Monday, 19 June 2017

Toomas Karmo (Part E): Philosophy of Perception, Action, and "Subjectivity"

Quality assessment:

On the 5-point scale current in Estonia, and surely in nearby nations, and familiar to observers of the academic arrangements of the late, unlamented, Union of Soviet Socialist Republics (applying the easy and lax standards Kmo deploys in his grubby imaginary "Aleksandr Stepanovitsh Popovi nimeline sangarliku raadio instituut" (the "Alexandr Stepanovitch Popov Institute of Heroic Radio") and his  grubby imaginary "Nikolai Ivanovitsh Lobatshevski nimeline sotsalitsliku matemaatika instituut" (the "Nicolai Ivanovich Lobachevsky Institute of Socialist Mathematics") - where, on the lax and easy grading philosophy of the twin Institutes, 1/5 is "epic fail", 2/5 is "failure not so disastrous as to be epic", 3/5 is "mediocre pass", 4/5 is "good", and 5/5 is "excellent"): 4/5. Justification: There was enough time to write out the  necessary points to reasonable length. 

Revision history:

  • 20170620T0001Z/version 1.0.0: Kmo was able to upload a fairly polished version. He reserved the right to make tiny, nonsubstantive, purely cosmetic, tweaks over the coming 48 hours, as here-undocumented versions 1.0.1, 1.0.2, 1.0.3, ... .  

[CAUTION: A bug in the blogger server-side software has in some past weeks shown a propensity to insert inappropriate whitespace at some points in some of my posted essays. If a screen seems to end in empty space, keep scrolling down. The end of the posting is not reached until the usual blogger "Posted by Toomas (Tom) Karmo at" appears. - The blogger software has also shown a propensity to generate HTML that gets formatted in different ways on different client-side browsers, perhaps with some browsers not correctly reading in the entirety of the "Cascading Style Sheets"  (CSS) which on all ordinary Web servers control the browser placement of margins, sidebars, and the like. If you suspect CSS problems in your particular browser, be patient: it is probable that while some content has been shoved into some odd place (for instance, down to the bottom of your browser, where it ought to appear in the right-hand margin), all the server content has been pushed down into your browser in some place or other. - Finally, there may be blogger vagaries, outside my control, in font sizing or interlinear spacing. - Anyone inclined to help with trouble-shooting, or to offer other kinds of technical advice, is welcome to write me via]

First, an administrative preliminary: I have had to change the previous posting in this (philosophy-of-perception-and-action) series. That was the posting from 2017-06-05 or 2017-06-06, under the heading "Part D". On reflection this week, I find I have to rip out my last two paragraphs. It is not that those paragraphs are untrue. Rather, they are premature, given what I this week realize to be the overall necessary flow of exposition - an exposition requiring, as noted in my "Part C" (2017-05-29 or 2017-05-30) and "Part D" (2017-06-05 or 2017-06-06), first working on some foundational points. I still have to take us through some foundational or prefatory material, being unable as yet to wade boldly into my official topics of perception and action. Worse, within the preliminary, and now rather protracted, foundational work, the two paragraphs just referred to got written a bit too early, and so have for the moment to be chopped out. 

The pair of premature paragraphs should be resurfacing rather soon, perhaps in the next week or so, as the preliminaries get padded out and finished off and tied up. For the moment, I have recorded my embarrassing excision (quoting what I excised from normal-font body text) in the small-font "Revision History", at the top of the now-shortened 2017-06-05 or 2017-06-06 posting. 


With the required amputation performed, the previous ("Part D", 2017-06-05 or 2017-06-06) posting ends simply as follows:

On one of the various possible outrageous Other-Minds parallels to the outrageous Intermittent Furniture and the outrageous Young Universe hypotheses, if I had, counterfactually, asked the sales clerk and the other customer not about pepperoni sticks but about - for example -  east African railways, they would have said the same things, thereupon giving them the air of mere robots playing back mere canned phonemes:

    ME, TO OTHER CUSTOMER: Sir, are you aware that Nairobi
    now has a thrice-weekly service to Mombasa?

    LADY AT CASH, TO ME: I already put your pepperoni sticks into the bag.

    CUSTOMER, TO ME: Yes, these two are my pepperoni sticks.

    ME: I am talking about this year's Nairobi-Mombasa railway launch,
    not about pepperoni, you doofus.

    CUSTOMER TO ME: The same thing.


As I suggested back on 2017-06-05 or 2017-06-06, it is a supposition consistent with all your actual observations that you (the reader) are within the species Homo sapiens the sole possessor of awareness, with all other humans mere mindless, walking and gesticulating, sleepwalkers. My little story from back then, just now reproduced, is a story on which if you were to have said things to your fellow humans other than the many things you have in fact over the course of your life said to them, those humans would have taken on the distressing aspect of mere gramophones, emitting mere canned phonemes. Now I have to look at this general kind of scenario a little further, as I unfortunately neglected to do on 2017-06-05/2017-06-06. 

In so looking, I have to confess a feeling of unease. I have, alas, not only to articulate a position more fully than hitherto, but also to put on public record my worry that my now-more-adequately articulated position may prove, in some way that I cannot now see, inadequate. 


Consider, once again, the 2017-06-05/2017-06-06 robot. 

Suppose first (as "Supposition One") that the cosmos really is rich in causal connections, of the kind affirmed by common sense. In terms of my "Part C" from 2016-05-29/2016-05-30, this is the common-sense supposition that various events do really make other events happen, and that therefore it is FALSE that (to reproduce language from "Part C") for every event E2 supposedly caused by some event E1, E2 would have happened even if E1 had failed to happen. On this "Supposition One", counterfactuals are abundantly supported. To give two common-sense examples: 
  • Had the airship Hindenburg been, counterfactually, filled with helium rather than with hydrogen, the 1937-05-06 explosion which in fact traumatized Lakewood, New Jersey, would not have occurred. 
  • Were the Sun's photosphere to have, counterfactually, a temperature of a mere 100 Kelvins, there would, contrary to observed fact, be no liquid water on Earth. 
Additionally, suppose (as "Supposition Two") that the robot from our "Part D", 2017-06-05/2017-06-06, discussion is so well designed as to render socially plausible not only the conversation you in fact have had with it, but all the alternative conversations you could, counterfactually, have had with it. In that previous ("Part D", 2017-06-05/2017-06-06) discussion I imagined (regretably making things too easy for myself) a counterfactual change in the conversational scenario, exposing the robot as a mere parroter of canned phonemes: 

ROBOT [smiling}: Hi, how are you?

YOU:  Fine, thanks. What is your name?

ROBOT [smile now fading]: Well, yes, this can be depressing.
But I have been getting psychotherapy.

YOU: Don't tell me about psychotherapy: please tell me your name.

ROBOT: I see a lady in private practice.

YOU: "I see a lady in private practice" is not a name.

ROBOT: Oh, on the other side of the Charles River, in Boston.

YOU: That isn't a name either, you doofus.

ROBOT [with a slight narrowing of the gaze, appropriate for the communication ofjuicy social gossip]: Well, as I was explaining just last week to a fine old local family when I got wheeled out to Hyannis Port after psychotherapy, regional accents sound contrived - affected, even - when robots use them.

[and so on]

This week, by contrast, I am imagining not only the entire actual conversation being socially plausible (it started off - to recapitulate - with banter about regional English accents, and morphed into bantering reminiscences of the ancient Rowan-and-Martin television show), but more radically am imagining that any other line of conversation WOULD have proved equally plausible HAD we (counterfactually) ventured on it. So had we, contrary to fact, asked the robot, "What is your name?" things would (so I am imagining this week, in a for-once-properly-radical spirit) have gone on just fine: 

 ROBOT: [smiling]: Hi, how are you?

YOU:   Fine, thanks. What is your name?

ROBOT: Would you like my originally given name, or the appellation that I have now taken as my Name in Religion?

YOU:   Gee whiz, don't tell me you are in Religion?

ROBOT: Well, I must not exaggerate. I am only a Tertiary.

YOU:   You mean, like a Franciscan or Dominican teriary?

ROBOT: Close. I'm a Benedictine Oblate.

YOU [with some irritation]: Oh fine, then. First tell me your originally given name, then your Name in Religion.

ROBOT [frowning slightly, then brightening]: When they first plugged me in here at MIT, I got called Klankton Polymer. But on affiliating myself with the Benedictines at the Archabbey of Saint Vincent in LaTrobe, Pennsylvania, where some nice Opus Dei guys so obligingly carted me in the Opus Dei cube van, I became... oh... er... um...

YOU:   You became what? Don't er and um and prevaricate like that, or I shall begin questioning the reality of mind behind vinyl.

ROBOT [now beaming broadly]: Groccor Buastonius Loquacissimus

YOU:   That's hardly a suitably Benedictine Name-in-Religion. In Benedictine life, one is supposed to become "Willibald" or "Wunibald", or something.  Or maybe "Ansgar".

ROBOT: Well, that's what I said to the Abbot, you know. But he was obdurate ... this was under the rule of the Panzer Pope, you know.... I mean, ai-ai-ai and VOTT can I say? When the Abott goes "Groccor Buastonius" I'm like, totally, Oy vey zmir, oy vey gevalt already - know what I mean?

[and so on]

Under this pair of suppositions, with Supposition One following untutored common sense but Supposition Two radical, is there any longer any logically coherent possibility of doubting that there is a "mind behind the vinyl"? 


Well, in my readiness to duck slings, arrows, and rotten tomatoes, I herewith go on public record as believing that yes, logical room for manoeuvre remains. (At least, I stress, until someone instructs me to the contrary - and being (I stress) admittedly unsure of myself, in these so-dense conceptual thickets.) Even here, I think (tentatively, I stress; being  less than sure, I stress) there remains the logical possibility that no mind is present. I am suggesting, that is, that even on this week's radical supposition regarding the robot's counterfactual capabilities, the meaning our language attaches to the term "mind" is such as to render the no-mind-is-present supposition logically coherent. 

To reiterate this key thesis, in a varying of terminology: although (I) the meaning our ordinary, common, English language attaches to "triangle" is such as to render the supposition of drawing a four-sided triangle logically incoherent, and although the meaning our ordinary, common, English language attaches to "bachelor" is such as to render the suggestion of meeting a married bachelor logically incoherent, nevertheless (II) such is the meaning our ordinary, common English language attaches to "mind" that the supposition of a mindless machine passing not only (as in my writing from the night of 2017-06-05/2017-06-06) all actual but (as in my writing tonight) all counterfactual social-conversation tests is logically coherent. 

Indeed, I suggest, the very supposition that all humans other than you, the reader, are eloquent sleepwalkers lacking awareness, despite their passing all actual AND counterfactual social-conversation tests, is logically coherent. 

I look at it in the same way as I look at the supposition that those wicker chairs and that mirror-topped coffee table exist only when inspected (from installment "C" of this essay, on 2017-05-29/2017-05-30). The meanings, I insisted back in installment "C", which ordinary human language attaches to "wicker chair" and "mirror-topped coffee table" are such as to make the outrageous supposition logically coherent, even when we build it up counterfactually: The wicker chairs and glass-topped coffee table exist when and only when inspected, in the sense that for every time t at which nobody is looking into my parlour both (a) the parlour is empty of furniture at t and (b) WERE, counterfactually, some observer x to be looking into the parlour at t, x WOULD in that counterfactual situation find two wicker chairs and a mirror-topped coffee table. 

(I should, to be sure, have written this point more clearly back on 2017-05-29/2017-05-30; I omitted to spell out the importance of considering both categoricals and counterfactual hypotheticals.)

My point is thus that there presently seems to me (I do stress my unease and uncertainty, as I struggle with dense thickets) no reason for treating the existence or nonexistence of Other Minds differently, so far as the space of logically coherent possibilities runs, from the existence or nonexistence of Uninspected Furniture. 


It must be conceded that our discussion is somewhat bedevilled by the primitive state of Artificial Intelligence (AI) and robotics. 

Here is a marvel before us - a housefly, a buzzing little specimen of Musca domestica, hairy in leg, loud in wing, and bulbous-red in compound eye. As we approach the fly, it takes flight from its bookshelf perch, only to land on a piece of our lunchtime cheese, four metres away. We stride in stealthy irritation toward the now-immobile creature, alarmed by an all-too-plausible imaginative vision of microbes lodged in leg hairs. The creature remains immobile until the very instant of our trying to detain it, beneath outstretched, slowly and quietly descending, palm. In the last possible hundred milliseconds, it cunningly eludes our ambush, now soaring, now diving, now soaring again - tracing first some geometrically easy plane curve, and then venturing, Spitfire-like, along the space curves which are the proper province not of univariate but of multivariate calculus. As we turn away for a moment, the unwelcome forager is back at our Monterrey Jack, proboscis now insolently extended onto the appetizing creamy-yellow surface. 

Can AI labs create anything like this - even, perhaps, as robots confined to the mathematically easy two-dimensional space of a lab floor, rather than flying about in three-space? I am agnostic. Perhaps the feat still eludes roboticists. Perhaps, on the other hand, it has somehow been accomplished, in some such leading-edge school as MIT. If possible at all, the feat must at any rate lie on the leading edge of what is currently possible. 

When it comes to robot conversation, the labs are (surely) nowhere near success. Although I am happy to be advised to the contrary, I suspect things now to differ in degree only, and not in kind, from the ELIZA psychiatrist-simulators that were amusing everyone a generation ago:  

The year is 1982 or 1983. I have become the proud owner of an Osborne 1, for a mere four or five or six weeks' pay. Here is a marvel before our eyes - an 8-bit CPU addressing 64 kilobytes of RAM! And the absence of a hard drive is amply compensated not by the presence not of one but of two floppy drives, each capable of delivering perhaps around one hundred kilobytes! 

I am showing off the "machine", as one in that era so loftily calls it, to a friend, "VUWX", from the Ukrainian diaspora, with his lady companion. As I expatiate on the "machine's"  "word processor" and its "spreadsheet", VUWX assures me that gee, he is sure glad this is not boring. 

Then, however, I launch the LISP-coded ELIZA simulator. It purports to be a robot psychiatrist. 

TELL ME ABOUT YOURSELF, says ELIZA, via the tiny CRT screen (not 80 but a mere 52 characters wide - this particular Osborne 1, an early model within Dr Adam Osborne's full production run, forces its operators into lots of horizontal scrolling). My friend VUWX types in response I AM UKRAINIAN. There is now a prolonged pause, amounting to perhaps around 30 seconds, as the valiant little 8-bit CPU chugs and chugs, processing its troubling, putatively mental-patient, input. Finally its output appears, in those tiny glowing CRT letters: HOW DO YOU SUPPOSE YOU GOT TO BE UKRAINIAN? 

Nobody now has, any more than anyone in 1982 did have, any idea how to design a robot to handle the full categorical-and-counterfactual gamut of conceivable social-conversation tests, of the kind partly sketched on this blog in the evening or night of 2017-06-05/2017-06-06 and more carefully (adding duly robust counterfactuals) sketched tonight. 

Whether it is possible in principle for engineers in Homo sapiens to design such a robot, I do not know. 

(1) One possibility is that yes, it could somehow be done, perhaps upon acquiring a bits-and-bytes-level understanding of the Homo sapiens brain. (This would be an understanding dimly akin to, and yet orders upon orders of magnitude more elaborate than, the understanding molecular-biology researchers claim to have of the DNA genetic code.) 

(2) A different possibility is that it could not be done without a kind of cheating - by taking a living, functioning Homo sapiens brain, and without understanding at bits-or-bytes level what the individual neurons are doing, simply reproducing its myriad possible information flows in electronics, bit for bit or byte for byte. It would be somewhat like cheating on a physics exam: when Bertie Balky copies that differential equation from his examination-hall neighbour Alice Astute, he does not really understand the Free Harmonic Oscillator, much though he may wish his professor to think he understands. 

(3) A still different possibility is that it could not be done at all - thanks to the absence, in mere robotic electronics, of a "Soul". 

The "Soul" can be something of a Catholic superstition. One winces just a little upon reading, in the twenty-second chapter of Gordon Thomas's and Max Morgan-Witts's gripping history Pontiff (Granada, 1983, on the three 1978-era pontificates) the following: 

The Camerlengo is faced with a ticklish problem. His question /.../ about when the pope died is linked to Absolution, the granting of the forgiveness of sins. The much-debated theological point is how long after death may total Absolution be granted. It revolves around the vexing question of how durable a soul can be. There are some Catholics who argue that if a Catholic succumbs following a long wasting illness, cancer for example, the soul might leave the body relatively quickly, possibly within thirty minutes of death. But if a person has been healthy before being fatally struck down, his soul could remain in the body three or four hours, perhaps even longer. To non-Catholics the proposition may appear fanciful, but it can afford great comfort to Catholics.

Still, what can be said for sure? Perhaps, for all that is currently  known, it is impossible in principle to endow a robot with awareness, since a "Soul" is lacking. 

I am of course liable here to attract, once again, a derisory shower of slings, arrows, and decaying veggies. But do consider what we find in secular writers, not promoting  specially Catholic agendas. I recall the account of a naturalist who befriended a number of wolves, I think keeping them in a kind of enclosure - in a kind of paddock, perhaps - close to his house. At the instant the ailing wolf-master died, his pack was observed to burst into a howl, physically isolated though they were from the deathbed. - Well, "was observed" provided, to be sure, that the now-forgotten author I was reading, perhaps from some magazine or other, did write the truth. 

And I recall recreational reading just eight days ago in Jim Robbins's The Man Who Planted Trees: Lost Groves, Champion Trees, and an Urgent Plan to Save the Planet (New York: Spiegel & Grau, 2012). Here is a report, from what seems a reasonably tough-minded professional science journalist, of forest conservationist David Milarch's 1992 near-death experience. Here is what I suspect, without having much wallowed in supermarket tabloids, to be the usual wearisome stuff - the usual tunnel of light, the usual luminous Beings, I think the et cetera with the et cetera. But I do wonder, on the strength of Mr Milarch's and (perhaps especially) Mr Robbins's credentials, whether we might not be forced, for once, to take such a pop-spiritualism thing seriously. 

In general, a denial of the traditional so-to-speak naive-Catholic-peasant conception of Soul is a bit like the denial that God exists. In both cases, the denial is respectable, and indeed is urged in eloquent - to some duly diligent readers, in compelling - terms by such respectable authorities as Richard Dawkins and Christopher Hitchens. Yet in the end what have we here but another (respectable, worthy) Creed, another (respectable, worthy) Faith? 

I remarked on this blog in the evening or night of 2016-04-18/2016-04-19, under the heading "Essay on Green Catholic Hermits", on some MontrĂ©al crutches: How wearisome - baffled non-Catholic people occasionally say, or at any rate in a polite way think - is that repeated Romish fiddling with rosaries and holy water, with incense, with altar cloths, with pilgrimage shrines, with alleged miraculous cures for one ailment and another (one recalls the tens or hundreds of crutches hanging from a ceiling at St Joseph's Oratory in Montreal). The atheist, Dawkins-or-Hitchens, position on the Oratory has to be that of all those tens or hundreds of crutches, each of them - each and every one of them, without exception - is the mute witness to some pious fraud or pious self-delusion: that the number of actual miraculous cures correctly attributable to the prayerful intercession of Blessed AndrĂ© Bessette (1845–1937) and his patron Saint Joseph is exactly zero. It may be so, reply I. But, say I, to affirm it so is to make a leap of faith, no less imposing than the bold leaps Catholics (including me) make on such things as God and the Soul. 

So even hypothesis (3) has to stay in the running, for all we now definitely know. 

Until we can choose on strictly rational grounds among hypotheses (1), (2), and (3), some fog of uncertainty will reduce us to some degree of groping as we work on the philosophical problem of Other Minds. I can only hope that what I have to write in later weeks on this blog, forsaking these present foundational explorations (including the current divagation into philosophy-of-robotics) for my official topics of perception and action, does not turn out to be vitiated by the uncertainties. 

[This is the end of the current blog posting.] 

No comments:

Post a Comment

All comments are moderated. For comment-moderation rules, see initial posting on this blog (2016-04-14).