Kelly wrote:
> What is the advantage of assigning consciousness to computational
> processes (e.g. UDA), as opposed to just assigning it to the
> information that is produced by computational processes?
> For example, to take Maudlin's "Computation and Consciousness" paper,
> if you just say that the consciousness is found in the information
> represented by the arrangement of the empty or full water troughs,
> then that basically removes the problem he is pointing out.
> Similarly, associated consciousness only with information seems to
> resolve problems with random processes interfering with the causal
> structure of physically implemented computations which then, despite
> having the causal chain interrupted, would still seem to produce
> consciousness.  (more on the irrelevance of causality:
> Bruno Marchal has mentioned this in his movie graph argument, where a
> cosmic ray interrupts a logical operation in a transistor on a
> computer that is running a brain simulation, but due to good fortune
> the result of the operation is still correct despite the break in the
> causal chain that produced the answer.
> Conscious being associated with information would also seem to address
> the problems with Davidson's "swampman" scenario, and the related
> quantum swampman scenario (
> 2009/03/quantum-swampman.html).
> So, many different programs can produce the same information, using
> many different algorithms, optimizations, shortcuts, etc.  But if all
> of these programs all accurately simulate the same brain, then they
> should produce the same conscious experience, regardless of the
> various implementation details.
> The most obvious thing that all such programs would have in common is
> that they work with the same information...the state of the brain at
> each given time slice.  Even if this state is stored in different
> forms by each of the various programs, there must always be a mapping
> between those various storage formats, as well as a mapping back to
> the original brain whose activity is being simulated.
> Therefore, it seems better to me to say:  Consciousness is
> information, not the processes that produce the information.
> What are the drawbacks of this view when contrasted with
> computationalism?
The main difficulty I see is that it fails to explain the sequential 
aspect of consciousness.  If consciousness is identified with 
information then it is atemporal.  There are attempts to overcome this 
objection by assuming a discretized consciousness and identifying 
sequence with a partial ordering by similarity or content, but I find 
them unconvincing because when you chop consciousness into "moments" 
then the "moments" have very little content and it's not clear that it 
is enough to define a sequence.  It seems you have allow each "moment" 
to have small duration - and then you're back to process.  Or instead of 
expanding consciousness in the time direction, you could get enough 
information by expanding in the "orthogonal" direction - i.e. including 
unconscious things like information stored in memory but not being 
recalled (at the moment).  But then you've slipped physics in.


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