This thread has been extremely helpful to me in terms of getting to
the heart of this problem and the whole issue of supervenience - thank
you Joseph for your clarification of the meaning of the term and for
your succinct and clear summary of the MGA, and to David for the nice
clarification of the 'qua materia'/'qua computation' distinction. But
I have yet to see why the MGA proves that consciousness can't
supervene on abstract computation +  concrete implementation. I can
see that Joseph's refutation misses the mark because the issue is that
the replaying of a recording, whether on a screen or within the
original mechanism, performs no computations. But why cannot the
materialist/computationalist merely counter that Alice *is* a zombie
during the playback of the movie, because the required instantiation
of  a computation is absent? Sure, he is committed to consciousness of
the machine if the physical activity is identical, but in the playback
of the film, the activity is not identical, since the connections
between logic gates are broken and/or overridden by the *projected*
activity (be it 'lucky rays' or the film). Although the sequence of
firings in the network is the same, the causal connection between
firings is removed - indeed this is the point: no calculation is being
carried out. But a sequence of firings in a logic network is not the
entirety of that network's physical activity. Or rather, the physical
activity of the sequence is not sufficient to define its activity as a
computation. That requires the casual connection between firings to be

Imagine a domino computer. I can't remember where I heard this first
(maybe on this list somewhere), but we can imagine a network of spring-
loaded dominos that are set up to spring back upright after a certain
time. By setting up rows of such dominos in a clever fashion, we can
use it to perform calculations. Let's say we perform a calculation
with a boolean output - either a domino at the end falls or it
doesn't. If we set up such a domino computer and push the first
domino, we initiate a causal chain reaction that performs the
calculation we have programmed it for. Now imagine we disable the
causality by gluing the dominos upright. Now imagine we have a set of
instructions telling us to lower and raise dominos in such and such a
sequence. Our instructions happen to tells us to raise and lower them
in exactly the sequence they would have if they had simply been pushed
without the glue. This could be a random set of instructions that just
happens to be the same (as per luck rays), or a description
(recording) of a previous actual run of the computer (as per movie
graph). This is a restatement of the MGA scenario. In that case, the
casual interaction between dominos has been removed, but the sequence
of 'firings' in the network is retained.

Now the materialist-computationalist already believes in the odd
scenario of a consciousness instantiated by a computation in which the
steps of the computation are performed in different places in time and
space - eg one step in a calculation is performed in Sydney on one
machine in 2011 and the next is performed on another in Melbourne in
2012  (local examples rather than Brussels-Amsterdam!). It is still a
potentially conscious calculation if a causal connection between
computational steps is retained. Remove the causality from the
scenario and it becomes meaningless and absurd - otherwise
consciousnesses would arise between all kinds of  unrelated things. A
bit of half written code on my computer in Melbourne could be
completed by some half written code on your computer in Sydney, even
though the computers and the programmers never interacted. And of
course, everything physical is Turing emulable, so everything physical
performs (at least trivially) calculations. Consciousness would arise
between all the random motions of particles that could be regarded as
performing a calculation *if* they were causally connected. Madness.

So, given that causality is physical (even if such causality is highly
indirect), then comp-phys can argue that Alice is a zombie in the
projected film scenario because of the severance of causality between
the activity of logic nodes. The computer no longer instantiates a
physical computation and comp-phys requires both a computation and a
physical instantiation.

Personally, I think the scenario of a physically atomised computation
does comp-phys in anyway. The notion of physical activity seems
stretched beyond breaking point when we extend it to the sequence of
causes that connects the steps of such a computation. No further
reductio ad absurdum is required. But the problem with any reductio ad
absurdum is that different people find different things absurd, and
seeing as comp-phys accepts the possibility of a temporally and
spatially atomised, conscious computer, so it can use the same
principle to refute the MGA. Sure comp+phys forces us into absurdity,
but the absurdity has already been accepted, and the MGA adds nothing

You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To post to this group, send email to
To unsubscribe from this group, send email to
For more options, visit this group at

Reply via email to