Orchestrated Reduction, cont.
(reply to Paul Wilkens)

Date: Mon, 19 Feb 1996 08:23:38 -0800
From: Stuart Hameroff <srh@ccit.arizona.edu>
To: quantum-d list <quantum-d@teleport.com>
Subject: Orchestrated Reduction, cont. (reply to Paul Wilkens)

with reference to Paul Wilkens' comments at:
http://www.teleport.com/~rhett/quantum-d/posts/paul_2-16-96.html

Paul Wilkens: 
                 The idea of coherent quantum states leading to a real 
neural events is quite attractive. however, i don't quite understand 
the way this event is presented. when do neurons fire? when microtubules
attain coherence, all happily postulating possible states in infinite
universes, they must choose a particular state to collapse to: when this 
happens, does this mean that their neurons reenact, in ionic potential, 
the exact state that the microtubules envisioned? 

SH: 
             It's likely that neural activity most directly relevant to 
consciousness occurs in dendrites ("dendritic processing" - related to 
the "summing post synaptic potentials - PSP's" to which Paul refered"). 
This notion has been eloquently argued by John Eccles and Karl Pribram. 
The cytoskeleton in dendrites is distinctly different from axons. In 
dendrites microtubules are short, and of mixed polarity arranged in many 
local circuits; in axons they are uniformly polarized ("plus end" pointing 
toward synapses, "minus end" toward axon hillock). What seems logical is 
that dendritic microtubules, influenced by post synaptic potentials 
("PSPs"), process in quantum superposition mode and self-collapse to a 
set of states which is conveyed to the axon hillock and regulates axonal 
firing. (We would argue that all-or-none axonal firing, even considering 
frequency coding and contextual patterns, is a mere shadow of the 
information conveyed along the axonal cytoskeleton.) Cycles of dendritic 
processing are reflected in extra-cellular dendritic potentials - the 
electro-encephalogram (EEG).

Now lets look at our Orch OR model. As we've said, 10^3 (one thousand)
neurons would be required to sustain coherence for 500 msec to have a
conscious event. With some form of excitatory input (e.g. you are
threatened, or enchanted) in which quantum coherence emerges faster,
then, for example, 10^10 coherent tubulins would "Orch OR" after 50
msec, or 10^11 after 5 msec. Turning to see a bengal tiger in your face
might perhaps elicit 10^12 in 0.5 msec, or more tubulins, faster. A slow
emergence of coherence (your forgotten phone bill) may require longer
times. (A single electron in quantum superposition would require more
than the age of the universe.) Each Orch OR may "bind" various super-
positions and their content into an instantaneous event (a "conscious
now"). 

Lets say you're mildy excited, and 2 x 10^10 dendritic tubulins Orch OR
repeatedly at 25 msec intervals. This would correlate with an EEG signal of
coherent 40 Hz! 500 msec Orch ORs would correlate with 2 Hz. Generally,
there will be a mixture of frequencies.

Paul Wilkens: 
                How do we know if microtubules did anything at all, if
the end state is merely another electrical pattern, one that would have
occurred anyway by normal means? 

SH: 
       First, irrespective of consciousness, it's unlikely that neural 
membrane and synaptic activities could occur without cytoskeletal support. 
But if they could (or could function without cytoskeletal Orch OR) we 
would argue the result would be a zombie.

Stuart Hameroff
srh@ccit.arizona.edu