Newsgroups: alt.consciousness,comp.ai,comp.ai.philosophy,sci.cognitive,sci.philosophy.meta,sci.logic
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!godot.cc.duq.edu!news.duke.edu!news.mathworks.com!newsfeed.internetmci.com!in1.uu.net!utcsri!utgpu!pindor
From: pindor@gpu.utcc.utoronto.ca (Andrzej Pindor)
Subject: Re: The physical basis of consciousness
Message-ID: <DqC1E7.6n5@gpu.utcc.utoronto.ca>
Organization: UTCC Public Access
References: <winter-0704962115440001@pm1-20.primeline.com> <4lectg$r8i@newsbf02.news.aol.com>
Distribution: inet
Date: Tue, 23 Apr 1996 20:39:42 GMT
Lines: 387
Xref: glinda.oz.cs.cmu.edu comp.ai:38425 comp.ai.philosophy:40557 sci.cognitive:12472 sci.philosophy.meta:27166 sci.logic:17942

In article <4lectg$r8i@newsbf02.news.aol.com>, Den1P <den1p@aol.com> wrote:
>pindor@gpu.utcc.utoronto.ca (Andrzej Pindor) wote:
>>>    While it is certainly true that we often can and do make statistical
>>>predictions in qunatum theory, the ability to do so there is largely
>>>dependent on the linearity of the equations used.  It is less obvious
>that
>>>we can make useful statistical predictions in the non-linear chaos
>>>theoretic cases.  We can certainly characterize in a qualitative or
>>>semi-quantitative way some general system features - strange attractors,
>>>threshholds of instability, and so on - but this are much less
>informative
>>>than even the statistical predictions of quantum theory.
>>
>>From what I know about these matters (and from what I heard Priggogine
>say), 
>>there is 'order' in chaos, i.e. we can even quantitatively characterise
>output
>>distributions from input distributions, even in quantum case.
>    This is certainly true.  Many consequences of micro-dynamics do
>penetrate cahotic dynamics to re-emenge at higher levels of integration. 
>Note that my anti-reductionistic stance does not deny this.  It simply
>questions the dogma that _all_ macro-phenomena are the dynamical
>consequents of the dynamics of the micro-constituents _studied in
>isolation_ from the macro-system.  In physics, we use perterbation theory

I think that you have a rather dogmatic view of reductionism. It may be
characteristic of some philosophers, but not of many practising scientists.

>to expand the generator of time-development (the Hamiltonian), thus:
>
>   H = Ho + Hi
>
>and approximate the solutiion thus
>
>   |psi(t)> = |psi-0(t)> + |psi-1(t)> + ...
>
>Where Ho is the hamiltonian in the absence of the interation potential Hi.
> One then solves the equations using the approximation
>
>    H = Ho   -> |psi-0(t)> = exp(-i Ho t) |psi-0(t)>
>
>Generally or hopefully, psi-0 dominates the solution.  Then, one
>calcualtes a correction,  |psi-1(t)>, due to the fisrst order effects of
>Hi, thus:
>
>    |psi-1(t)> = exp(-i Hi t) |psi-0(t)>
>
>Thus, the solution is dominated by the time-development in the absence of
>Hi with small, but interesting, effects due to Hi.
>   The moral here is that the fact that the non-interaction micro-dynamics
>have macro-consequences that survive the cahotic deduction black-out is no
>reason to assume that Hi is zero, i.e. that there are n emergent effects.
>...
Who assumes this? In fact there are many cases when it is known that
perturbation expansion is invalid and you have to solve exactly the problem of
Ho+Hi and the states of such system cannot be mapped to the states of Ho.
Superconductivity comes immediately to mind as one example. I have nothing
against calling the s/c properties as emergent properties (and often do so
myself), but it is still true that the properties of the 'emergent' system can
be _reduced_ to Ho and Hi and traced to the properties of constutents and their
interactions.
..............
>>
>>If we get the same predictions and they agree with empirical data, then
>models
>>are equivalent at this stage, so what? You cannot distinguish two
>potential
>>fields differing by a constant either.
>    They are equivalent in whatever prediction they make, but the
>reductionist dogma is not confirmed, nor is it falsifiable.  Further, it
>is pragmatically wothless!

Again, 'reductionist dogma' is mainly in the minds of philosophers and not
practioners of science. Hence I do not see what you mean by 'reductionist
dogma is not confirmed'. And dogma in itself is always contradictory to
a pragmatic approach.

>    Now contrast the hylomorphic or heirachical approach.  Its claim is
>that there may well be new, logically emergent, intelligibility at higher
>levels of integration.  Thus, it is worth studying phenomena at there own
>level.  If biology or psychology can not be deduced from physics, what use
>is strong reductionism?

First of all, how do you know that they cannot be? Perhaps we cannot deduce
them at the moment, but this is not the same as saying that they cannot be.
Secondly, no one objects to studying phenomena at their local level, but
this does not mean that it is impossible or not allowed (?) to investigate how
these phenomena arise from the lower levels. Studying superconducting
phenomena is not a reason to abandon a search for the ways how these phenomena
arise from the lower level properties of the materials.

>     In fact we do study things at their own level and learn
>scientifically interesting things that are not deducible.  Thus the

What you mean "are not deductible"? Can you provide a provable example of
such situation?

>Aristotelian model, with its dynamically emergent forms, and heirachical
>atructuring of matter and form, more closely fits the logical structure of
>science as it actually has been done, and now, with the admitted failure
>of deducibility, it also fits the theoretical foundations more closely
>than any form of mechanistic reductionism.  Let's come into the

Admitted in what sense? Perhaps in your very narrow view of decuctibility.
And what is "mechanistic reduction" (as perhaps opposed to a "non-
mechanistic")?

>post-Enlightenment period and get past our prejudices.
>   From a logical point of view whatever is deducibble from principles is
>logically dependent on those principles.  Whatever is not so deducible, is
>logically independant, and must be known, if it is known at all, by direct
>investigation.  The logically emergent part of human nature, called by
>Aristotole the psyche or soul (a term broader than nous), is such an
>object of enquiry.
>
I have nothing against direct investigation, and I am certainly opposed to 
a view that we can learn about nature just by sitting in an armchair and
thinking, i.e. doing logical deductions. This does not however mean that
our investigations should not concern the problem of how phenomena at one
level lead to phenomena at the higher level.
....................
>>Chaos theory tells us that this has to be
>>accepted even for classical systems. It still remains true (at least in
>the
>>modified reductionist sense) that the output distribution can be
>determined 
>>from knowledge of input distributions. 
>   The is not clear at all!  In the case of the mandelbrot set, the
>mapping is so fractured in the boundary region, that it is not clear that
>sensible distribuion functions can be derived. As I said in my last post,

Mandelbrot set itself is such a dsitribution. It is a very narrow point of view
to demand "sensible" functions, whatever it means. The point is that this
distribution is computable with arbitrary precision.

>the ability to do this in QM is largely a consequence of linearity -- and
>that does not apply here.

Even for non-linear systems you can determine what output distributions come 
from given input distributions. Ability to do so numerically is just as good.
 
>   Suppose that consciousness is a dynamically emergent feature.  What
>kinds of distibution functions would shed light on it?  I might be able to
>calculate distibutions of behavioural response, but sociologists can do
>that to a certain degree now without sheding a scintilla of light on
>consciousness per se.
>
Since there is no agreement what consciousness is, speculations like above
are meaningless.

>>>the possibility of a dynamical sense in which macro-properties may be
>said
>>>to follow from micro-properties, i.e. even though we cannot deduce the
>>>macro-properties, still there is no more physics in macro-systems (such
>as
>>>organisms) than there is in their constituents taken is isolation. 
>>
>>Exactly.
>>
>>>Clearly, in the absence of logical reducuibility, such a position is not
>>>scientifically confirmable or falsifiable.  It is merely a philosophic
>>>stance or a faith tradition.
>>
>>True, the same way as the traditional reductionism. However, one can say
>>that it is justified by Occam's razor - as long as there is no need to
>invent
>>additional axioms, we stay with what works.
>    Works how?  Motivation asside, I know of no instance of the
>reductionistic stance entering in an essential way into any scientific
>conclusion.  The only thing it can do is close use to the possibility of

I am not sure what you mean by the last statement. By "what works" I meant
that we have found out in many instances that the properties of a higher
level are reducible to the properties and interactions of the constituents
at the lower level. I already gave the example of supercoductivity and 
the most typical one is thermodynamics vs. statistical physics. In no case
we know of a principal reason why this could be impossible, certainly not
at quasi-macroscopic levels you are talking about (biology for instance).

>multi-body force laws, new forms of interaction, and other emergent
>features.  Tossing reductionism results in no loss of analytical power. 
>If a real reduction can be done, there are no significant emergent
>features.  Good.  If it cannot, then there may be.  Good again.  If there

It looks to me like a matter of vocabulary - what you call _real_ reductionism
or _significant_ emergent features. There is no reason to get emotional about
words.

>may be, we will look fo them, but if we exclude them a priori, we will not
>look.  So how is science better off?  I think with increased openness and
>respect for the legitimacy of both high level (e.g. biology) and low level
>(physics and chemisty) approaches to reality.

Who is suggesting excluding for instance multi-body forces? We know that 
they exist. If this is how you understand reductionism then you are attacking
a strawman.

>  From another point of view, however, it seems that dynamical
>reductionism, in the sense that there is no more physics in macro-systems
>(such as organisms) than there is in their constituents taken is
>isolation, is meaningless.  Physics is a science and so the physics in

As I said you are erecting a strawman. Even in case of two-body forces,
investigating a consituent in isolation will not be enough. In fact it is
meaningless to talk about investigating something in strict isolation, since 
to investigate an object you have to make it interact with other things.
Consequently, reductionism as you seem to interpret, it is logically 
inconsistent.

>anything is, in a very real sense, what we can know and deduce.  If this
>is what we mean by physics, then dynamical and logical reductionism are
>identical.  If we mwan more than this, we are saying something that can
>never be known by human reason, and both my own, Thomist, tradition and
>the postivist tradition are of a mind that such a postion can ony be a
>faith statement.
>...
I am not sure what is 'dynamical' reductionism. To 'know' we also have to
'experience', i.e. do empirical work. Logic by itself is not enough for
science. It looks to me like you are atacking more strawmen.

>>As long as there is no counter example, pragmatism (as expressed by
>Occam's
>>razor) seems to be preferable. We should of course keep our eyse wide
>open,
>>this is the essence of doing science.
>     But there are counter examples, which I mentioned, to wit:
>>>theories of nuclear binding predict multibody forces over and above the
>>>sum of all the two body forces taken in isolation.  Further, the
>>>gravitational effects of multiple particles is not simply the sum of the
>>>gravitational effects of the particles taken separately.
>...
>>>it gatuitously assumes that all forces are two-body forces, but meson
>>
>>Such assumption simply indicates ignorance, why bother talking about it?
>  Because the whole reductionist position is the reification or projection
>of human noetic limitations in nature.  It is the ignorant assumption that
>nature is as simple as the isolated systems we have managed to comprehend.

More strawmen.

> If there were no ignorance on these points, no one would be a
>reductionist.

Again, your view of reductionism is perhaps characteristic of some 
philosophers, but hardly can be said to characterize attitude of scientists.
>...
>>All true, but we proceed by making most economical assumptions - we know
>>that in many cases the matter of our inability to deduce behavior of
>systems
>>is caused by the sheer complexity and when we have many examples that
>when
>>we manage to overcome this problem the said behavior becomes
>understandable
>>(or we discover that there were factors we did not know about like an 
>>additional planet perturbing the orbit of Neptune, for instance). We will
>
>>assume multibody interactions when we cannot in any way explain the said
>>behavior by the forces known so far.
>    That is fine, but it is not the reductionist stance.

Then why attack a strawman?

>>>   Whereas it is often aserted that reductionism is an essential axiom
>of
>>>science, in practice only a very limited form of reductionism is ever
>>>used.  It is only when two areas of study apply to a common object that
>we
>>
>>I disagree. Science looks for underlying reasons of phenomena, does it
>not?
>  Yes it does.  Aristotle says science is the understanding of things in
>terms of causes.

Good for him, but he also said many incorrect things.

>>Every time you ask "why?" (or "how?") expecting to find underlaying
>causes, 
>>you practice reductionsism. Isn't science this search for "why?" nad
>"how?"?
>   No.  If I understand the cogency of an argument in terms of the
>validity of its reasoning processes, I have not thereby reduced logic to a
>more fundamental science.  If I understand an ecosystem in terms of
>competition, cooperation, predation, species diversity, etc.  I have
>grasped valid causal factors without making ecology a branch of physics or
>chemistry.  Causal explanation is one thing, and reductionism quite
>another.
>...
Branch or no branch is irrelevent. The more important point is if the higher 
order phenomena and laws are independent of the lower level phenomena or
are they a _result_ of interactions at the lower level? It is still convenient
to talk in terms of temperature or pressure even if they are a result of 
the distribution of particle velocities.

>>> In fact, even the deduction of Newtonian mechanics from
>>>classical quantum mechanics is not rigorous, but merely suggestive. 
>>
>>Yes, we do not very well understand the transition from quantum to
>classical
>>regime, but it does not mean that we do not try. Are you suggesting that
>we
>>do not concern ourselves with this? It is an element of science that we
>ask
>>"how" this transition happens (and in recent years thae answers started
>to
>>emerge).
>  No, I am merely pointing out that the exagerated promises of
>reducrtionism are very hard to make good on.  We can't even do it within
>the boundaries of physics, so it is rather absurd to claim that we can
>reduce all other sciences to physics.  If we cannot even deduce Newton's

'Physics' spans phenomena from a huge (in fact infinite) range of scales.
That we have diffficulties at the limits of the presently observable ranges
is not surprising.

>Second Law from classical quantum mechanics, what hubris it is to claim
>that we can reduce human behaviour (which depends on N-2) to fundamental
>physics, let alone consciousness.
>...
Presumably human behavior is driven by macroscopic phenomena (although some,
like Penrose, think otherwise). If this is true then there does not seem to be
a fundamental reason why this should not be possible, at least in priciple
(and in the sense of understanding how distributions of inputs transforms into
a distribution of outputs). Consciousness is different mainly because we
do not have a good idea what it is. Untill we do it is pointelss to 
speculate. If we accept Dennett's explanation (which I think has a lot of
merits), even this could be explainable.

>>>  Thus, gene sequences are correlated to
>>>physiological characteristics, and brain scans are correlated to
>>>behaviorial characteristics, but in neither case is the one deduced from
>>>the other.
>>
>>True, but this is because these links are very hard. 100 and more years
>ago
>>chemistry has been studied in isolation from physics too. Are you saying
>>that the correlations between brain scans and behavior is something which
>>cannot be understood? I do not understand your point.
>  They may be hard, but they are more likely impossible.  In either case,

This is a statement of faith, at least as much as the opposite view.

>strong reductionism makes no logical contribution to actual science --
>that is the point.
>
This 'strong reductionism' you seem to be presenting is a philosophers
invention, let's them fight over it themselves.

>>>   Reductionism is a good motivational stance, serving to motivate
>>>research and increase the unity of science, but it fails as a basic
>truth.
>>
>>I am sorry but your last sentence follows from your examples the same way
>as
>>someone saying that all this claims of progress in medicine are crap 
>>because his/her friend died at the age of 20 inspite of being treated by 
>>the best specialists.
>     I am afraid that you miss two points: (1) Unlike medicine, which had
>many successes to counter balence its failures, reductionism has none. 
>(2)  When I say "it fails as a basic truth," you have to undersatnd that a
>basic truth is a principle from which others logically follow.  Nothing
>logically follows from reductionism.  When it works it works, and when it
>does not work, we simply invent a new form of interaction or cite
>difficulties of computation.  In short, it does not have the invariant

The above applies to any ideology and I certainly oppose forcing science into
any ideological straightjacket.

>logical power requires of a basic truth, like its Aristotelian replacemet,
>the principle that whatever happens has a cause.  (Note that this last
>applies event to unpredictable quantum events such as decays, because (1)
>the claim is not one of predictability, and (2) the various distinct half
>lives show that there is an undelying principle of regularity for each
>nuclear and elementary particle species.)
>
The claim which is not the one of predctability is totaly useless - does not
tell us anything of empirical importance. I also do not understand your
point (2) - what principle of 'regularity' are you talking about and 
how would this show that there is a 'cause'?

>Dennis Polis

Andrzej
-- 
Andrzej Pindor                        The foolish reject what they see and 
University of Toronto                 not what they think; the wise reject
Information Commons                   what they think and not what they see.
pindor@breeze.hprc.utoronto.ca                      Huang Po
