Date: Thu, 10 Jun 1999 17:03:33 -0700 (PDT) From: Henry Stapp To: Pat Hayes Cc: Brian Josephson , brings@rpi.edu, brucero@cats.ucsc.edu, chalmers@paradox.ucsc.edu, ghrosenb@ai.uga.edu, hameroff@u.arizona.edu, jmschwar@ucla.edu, keith@imprint.co.uk, klein@adage.berkeley.edu, patrickw@cs.monash.edu.au Subject: Re: Classical vs Quantum Dear Pat, It has been a long time since you sent your last letter. It covered in a compact way the main issues between us. I have been busy on other matters, and had no time to reply. In any case, a six months "cooling-off" period is probably a good thing. The matters you raised have are crucial, and I thought answered satisfactorily by me in earlier correspondence. But you raise them now again in concise ways, and I would like to reply. Perhaps the time lapse will allow me to respond in a way that you will find more convincing. On Tue, 5 Jan 1999, Pat Hayes wrote: > > Hi Henry > > In reply to your message of a few weeks ago. > > >We already know enough about consciousness and about classical mechanics > >to know that the occurrence or nonoccurrence of the painfulness of > >an experience is NOT LOGICALLY ENTAILED solely by the standard principles > >of classical mechanics in conjunction with a complete description of all > >the spacetime trajectories. > > Forgive me if I reply in a similar tone. NO WE DON'T. WE KNOW SO LITTLE > ABOUT CONSCIOUSNESS AND THE MACHINERY OF EXPERIENCE THAT WE HAVE ABSOLUTELY > NO WAY TO JUSTIFY ANY CLAIM THAT ITS PROPERTIES *CANNOT* BE DERIVED FROM > ANYTHING. Ahem. We do indeed have no *theory* of the relationship between consciousness and classically conceived brains: that is exactly the problem. But that does not mean that we know nothing pertinent about consciousness. The structure of our conscious thoughts, as revealed by our examination of these thoughts by further conscious thoughts, is something that we do seem to know something about. (Even this could be challenged. So you might claim that we *know*, for sure, absolutely nothing about anything. But that position is not productive) As Wm. James emphasized, all observation is fallible. Still, our most basic knowledge is about the structure of our conscious experiences: all other knowledge is essentially a theoretical construct that we use to bring order to that most elemental type of knowledge. Even that most elemental kind of knowlegde has some theory built into it. So we are in a boot-strap situation right from the start. Yet we human beings who are reared in our culture, and who communicate with one another, do seem to possess, in the structure of our experiences, certain recurring features, which we can describe . So we are not totally devoid of all scientifically usable knowledge about the structure of our conscious experiences. > Your assertion is exactly analogous to someone who might > have claimed, in the 17th century, that since was no way to explain how > birds flew, this meant that the fact of bird's flying could not be deduced > from physical principles, and that therefore there was no way to account > for flight within the principles of science. Several thoughtful people did > come to this very conclusion, in fact, Descartes among them. Arguments of > this kind contain, I think, a measure of megalomania about them. They have > the general form: I can't see how it could be explained, so therefore it > can't be explained. The missing premis in this, of course, is that if there > were an explanation, then I would know it. The remedy is to meditate deeply > on the possibility that there may be things that even a physicist does not > yet understand. > My assertion is in no way analogous to your example. Your example pertains to a lack of *a physical theory* that will account for the structure of a certain kind of phenomena. But the structure of the phenomena "birds are perceived to fly" is something that is prior to the physical theory that one is seeking at this point. This situation illustrates a key point that I (following Bohr) have repeatedly emphasized: high level theories are based on more primitive theories, which are needed just to allow us the describe to each other the structure of the phenomena that we perceive: we perceive, and describe what we perceive, only with the aid of primitive theories. Classical physical theory accounts for the phenomena of "falling apples" and "the motions of stars", which were descibable prior to the creation of classical physics by using a prior theory: geometry. And quantum theory accounts for phenonema that were describable prior to the creation of quantum theory in terms of the concepts of classical physics and geometry.. So the point is that even though we lack, at the outset, a (high-level) theory of the connection of our thoughts to our brains, that does not mean that we can say absolutely nothing pertinent about the structure of phenomena. WE DO HAVE A POSSIBLE BASIS FOR CLAIMING THAT CERTAIN PROPERTIES OF CONSCIOUSNESS CANNOT BE DERIVED FROM SOME PARTICULAR THEORY> > To return to your above response to my point to Brian. As I have been > pointing out to you now for nearly two years, you must make clear what you > mean by 'logically entailed'. In one, strict, sense you are perfectly > correct, but then in that sense almost nothing (except physics) follows > logically from physics alone. In a (different) sense which allows much of > scientific knowledge to 'logically follow from' classical physics, we > simply do not know - indeed, we *cannot* know, as a matter of logic - > whether facts about consciousness follow from them or not, until we have > some scientifically accepted theory of consciousness. > We have indeed been debating this issue for nearly two years. My answer has been that classical physical theory must be seen as imbedded within a more primitive language and theory that allows us to describe the phenomena under investigation to our colleagues. Classical physics is an "objective" science that accounts for phenomena that we all can observe, and describe to each other, by means of a prior language and theory. It has no physical meaning apart from this link to empirical phenomena. This link is the identification of the geometric features of the theory with the geometric features of phenomena describable in ordinary language. Classical physics acquires, via this link, a close connection to a wide variety of physical phenomena. So when I speak of what `logically follows' from the principles of classical physics, I mean what logically follows from the mathematical formalism together with the principles that connect these formal properties, which are expressed in geometric terms, to the features of empirical phenomena that are expressed in geometric terms. This includes all of what nineteenth century scientist could say followed from the principles of physics. The operation of steam engines and barometers would be included. Science is based on phenomena, but that does not mean that there can be no science until the connection between mind and matter is understood: explaining that connection will be based on earlier scientific theories that do not explain that connection. > (By the way, I think that the correct form of the argument you are groping > for is as follows. We already know all there is to know about our > experience of, say, pain, since the very act of having the experience is > all there is to 'know' about it; any knowledge of a pain must be causally > downstream from the experience itself, which is not itself polluted by > propositional content but is, as it were, acquired whole and unsullied by > interpretation in the very act of raw experiencing itself. (Personally, I > find this account of experience deeply unsatisfactory, but many > philosophers- especially those who talk of qualia - accept it as a given > truth.) On this account, therefore, we already know the full meaning of > 'pain' just by having pains, and there isnt any more to 'pain' than that > experiential acquaintance. While a `pain' is the same thing as a `painful experience', and is *defined* in terms of that experience alone, this does not mean that there is nothing more that can be said about it. One might to be able to say *also* something about the connection of this experience to brain activity. > Now, if one accepts this doctrine, then indeed > one might make a case that facts about experiences can't be derived from > CM, since my rejoinder - that we can't tell if they are or not until we > have a better account of pains - is rendered impotent: we already, on this > view, know everything about pains. (Maybe not about the neural machinery of > pains; but thats a different matter, on this view.). Yes! What we know about `pains' is exactly the same as what we know about `painful experiences': they are the same things. > But now, notice that > this argument isnt about the nature of the vocabulary involved, but is > quite absolute: it says that *no* theory can account for pain without > simply mentioning pain itself. It also applies to QM, since QM doesn't say > any more about pain than CM does. Its not enough to say, but we might be > able to weave an account of pain into QM since its already talking about > "experiences" in some vague sense: it has to be talking about *these* > experiences. I challenge you cite me the paper in Physics Letters where > someone derives a theory of pain from the basic equations of QM.) > The quantum theory of pain is in the process of being developed. Each journal has its purview, and the purview of Physics Letters does not cover this nascent subject. The essential point is that a limitation to geometric features is built into classical physical theory by the dynamical disjunction of the worlds of mind and matter, with a correspondence link only via macroscopic geometric properties, whereas in quantum theory there is potential dynamical link via any property that can be used to describe the structure of an experience. > >The WHERENESS of the "redness" or the "pain" is experientially different, > >---now, and to a child of five---from the "redness" or the "pain" that is > >located THERE. These aspects, geometric and nongeometric, that can be > >experientially distinguished today, are not going to become > >indistinguishable as we learn more: we are not going to become unable > >to see that the "red" spot lies ABOVE the "green" spot. > > > >Classical physics is based on a certain very narrow part of our > >experiential space, and it is designed to place restrictions on our > >experiences in that narrow domain. We can certainly add on extra > >parts: a temperature of 110 degrees Fahrenheit FEELS HOT, to most of us > >under otherwise normal conditions etc. But these extra add-ons are not > >derived from classical physical theory alone. Yet the rate of acceleration > >of a falling apple is! > > Well, lets look at this case in more detail. In fact we know enough about > neural machinery, etc., by now to give a pretty detailed account of what > goes on when our finger touches something hot. Energy is transferred, > average particle velocities in the cell membranes, and then the cells, > increase, eventually some neurochemical reaction's rate increases past some > limit and the resulting phase transition triggers a release of some > neurotransmitter, etc. etc. . They may know all this stuff right down to > the shapes of the molecules by now, and certainly they eventually will. But this does not entail or explain why there is a concomitant FEEL OF HOTNESS: no logical link is established, via the principles of classical physics, between the physical activities of the brain and the `feeling of hotness', and none is possible because the phenomenal/experiential correlates of physical properties in classical physics are confined, in fact and in principle, to perceived geometric features of large aggregations of the microscopic localized entities postulated in classical physical theory to be the only real physical elements in nature. > Since we are talking about very small things here, QM is the better theory > at times, no doubt, though once we are talking about anything the size of a > neuron or larger, the distinction between CM and QM is largely irrelevant, Highly questionable and, I argue, false in this context! > especially at 110 degrees. Seems to me that science may well have an > account of why we feel hot at 110 degrees which is largely classical: There is absolutely no logical implication in classical physics that makes the transition from physical properties to "it feels hot". Classical physical theory is about perceptions of macroscopic geometric properties, and features a theoretical construct based exclusively on microscopic localized entities that accounts for many macroscopic geometric aspects by considering aggregates. > or > better, for which the disntiction between CM and QM is largely irrelevant; > or better yet, for which the CM approximation to QM is almost completely > adequate. > The adequacy of the classical approximation to a quantum system observing itself is not known. (See "Attention, Intention and Will in Quantum Physics" (June 6) for an extended discussion of this point.) > Now, is all this stuff about neurons, neurotransmitters, etc. an "extra > add-on" in your view? Not at all! The extra add-on, relative to classical mechanics, is the "feeling of hotness": there is no logical linkage via the principles of classical physics to *this kind* of perception. > >Mathematics is my game/profession. If you have a theorem that is relevant > >please state it, and the relevant definitions in terms of which it is > >defined, and how it is to be applied to the present case. > > I have already, long ago, cited Beth's definability theorem, but a more > direct and straighforward citation for present purposes would be Craig's > interpolation lemma. First, let me note that any finite set of sentences > can be regarded wlog as a single sentence consisting of their conjunction. > (This keeps things simpler as there is no need to distinguish theories from > sentences.) Now, Craig's lemma is that if a sentence S with vocabulary V > logically follows from a sentence T with vocabulary W, then there is a > third interpolating sentence X **written using symbols from the > intersection of W and V ** such that T |- X |- S; that is, there is a > logical proof of S from T via X. > > The point of this for our present purposes is that if V and W are disjoint- > that is, if the two sentences share no vocabulary - then because of the > highlighted condition, X cannot have any vocabulary at all. Which means > that there is no way to deduce S from T (except in a trivial sense in which > S is a tautology or T a contradiction.). Applied to the present discussion > , let S be classical physics, which, as you have now made vividly clear, > speaks of nothing but the geometry of particles and fields, No! The principles of classical physics are about aspects of perceivable phenomena describable in geometric terms, and a theory of localized particles and local fields that can account, via entailed geometric properties of aggregates, for certain macroscopic geometric aspects of those phenomena. > statement about almost anything else but particles and fields: say, about > consciousness, the color of leaves or the tides. Since V - the vocabulary > of S - does not contain 'color', 'tide', etc.., and since these > observations - T in our example - do not mention subatomic particles or > fields, it is simply impossible for them to *logically follow* from > classical physics. > This statement reveals an element of confusion in your perception of he principles of classical physics, which are precisely a way of making predictions about certain geometric properties of our perceptions via a theory of localized particles and local fields. > I presume that, given your respect for mathematics and the nature of > theoremhood, you must mean something different by the words 'logically > follow'. There is no problem with Graig's interpolation lemma. The language of the pertinent observations is about the macroscopic geometric properties of observered phenomena. The language of the theory include descriptions of macroscopic geometric properties of large aggregates of localized microscopic postulated entities. The basic interpretive postulate identifies these two macroscopic geometric properties. > Indeed, at various times in the past you have suggested various > novel alternatives, but none of them have yet been subjected to the rigors > of detailed analysis, as far as I know. In the meantime, maybe it would be > wiser for you to to avoid phrases like "LOGICALLY NECESSARY connection". > ...... I believe that I have adequately explained what I mean by logically entail, and that it is so transparently clear as to need no further "detailed rigorous analysis". Your illustrations of possible problems have revealed no hint of any difficulty. > >I believe that a great deal of obfuscation has occurred in the discussion > >of "physicalism" and "identity theory" by insufficient care in the > >definition of "physical": the ideas of classical mechanics are taken over, > >but only in a loose and imprecise way, without sufficient rigor in the > >definition of "physical". Clearly, a sufficiently loose definition would > >not exclude consciouness. Indeed, quantum mechanics shows how the physical > >and mental aspects of a "physical" theory of nature can become enmeshed if > >one drifts away from strict CM. > > I suspect that (like in much of your prose) there is a shell game going on > here. You wish to be very strict and rigorous about what is counted as > 'physical' when talking about the limited vocabulary of CM, but > impressionistic and vague when it comes to talking about what you claim > cannot be derived from it. I believe the distinction between the geometric properties (of relative locations and shapes), and non geometric properties (such a `pain' and `yellow') is not vague. > Let me pose you an intellectual challenge. Give > us a sufficiently detailed and precise account of what you think 'logically > entail' means so that you can conclusively and rigorously demonstrate that > facts about consciousness are not entailed by "strict CM". You may use > either proof-theoretic or model-theoretic methods. Then my challenge is > that simply by substituting the expression 'green leaves' (or any concept > not mentioned in the strict vocabulary of CM) for 'consciousness' > throughout your argument, it will become a rigorous demonstration that > facts about green leaves (or whatever) cannot be so derived. > Since there is a clear distinction between phenomenal properties that are defined in the language of geometry (shapes, relative locations, and angular orientations) and phenomenal properties such as `pain' and `yellowness' and `feeling of willful effort' I will claim that the substitution you describe does not have the consequence you claim. > (I can give you a premedial meta-argument why I think this challenge is > likely to succeed. In order to escape it, your demonstration must somehow > distinguish between various kinds of non-CM vocabulary: it must apply to > the vocabulary of consciousness but not to that of the everyday physical > world. But you have no formal basis to make any such distinction, since the > premises of your argument only distinguish the vocabulary of CM from the > rest, making no other distinctions. You must therefore supply this > distinction from elsewhere. (Your recent category of 'geometric' properties > is one such attempt.) But any such move on your part is vulnerable to a > simple denial: I can simply insist that the vocabulary of consciousness > *might* have the property in question. How could you possibly refute such a > counter-claim, since by everyone's admission, we don't yet have a firm > conceptual theory of consciousness?) The vocabulary of consciousness is the vocabulary that we do in fact use to describe the features of our phenomenal/conscious experiences. It includes `painfulness' and `yellowness' as well as `parabolic' and `ellipsiodal'. The former are non-geometric, the latter are geometric. Descartes, the inventor of analytic geometry, perceived an important distinction between geometric properties and the others, and attempted to build a dynamical theory based on the microscopic idealization of the geometric aspects of our macroscopic experiences. Newton, and later Maxwell and Einstein, completed that work. This theory is tied into the same geometric aspects of our experiences whose idealization formed the basis of the theoretical undertaking. It is not enough, in this situation, for you to simply assert that this theory that is geometric in its inception, conception, and application, is nongeometric. There simply is no basis for such a claim, or any way to see how something so clearly based on clear and clean geometric concepts is going to jump outside this natural domain. Certain philosophers have propounded sophistries that try to hide this plain fact by way of analogies, but these are all rendered irrelevant by the fact that the issue in question is a distinction between geometric and nongeometric aspects of human experience, not between two geometric aspects. > > >My approach is to pursue this superior theory that features enmeshment, > >and clarify it, rather than adhere, mindlessly I think, to a false theory > >that has arisen from the exclusion of mind from our theory of nature. > > My point is not to adhere to a false theory, but to question your > oft-reiterated claim that some other physical theory is in a position to > replace it. All the evidence available suggests strongly that the > conceptual language of an adequate theory of consciousness cannot be > reduced to *any* physics, quantum or classical. No amount of vague Bohrian > incantation can disguise the fact that quantum physics says nothing > whatever about the structure of experience; or perhaps, whatever it does > say is utter nonsense (eg. that all of experience must ultimately consist > of noticing the positions of pointers on dials.) > Bohr, if read with care and understanding, said many insightful things. But the essential core of the Copenhagen interpretation, namely that the quantum formalism is about "our knowledge", is the combined wisdom of a whole group of hard-headed physicists that include Dirac, Heisenberg, and Pauli, and this insight was amplified by von Neumann and Wigner: the most rigorously and mathematically oriented of the quantum physicists were very much in the vanguard as regard the recognition of this profound truth. > Best wishes > > Pat Hayes > > --------------------------------------------------------------------- > IHMC, University of West Florida (850)434 8903 home > 11000 University Parkway (850)474 2091 office > Pensacola, FL 32514 (850)474 3023 fax > phayes@ai.uwf.edu > http://www.coginst.uwf.edu/~phayes > > Best regards, Henry