From [Marc Abrams (2005.10.17.1536)]
In a message dated 10/17/2005 1:38:05 P.M. Eastern Daylight Time, r-robertson@NEIU.EDU writes:
Bill Powers wrote:
[From Bill Powers (2005.10.16.1851 MDT)]
I guess both acting and observing involve perception. Why didn’t I
think of that?
But then, shouldn’t the whole motor cortex light up when you observe
somebody doing something? Why just Broca’s area?
Category perceptions? What, exactly, was the goal-directed action in
the experiments?
Category perceptions. That’s interesting, I was looking for what
hierarchy level , and I couldn’t resolves upon one. That helps, I’ll
think about it. As for the experiment, as usual they don’t report what
a PCT approach would like to know. I’ll see if I can find out more about it.
Bill makes a good point about the cortex and herein lies one of the problems I have with the hierarchy as it’s currently postulated. Are neurons at different levels somehow differentiatable? That is, can we tell the difference between a ‘category’ neuron and a ‘relationship’ neuron? Are there any such things? If so, how can we go about testing that idea?
I have always been confused about whether the hierarchy represents a physical physiological structure (i.e. from spine, level one; to cortex, level 11) a functional one that represents how a perception is constructed, or something in between or different than those two.
Another problem I have with the PCT hierarchy is the notion and concept of dependency.
In a hierarchy each level is independent of any other levels and each ‘higher’ level; is fully dependent on its existence from the level below. Hence the neat concept of up=Why and down=how. Again, in a hierarchy we are not talking about mere influence as in a network. We are talking about full dependence. Which leads too…
What data do we have to support this notion of dependency? A theoretical model? Hardly, because if that were the case I could justify the existence of Alice and Wonderland.
A model is useful if it can support empirical data. To date we have not been able to get any data on the existence of any levels. Rick’s spreadsheet model simply states that a hierarchy is theoretically plausible. It does not show the existence of one. Huge difference
As much as I have tried to think about this and even drawn it out with diagrams, I cannot construct perceptions from this structure.
First of all, perceptions as I and most of the non PCT universe defines them, are interpreted understandings of sensed data. Simple sensations are not meaningful because in order to be ‘controlled’ they must be ‘compared’. If two things are compared there must be something that differentiates between the two and each one must have some kind of ‘value’. That means the ‘values’ must either come from the nature of the neurons themselves or the ‘structure’ the neurons take on when representing something. By structure I mean the cohesive signaling or oscillating waves neurons produce in ‘talking’ with one another.
My question; What justifies the need for ‘levels’. Think Occam’s razor here. What benefits would levels provide a communicating nervous system?
This is a question not a challenge. Anyone have any ideas?
Regards,
Marc