Computers and the Unity of Consciousness

I wish I had the slides that Searle used. One was a short list of characteristics of consciousness he used to delimit the subject matter. One characteristic was sequentiality or unity. By that I take it that we are aware of one thing at a time rather in the sense that a movie is sequence of frames, each pointing in some particular direction. There may be several things in the frame and our consciousness is not limited to visual inputs, nor even external inputs. The sequence of conscious events seems simply ordered. These include signals from our subconscious which does indeed proceed asynchronously as far as we know. All of this is compatible with the notion that the stream of consciousness is serial stream deposited in our memory which is a record of our ‘conscious processing’, and that our ‘awareness’ is via reference to this memory. Of course, this must be very short term memory in order to comport with our subjective experience. (more here) Computer architecture teaches us that there are only documentation reasons to draw a boundary between CPU and memory. (Is the cache part of the CPU or the memory?)

To me there is a perfectly adequate and naïve explanation for this: Consciousness is rather like a computer which, since von Neumann, finishes one thing before it starts the next. Further the reason for this is the same for the brain and for the computer—there is some fairly general hardware that can be applied to any of many subjects but only one subject at a time. (quibble) The 6600 was perhaps the first computer that could begin an addition before it had finished a preceding multiplication, at least when the sum did not depend on the product. Even there there was an ‘illusion’ of sequentiality. A further analogy with the computer may be useful: Consciousness relies on evoking similes and metaphors to formulate new ideas. I contend that the brain’s representation of such is, like computer memory, serial. In this regard Kanerva’s brain-memory model is correct, I think. While we have evolved pattern matching hardware that is specialized to each of our various senses, the hardware that finds a match between the mathematician’s group, and Lorentz transformations, does this by evoking metaphors which are stored in memory. Like the computer, memory and CPU are naturally serial and nature has exploited this. It would be fun to speculate on the adaptive advantage of the ability to find such general patterns 200,000 years ago.

Patten matching specialized to vision can proceed in parallel to other specialized processes, because there is hardware with its own associative memory devoted to it. We are unaware of that process until a significant pattern is recognized, rather like a computer’s IO device interrupting a program upon an asynchronous event.

All of this is so obvious to some people, I suspect, that they find it too trivial to speak of. To others, I suspect it is very alien, or even sacrilege.

Parallel to all of this is the high level guidance system that leads us to seek these patterns. It feels like it is for the fun of it. How did that evolve? What complex problems were solved 300,000 years ago that allowed our ancestors to survive? Some suggest problems of social organization. I don’t like that answer but I have no better.