is self-contradictory to suppose I’m not feeling when I am in fact feeling. But the cogito is not a formal deduction, it is a felt fact: My feelings can deceive me about everything but the fact that I am feeling (if/when I’m thinking clearly and distinctly: i.e., if/when I’m compos mentis).

The “I” is equivocal. The indubitable proposition is that there is indeed feeling going on, when there is feeling going on, and to ascertain that, one must be the feeler. Nothing about being it across time, or about being it “coherently”. Feelings are felt, and “I” just means feeling the feelings… And it is necessarily a momentary, instantaneous thing. Once it becomes a memory of prior feelings or instants, it’s again open to doubt. (Did I really feel that, then, or does it just feel that way, now? Is there even a “then”? All uncertainties.)

Trying to hedge those uncertainties with a divine fudge-factor, arbitrarily stipulated as its guarantor of validity, is certainly not Descartes’ proudest moment! It goes from the rigorous to the ridiculous… (Some say he only pretended he believed it, to fend of the Inquisitors…)

So the momentary “res cogitans” is just that, a second indubitable kind of fact, but no more — certainly no guarantor of a reliable world-view, or world…

JE: To say this does not in the least deny the momentary awareness. But there we are stuck. I want more. I want the world, I want the past, I want memory and other people. I cannot be certain of any of it, but the one thing I can be certain of is a barren flicker, not a self, nothing that lasts, not a substance (or if a substance, a brief one, since it has no continuity, indeed no temporal extent at all, by definition no spatial extent, and since there is nothing else there in nothing it stands in relation to).

You got it: You can indeed have all the rest (world, people, memories, continuity, etc.) but not as a matter of certainty, like the other two. (So what? Who said truths had to be certain? They just have to be true, and learnable…)

JE: You cannot say you are you from time to time because it feels like it. Doesn’t guarantee a thing. In fact, the glibness of that whole paragraph is breath-taking. I don’t think it will hold up.

Which paragraph? Rene’s or mine? I never said that anything other than logic plus the fact of feeling was certain.

JE: I was less interested in talking about moral and legal problems per se than in using them to bring in one of many important understandings of “me” . There are increasingly comprehensive uses of the concept, and all of them are worth entertaining. The res cogitans is about the least interesting fictions anybody ever came up with, once you stop dressing it in the familiar clothes of reference and relationship to look like a mind. It is not a mind.

I can’t agree. I think it is the one sure fact about the mind (a fortiori, since it is the second of the only two sureties we have, and it is about the mind!). It is also a very useful bottom line — not only on what we cannot know for sure (the causal truths of empirical science), but also on what we cannot know at all (the causal status and role of feeling).

What I say is that feelings feel as if they have an effect, but we have no idea how or why — and there is zero evidence that they have any independent (telekinetic) causal power, and 100% evidence that other things (Newton, neurons) can and do cause everything we do and can do. So (to repeat): no idea of how or why we feel, or could.

And I’ve managed to say it in the ordinary anglo-saxon tongue, with no opaque jargon or weasel words to lose the simple point in — a point (dare I say it?) that thanks to 10+ years in the raincoat business I am at last daring to see and say is not quite as abstruse and complicated as everyone keeps making it out to be…

Two experiences (e.g., of seeing a stop sign) would be similar. That’s what makes us different from Funes the Memorious: we can selectively forget, and thereby we can selectively abstract what two non-identical experiences have in common, the underlying invariant across all instances — and as a consequence we can categorise, name, and, in general, learn and talk. (If we were stuck in Funes’s world of infinitely faithful rote memory, every instant would be unique and incommensurable with every other except perhaps along a fretless continuum of degrees of similarity (that we would be powerless to do anything about or with).

Harnad, S. (2005) To Cognize is to Categorize: Cognition is Categorization, in Lefebvre, C. and Cohen, H., Eds. Handbook of Categorization. Elsevier.

Real human beings hardly notice and quickly forget the details of experience, and especially the irrelevant details. (Relevance depends on our purposes, and on consequences: like it or not, we have to learn what features distinguish toxic and edible mushrooms, if we need to eat mushrooms to live.)

So selective perceiving and remembering is going on, and must. (That is why there could never be a Funes the Memorious, except perhaps on a passive life-support system — and even then he could never think or talk: he could just feel!)

So, yes, there is a lot more than merely the passive recording of successive, unique, incommensurable instants, all grading continuously into one another. But we can’t take credit for it, because all that selectivity — even the active, consciously learnt part — is basically done for you by brain mechanisms for category learning and detection that we are only beginning to understand and that our armchair phenomenology simply takes for granted without even realising it.

Feelings would be unanalyzable wholistic chunks if we didn’t have a lot of innate and learned selectivity and abstraction going on. But we do.

But let’s not lose sight of the point at issue: My belief P that the distance from (x1,y1) to (x2,y2) = SQRT [(x2-x1)**2 + (y2-y1)**2] (i.e., Pythagoras, your example), or, to take a simpler case, my belief Q that a mouse is smaller-than-a-breadbox.

First, does one specific instance of my believing P (understanding P and feeling that P is true) differ from one specific instance of my believing Q (feeling that Q is true)? Yes, of course, just as one instance of tasting chocolate ice-cream differs from one instance of tasting vanilla ice-cream. And what about two instance of believing P? They differ in the same sense that two instances of eating chocolate ice-cream differ.

Do they have something in common? Sure, all four feelings do: P, Q, chocolate and vanilla. In fact, every feeling has something in common with every other feeling — and something different too. We are tempted to say, however, that some feelings have more in common than others, and that is true, but we should also ask how and why, rather than take it for granted.

(The pertinent exercise here is to remind ourselves of the diagonal argument in Watanabe’s “Ugly Duckling Theorem“: The only reason the 5 small yellow ducklings look more like one another than like the large gray gosling is that our perceptual and computational systems “privilege” differences in size and colour (i.e., weight them differentially in feeling space: Watanabe calls this a “bias.”) If our brains instead coded all differences on a par — spatial position, being closer to duckling X than Y, having the same number of feathers — including, by the way, subtle “higher-order” features such as not-having an even number of feathers, or having the same number of feathers as the middle duckling has epidermal cells, etc. — if all classifiable differences were felt to be on a par, then all ducklings would indeed be infinitely similar and infinitely different, and incommensurable.)

In reality, however, the reason everything does not feel infinitely similar/different to/from everything else is because features are not coded on a par (we are not Funes the Memorious). Some things look, or come to look, more alike than others, some features “pop out,” etc., and this is what allows us to categorize our experiences instead of being awash in an intractable, unnavigable flux.

But none of this resolves the zombie (feeling/function) problem, because our brains could just as well have privileged and abstracted certain useful traits over others in our adaptive interactions with the world without ever having bothered to make any of it — whether tasting an ice-cream or judging whether or not it is true that a mouse is bigger-than-a-bread-box — feel ike anything at all! The feeling, “veridical though it may be” is functionally supererogatory: As always, only the doing capacity is functionally needed, or causally efficacious. The feeling is just a superfluous frill.

(And the mind/body problem is and always was no more nor less than the “feeling/function” problem: why/how are some functions felt functions? That is also the “zombie” problem: Not imagining whether there could be zombies, but explaining how and why we are *not* zombies — without resorting to telekinesis.)

Moreover, I’m inclined to say that, qua feeling, unconnected with differential action, all feelings are incommensurable with the objects of which they are meant to be reflections or representations. At best, they are correlated with them (in that I feel chocolate when I eat chocolate and not vanilla). But I am sure there is a higher-order version of the inverted-spectrum problem that can be resurrected for all of feeling-space (qualia space), according to which no feeling really “resembles” the real-world property it is meant to be a feeling of (or caused by): it only feels as if it is. (Feeling is just seeming.)

This is probably a variant of Wittgenstein’s Private Language Argument: What could it mean to err in one’s subjective feelings? One can act erroneously; one can speak erroneously (saying something is True when it is False, or when others disagree). But what does it mean to say that chocolate does not really taste like chocolate (Dan Dennett’s Mr. Chase and Mr. Sanborn come to mind)? or that THIS feels more like what it really is than THAT? Again, we seem to be back into the incorrigible circle of “feels-as-if-it-feels”: Descartes is right that I cannot doubt that I am feeling what I am feeling when I feel it; but the incorrigibility of my feeling that I have a tooth-ache — even when I have no tooth, and when neurology says I am experiencing referred pain from an ear infection — does not imply anything about any commensurability between feelings and whatever they might be caused by or correlated with or representative of in the real world.

For present purposes, not only does believing that P feel different form believing that Q, but believing that P this time feels different from believing that Q the next time. Thanks to our Watanabe “biases,” we are not Funes the Memorious; we can detect the recurrent invariants, so we can feel that although my P-feeling this time is not identical to my P-feeling that time, they are nevertheless both P-feelings rather than Q-feelings. In other words, yes, we can categorize beliefs just as we can categorize concrete objects.

Yes, feelings are unique, hence different, every time, but no, that doesn’t mean we cannot ignore the differences and selectively categorize and identify them anyway.

But when I challenge the idea that there is nothing that it feels-like to believe that P, I am not trying to suggest that it follows that we cannot abstract and categorize beliefs, just as we abstract and categorize perceived objects. I am only trying to point out that our “cognitive” activity is just as felt as our “perceptual” activity, and that this fact — the felt nature of these functions — is equally problematic in both cases.

It is not possible to bracket the problem of cognition (or intentionality, or belief, or whatever you want to call it) and treat it separately as unproblematic mental territory, zombie-immune. Either it is just as problematic as tasting ice-cream, or it’s not mental (and merely functional, i.e., zombic). Feeling suffuses all of mental life, perceptual and “cognitive”; it is what makes it mental, and problematic. The rest is just unproblematic function: doings rather than feelings.

Let’s compare “twoness” as a perceptual experience (a direct perception of numerosity) with “2+2=4” as a piece of propositional thought. I say both “there’s 2 (things)” and “2+2=4” are thoughts we can have, that feel-like something to have, and mean, and experience (all those are the same thing).

They do feel different from instance to instance, but we can overcome the Borgesian uniqueness with a Watanabean bias, selectively abstracting some “aspects” of the feeling over others. (I doubt, though, that the exercise actually takes place at the felt level: I think our brains do the selection/abstraction for us, then hand the outcome to us on a platter, garnished with the right feeling, and give us a free co-authorship, as if it had been our discerning tastes that had generated the dish…)

I am not adept at phenomenology, so I can’t really introspect and describe what’s happening in my head very well, but I suspect that analysis is unconscious (zombic) and our feelings are after-the-fact, always gerrymandered to square with the computations (or other dynamics) that they did not themselves engender (or constitute: how/why would feelings be computations?)

If tasting chocolate, and recognizing its chocolate while at the same time recognizing that tasting the same chocolate on a number of feelings is always a different feeling too, then exactly the same goes for believing that “the cat is on the mat.” “The cat is on the mat” is a proposition, P. It feels like something to understand what P means (including, no doubt, knowing what a cat is and a mat is, so no doubt images are involved). To believe P is to have the feeling of knowing what P (and its components) means and are, and to feel that P is true. (It feels different to feel that P is false. And, yes, there is something all true propositions have in common, but there is also a lot about them that is different; and different to is believing that P on different occasions.)

The zombie problem, meanwhile, just perdures: How and why does it feel like something to believe that P? Why is it not enough simply to have the datum and just act accordingly, zombily?