From Aaron Sloman Tue Sep 1 23:25:59 BST 1998 Henry wrote > Reply to Flanagan > .... > The key point that makes the way I speak possible, and fully rational. > depends on "the nature of the beast". If the physical description were a > classical-type description of particles and fields, then it would, in > my opinion, be unsatisfactory to assert that bringing brain into the > physical description also brings in mind. This is because conscious experience > is not logically within the classical conception of the physical universe. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ .... You keep saying this, as if it were obviously true, but every attempt that you make to defend it seems to me (and Pat, and possibly others) to be fallacious. In particular, I think you often fall back on the "nothing buttery" fallacy. Part of the problem is that there are two interpretations of the underlined bit, one of which is correct (see below) while the other is unwarranted. (I think it is false in the second interpretation, but at present all I'll say is there are no adequate arguments to support it.) The sense in which it is correct: Of course our ordinary concepts of "experience", "desire", "itch", "belief", and other mental states and processes cannot be defined in terms of the concepts of physics. (That needs argument, but we agree on it so I'll not bother.) Also if there are any laws of psychology linking these phenomena they will not be derivable from those of physics. (That also needs argument, but I guess we agree on that too, so I'll not bother. It has often been argued by philosophers and others.) The sense in which it is incorrect: There are many high level "virtual machine" concepts which we know can be fully instantiated physically in virtue of careful design of information processing machines, e.g. unification (in prolog), checkmate (in a chess game), addition and multiplication (in most programming language virtual machines), spelling checkers (in many word processors), and so on. (I've given many examples previously.) So we know that there are many kinds of physically implementable states, events, processes which are not reducible to those of physics in the sense that: (i) the concepts which are adequate for describing them cannot be defined in the language of physics and (ii) the laws of their behaviour cannot be logically derived from those of physics. (You can't derive winning configurations or legal moves in chess or Go from laws of physics.) But they are physically *implementable* in the sense that there are physical configurations which *suffice* for their existence, usually very many different configurations which suffice (e.g. different types of computers using different technologies can be used.) In that sense instances of these irreducible concepts are "emergent". (Ultimately this depends on the fact that the implementations require physical "boundary conditions" which cannot be described with full accuracy and generality in the language of physics. For example, we know that there might be future implementations using physical mechanisms that we can't even conceive of yet, as was the situation in the past. So current concepts and laws of physics are obviously inadequate to the task of definition and derivation even if they are adequate to the task of implementation.) So there are things which are logically within the classical conception of the physical universe. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ in the sense that they can be fully implemented in a classical physical universe (certain classes of physical conditions suffice for their existence, even if we cannot characterise those classes with full generality), but are not logically within the classical conception of the physical universe. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ in the sense that the concepts we need to use to describe them cannot be defined in physical terms nor can their laws of behaviour be derived from those of physics (classical or non-classical). (It's arguable that even within the so-called physical sciences there are things like that: e.g. chemistry may in that sense be emergent from physics also, and likewise the organic (physiological) mechanisms of living organisms. I've heard Tony Leggett and other well known physicists make such claims, with good examples.) In that (non-mushy) sense the phenomena are "emergent" though implementable. So, since some things are like that, the question arises whether mental phenomena fit the same pattern. I have not come across any argument that rules this out. Arguments attempting to rule it out either: (a) merely support the conclusion that the mental concepts and laws of mental behaviour (if there are any) are not reducible to those of physics, which I've already granted in the characterisation of "emergence", above. (I.e. the arguments merely support a kind of non-reducibility which is not in dispute (by me).) or (b) merely *assert* that mental phenomena are sui-generis, without any argument, or use arguments starting from theological or metaphysical premises (e.g. dualist theories) which are themselves at least dubious. (I.e. they START from premises which cannot be used while lots of people reject them.) or (c) presuppose a form of concept empiricism (all concepts are fully grounded in and abstracted from experience of instances) and start from the claim that we have some kind of direct intuitive understanding of what experience is because we have experiences ourselves, and therefore our conception of the nature of experience cannot be mistaken and therefore since we KNOW that experiences do not require anything else (especially any physical mechanisms) to produce them, their existence cannot emerge from any physical configuration. (This sort of claim, with minor variations, tends to be bolstered with lots of stories about things that are allegedly imaginable, e.g. your experience surviving after the total destruction of your body, flipping qualia, zombies with all the internal information processing virtual machines that we have but without any experiences, etc. etc.) The attraction of (c) is very powerful, among thinkers of many kinds (including those as different as Penrose, Searle, Chalmers(??), Stapp(??) nearly all first year philosophy students and many professional philosophers). Wittgenstein discussed at length, but inconclusively, this belief in the power of "private ostensive definition" in his Philosophical Investigations. He saw correctly that it was illusory but did not manage to assemble powerful enough counter arguments: for that task he needed a theory of mind, including mechanisms of concept formation, but he was against theories in philosophy! For now all I'll say is that however powerfully all sorts of intelligent people are drawn towards (c) that does not make it correct. I've argued at length (but still inconclusively) elsewhere that we are simply mistaken in thinking we all have a clear intuitive, reliable, idea of what we mean by "experience", "consciousness" etc. and compared it with other cases where people thought (wrongly) that they understood some concept clearly, solely on the basis of experience of instances (continuity, simultaneity, causation, etc.) E.g. it's one thing to know that something happens now here, but something happening now and far away is another matter. Likewise it's one thing to notice a particular kind of itch in your foot, but wondering whether exactly the same kind of itch occurs in another person's foot is a different matter. The first is a useful application of a biologically evolved information processing system with various kinds of self-monitoring. The second is not, though a subtle variant of it might be part of good medical procedures. [My arguments will also be inconclusive until more details of the architecture of mind have been worked out, to show precisely how these self-descriptive concepts actually work and how they lead both to biologically useful processes and to philosophical confusions. The same will occur in human-like robots. Convincing people may require something more like therapy than like argument, because of the way philosophical convictions establish themselves, alas.] Henry's position seems to be a combination of (c) plus the additional hope that if he takes this irreducible concept of "experience" and combines it with those of quantum physics he can patch up quantum physics. My claim is that if he did succeed in adding something useful to quantum physics his use of words like "knowings", "experiences" when doing physics, would be as much of a *pun* as the existing uses of "string", "charm", and perhaps "spin" and "particle" already are, even if he THINKS he is using our ordinary concepts of mentality. Partial self-delusion regarding our own mental states and processes, including our own concepts of mind, is an inevitable consequence of (a) the need for useful self-monitoring mechanisms (like all perceptual mechanisms) to abstract away from details, and (b) the impossibility of any perceptual mechanism having direct and full access to all aspects of the things being perceived. Cheers. Aaron === Aaron Sloman, ( http://www.cs.bham.ac.uk/~axs/ ) School of Computer Science, The University of Birmingham, B15 2TT, UK EMAIL A.Sloman@cs.bham.ac.uk Phone: +44-121-414-4775 (Sec 3711) Fax: +44-121-414-4281