From: Bruno Marchal <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] Subject: Re: Determinism Date: Fri, 16 Jan 2004 12:18:40 +0100

At 22:25 15/01/04 -0500, Doug Porpora wrote:

There have been two main reductionist strategies to deal with mental states, and they both -- to say the least -- have stalled. The two strategies are:

1. Eliminative materialism 2. Identity theory

Well, that is the two *materialist* strategies. Buddhism, Hinduism, Plato, Berkeley and many others have proposed what we could

call Eliminative spiritualism or immaterialism, sometimes called "objective idealism" (sometimes confused with subjective idealism, i.e. solipsism). I think the 20th century (perhaps by the dazzling success of physics) is the first (and probably the last too!) during which

immaterialism has no more been defended (with some exceptions).

My work shows that comp entails a form of Eliminative immaterialism, and also that comp (but actually this works for Everett formulation of QM too) is incompatible with identity theory. We can still ascribe a mind (a first person with her private life) to a body (a third person describable object), but we cannot ascribe a body to a mind, we can only ascribe an infinity of "possible bodies" to a mind (like our counterparts in the QM parallel "universes": i.e. the mind body relation is not one-one.

In this setting the word "reductionism" is misleading: comp *reduces* both minds and bodies to arithmetical relations between numbers, but

it leads also to a machine psychology (or theology) which is everything but reductionist, for that psychology/theology is a sort of

"negation theology" (to use the theologians expression) in the sense that the machine self is able to "diagonalize" against any proposed and well defined definition of the self. The real self being necessarily undefinable by anyself. This looks weird and perhaps even somehow mystical, but can be related to the incompleteness phenomena, even easily so once we work a little bit on the modal logic formalization of those incompleteness phenomenon (cf Solovay's work).

A lot of problems in the "modern" philosophy of mind are due to a reductionist conception of machines which is just impossible since Godel&Al.

In some sense Godel's theorem is the big killer of any reductionist conception of machines. And as I said, some quasi dogmatic attachment to materialism (even in the weaker sense of a doctrine reifying matter) will not help.

Of course, just to mention them, there are also the dualist approaches (cf "the brain and its self" by Popper and Eccles, or even Chalmers).

Is Chalmers really a dualist? Although he does label his views this way at times, from his writings he does not seem to believe in "matter" per se, rather he thinks the fundamental stuff of reality is likely to be something like "information" which has both an objective description (a particular bit string, computation, whatever) and a subjective what-it-is-like to *be* that bit-string/computation/whatever.

Is Chalmers really a dualist? Although he does label his views this way at times, from his writings he does not seem to believe in "matter" per se, rather he thinks the fundamental stuff of reality is likely to be something like "information" which has both an objective description (a particular bit string, computation, whatever) and a subjective what-it-is-like to *be* that bit-string/computation/whatever.

`It seems to me that any formal mathematical theory of consciousness or of "observer-moments" must work the same way. If you want to have a mathematical theory that assigns measure to different observer-moments, for example, you need to have a mathematical framework for listing all possible observer-moments, perhaps something like treating each distinct computation (or any finite sequence of steps in a distinct computation) as a distinct observer-moment. And yet, even if I understand this mathematical framework, "from the inside" I will not be sure which of these formally-described observer-moments corresponds to my own current experience, the "qualia" that I am percieving at this moment. So just as in Chalmers' system, there is a difference between the "objective" mathematical description of an observer-moment and the subjective "what-it-is-like-to-be" of the observer-moment corresponding to that description. There's a case for calling this "dualism", but also a case for labelling it as a monist theory, an "eliminative spiritualism" as you described it (although I'd prefer the label 'eliminative idealism', since 'spiritualism' has mystical connotations).`

But I don't think a lot in this list adhere to dualist positions, but please correct me if I'm wrong).

I think there are people on this list who *implicitly* hold dualist positions. There are a number of people who would use the following sort of procedure to find the first-person likelihood of experiencing a universe with a given set of properties:

I think there are people on this list who *implicitly* hold dualist positions. There are a number of people who would use the following sort of procedure to find the first-person likelihood of experiencing a universe with a given set of properties:

`1. First, find a measure on all "universes", regardless of whether a given universe is capable of supporting complex observers`

`2. Then use the anthropic principle to take into account the idea that you're more likely to experience a universe with lots of observers than one with few or none, assuming each universe's measure is equal`

`(see, for example, Hal Finney's post at http://www.escribe.com/science/theory/m5006.html on how to find the likelihood we will find ourself in a universe with no other intelligent life within communicating distance)`

`If this is just taken as a heuristic procedure, in lieu some more fundamental procedure that does not involve two separate steps, then perhaps it need not be labeled "dualist". But if this is really seen as the way the ultimate "theory of everything" would work, with no more fundamental theory to be found, then I think such a view is committed to a fundamentally dualist metaphysical view. Since I find dualism inelegant but I do think the anthropic principle has to be taken into account somehow, I prefer a TOE which only involves a measure on observer-moments rather than "universes", with this measure determined by a theory that already takes into account the anthropic principle somehow (see my posts on the 'Request for a glossary of acronyms' thread at http://www.escribe.com/science/theory/index.html?by=OneThread&t=Request%20for%20a%20glossary%20of%20acronyms for some speculations on what such a theory would look like).`

`Jesse Mazer`

`_________________________________________________________________`