Re: Significance (contd)

On 16 Aug, 14:33, Ymir <inva...@xxxxxxxxxxxxxxx> wrote:
In article
<5832114a-c4e9-440d-bb0c-5e813f74d...@xxxxxxxxxxxxxxxxxxxxxxxxxxx>,





 someone2 <glenn.spig...@xxxxxxxxxxxxxx> wrote:
I had said:
------------
I'm sure they could determine it sufficiently accurately to state the
outputs, by working it out from schematics such as logic circuits
where the logic actions are broken down into AND, OR, and NOT gates,
and circuit diagrams, providing they were sufficiently detailed so
that no experimentation was required to fill in any omitted detail.
Such that if the calculator was experimented on, the results could be
explained to be the only expected results for the experimentation
given the schematics.
------------

To which you replied about the logic gates:
------------
Here you are not attempting to derive its behaviour from the rules
contained in C, but rather from a version of functionalism.

You're also assuming that the researcher understands that they are
dealing with an electronic device and that the behaviours which they
are trying to explain are limited to what will happen when currents at
specific voltage levels are applied to specific portions of the
device.
-------------

It isn't a philosophy of mind version of functionalism. Just to be
clear so as no confusion should exist between you using the word
'functionalism' there, and the numerous other times it has been used
during the conversation.

There can also be layout diagrams and blueprints and other schematics.
The behaviour they would be trying to predict wouldn't be limited "to
what will happen when currents at specific voltage levels are applied
to specific portions of the device"

My point was that a schematic showing the logic gates isn't going to
tell you what happens when you hit the robot with a bat -- they only
tell you about one aspect of the machines behaviour, namely how it will
process signals.

It may be that that processing is the only aspect of its behaviour
relevant to whether it is conscious (that is essentially what
functionalism says). It may be that it is not the only aspect of its
behaviour relevant to whether it is conscious (that is what your
identity theorist would claim).







I had said:
-------------
We do know that both the functionalism of the theory B, and the
identity theory of step 2, would predict X, since they make no
predictions as to what those laws governing behaviour would be, nor
would conflict with whatever they could be discovered to be. They just
take whatever the known laws of physics governing the behaviour of
things that it isn't anything to be like are at the time. Which is
why, if we consider in our thought experiment that behaviour X could
be worked out, just to consider the implications, we can see that both
the functionalism of the theory B, and the identity theory of step 2,
expect behaviour X, simply because they both take whatever the known
laws of physics are at the time. If you disagree, then state which
suggestion you think would conflict with behaviour X.
-------------

To which you replied:
-------------
An identity theorist who didn't consider the robot to be conscious
would predict that it wouldn't exhibit behaviour X, even if the
functionalist could show them that it instantiated all of the
functional relationships necessary (according to the functionalist) to
produce consciousness. The functionalist, on the other hand, would
predict that it would exhibit behaviour X.
-------------

You seem to be claiming that if the identity theorist could predict
from the schematics, and by using a simulator or other means, that the
build would perform behaviour X, that they would expect other than
behaviour X. This simply isn't the case. You're wrong. They always
expect the build to follow the known laws of physics.

The point your missing here is that unlikely your omniscient theorists,
real world theorists don't have the ability to predict behaviour
directly from the laws of physics. They assume that the behaviour
ultimately derives from these laws, and, as physicalists, they assume
that these laws are the only ones which ultimately exist, but they can't
actually work out exactly what applying those laws to a complex system
will result in.

So instead they come up with higher-level explanations which attempt to
model the behaviour of the robot. Both have come up with different
models, and both assume that those models are consistent with the laws
of physics, but they can't actually derive those models directly from an
application of low level laws -- that would constitute methodological
reductionism, not physicalism, and I don't know of anyone who takes
methodological reductionism as a realistic approach to understanding the
universe.

Functionalism makes certain predictions.

Identity theory makes certain predictions, though not necessarily the
same ones. In particular, the identity theory which your theorist
advocated denies the possibility that the robot could pass the turing
test, whereas a functionalist who asserts the robot is consicous claims
that it will.

Both assume that these predictions are what the laws of physics will
ultimately lead to. At least one of them is, of course, wrong. But
neither of them has the ability to actually provide a reductionistic
proof that their predictions are what indeed follow from the laws of
physics.

When you magically endow them with the ability to actually derive the
robot's behaviour from lower level physical laws, then one (or both) of
these two theorists is going to come to the realisation that the
predictions which their model makes are not in fact consistent with the
behaviour that the laws of physics determine. But neither of these
theories originate with people with such magical powers, so in the real
world we go out and test the predictions against observable data rather
than attempting to show how the interactions of individual molecules, or
quarks and leptons, lead to the same predictions as their models.

In case you were trying to suggest that the identity theorist has to
work out what behaviour they would expect without using the schematics
or the simulator, that is rubbish, they could use whatever was
available. It would also be plainly wrong to suggest that either the
identity theorist or functionalists, could understand some schematics
that the others couldn't.

I'm suggesting that neither has access to anything that will let them
work out the behaviour in the way that you are suggesting. Both work
with the predictions generated by their own theories. If they construct
simulations, they will be simulations built around their own respective
theories and not ones built around low-level physical laws.

You don't seem to be able to come to terms with neither the identity
theory of step 2, or functionalism in theory B, making any predictions
about physical behaviour other than the universality assumption, and
the predictions of behaviour given the universality assumption would
depend upon the known laws of physics at the time.

What you call the universality assumption is simply the claim that there
is ultimately a single, unified set of physical laws and that all
things, living or non-living, conscious or non-conscious are subject to
those laws -- at least that's what it is if it is an idea that
physicalists actually subscribe. You seem to be treating it instead as a
committment to reductionistic explanations. If that's what you're
arguing against, then you won't get any argument here, but since neither
functionalism nor mind-brain identity theory purport to being
reductionistic theories (in the methodological sense), it certainly
doesn't lead to your absurd claim that physicalist theories must always
makes the same predictions regarding the universe.

Look at the history of science -- it is replete with examples of
competing theories making different predictions, predictions which
ultimately allowed us to decide between them. Are you claiming that
because different predictions were made, that in all such cases at least
one of the theories involved was not physicalist in nature?


You say:
--------------
Functionalism makes certain predictions.

Identity theory makes certain predictions, though not necessarily the
same ones. In particular, the identity theory which your theorist
advocated denies the possibility that the robot could pass the turing
test, whereas a functionalist who asserts the robot is consicous
claims that it will.
--------------

You had already agreed to
----------------
Step 5:
If for a particular robot construction behaviour X would be for it to
pass the Turing Test, then those believing the suggestion put forward
in step 2 consider that reality would either be that would be like
something to be the robot, or that it wouldn't.
----------------

So you knew that they weren't denying a robot could pass the Turing
Test.

Since this is just a thought experiment we can consider that there is
a computer simulation that can accurately model the robot build
following whatever the known laws of physics are at the time. In other
words it models behaviour X. This can be 5000 years in the future, or
whatever, it doesn't matter. Are you claiming that either the
functionalists of theory B, or those believing in the suggestion put
forward in step 2 would not expect behaviour X?

.

Relevant Pages

  • Re: Significance (contd)
    ... wouldn't be expecting the subject matter to follow the *known* laws of ... predictions directly from the known laws of physics any more than they ... there which make different predictions. ...
    (talk.origins)
  • Re: Significance (contd)
    ... It isn't a philosophy of mind version of functionalism. ... predictions as to what those laws governing behaviour would be, ... laws of physics are at the time. ... Functionalism makes certain predictions. ...
    (talk.origins)
  • Re: Significance (contd)
    ... wouldn't be expecting the subject matter to follow the *known* laws of ... predictions directly from the known laws of physics any more than they ... there which make different predictions. ...
    (talk.origins)
  • Re: Significance (contd)
    ... predictions directly from the known laws of physics any more than they ... there which make different predictions. ... very frequently cannot predict how a given program will behave -- if we ...
    (talk.origins)
  • Re: Problem for physicalist evolutionists
    ... Functionalists talk in terms of functions rather than physics ... wouldn't be *intrinsically* incompatible with functionalism. ... from the known laws of physics. ... asserts that photons possess mass and one which does not. ...
    (talk.origins)
Loading