David Nyman wrote:
> 1Z wrote:
>
> > Whatever properties are picked out by a computation
> > will be relevant to it as a computation.
>
> Yes, of course. But how are these properties supposed to simultaneously
> produce a state of consciousness stably linked to the 'computation'
> when this self-same computation could have been instantiated in
> arbitrarily many physically distinct ways?
Why not? A ordinary numerical computation can be instantiated
in arbitrarily many physical ways, and still produce the same result.
> The computations would be
> equivalent, but you appear to be claiming that however they are
> implemented, arbitrarily many distinct physical properties somehow
> become equally 'relevant' to generating the same state of
> consciousness.
I still don't see why you think this is a problem. If different
physical states always produced different mental states, there
would be no mental commonalities between people,
since all brains are physcially different. Mutiple realisability is a
feature,
not a bug!
> > There is no requirement that
> > the same connscious state is implemented
> > by the same physical state, so the multiple
> > reliasability of computations is not a problem
>
> So you say, but just *what* physical properties are supposed to be
> relevant and *how* do they contrive always to manifest equivalently
> within totally different implementations of a computation?
How do they for a non-conscious computation?
That is no a mystery, it is computer engineering.
> Is this just
> supposed to be a mystery?
The mystery is which computations are conscious.
> My point is that under materialism,
> 'computation' is just a metaphor and what is directly relevant is the
> activity of the physical substrate in producing the results that we
> interpret in this way.
Under physicalism, *all* the activity is relevant.
Under computationalism, a subset is.
> What's critical to computational equivalence is
> not the internal states of the physical substrate, but the consistency
> of the externalised results thus produced.
That's a broad definition of equivalence.
Running the same algorithm -- rather than producing the same results --
is generally more relevant.
> But with consciousness, it's precisely the internal states that are
> relevant.
Yes.
> And here your reasoning appears to become circular - a
> particular set of physical properties can be construed as
> 'externalising' a particular set of computational results at a given
> point in time (fair enough) so, whatever these properties happen to be,
> they're must also be 'relevant' in generating a specific internal
> conscious state - and so must any arbitrary alternative set of
> properties that externalise the same computational results. Only
> because you say so, AFAICS.
It's a claim of computationalism. I am just explaining how
computationalism is compatible with physicalism. You
are complaining about circularity, not contradiction!
It's not an argument that computationalism is actually true,
nor meant to be.
> By making the rationale for supervention of
> consciousness on physical activity completely arbitrary in this way
It's not completely arbitrary. I don't subsribe to
the idea that every physical system implements every computation.
I don't even think computer-like systems are particularly common.
> (it
> just *somehow* tracks a 'computation' however instantiated), you've
> effectively abandoned it as a materialist explanation. Didn't
> Hofstadter use this sleight of intuition to conjure consciousness from
> anthills and books - or was he perhaps just joking?
Well, there *may* be too much multiple realisability. But remember
that I have a narrowish view of what is a computer. And remember
that consciousness is not held to be any old computation.
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to everything-list.domain.name.hidden
To unsubscribe from this group, send email to everything-list-unsubscribe.domain.name.hidden
For more options, visit this group at
http://groups.google.com/group/everything-list
-~----------~----~----~----~------~----~------~--~---
Received on Mon Oct 09 2006 - 21:52:18 PDT