"Solving the Brain
Published: November 10, 2003
Zach Mainen, Assistant Professor, Cold Spring Harbor Laboratory"
"This vast ensemble of elements somehow works together in a community whose organizational principles make it all come together in the right way. It's our job to figure out how. This is where the deepest mysteries and challenges lie. Only by understanding this level can we start to forge the links between the molecular world of the `bottom-uppers' and the psychological world of the `top-downers'."
It's TD E/I-minimization, top to bottom, through and through.
"Spikes are like the 1's and 0's of the digital computer and we need to know how they should be read."
Sorry, this's not True.
The ionic conductances are infinitely-divisible. "Spikes" are 'just' thresholding energydynamics within the ionic conductances' infinite divisibility.
"In particular, if a group of neurons synchronizes their spikes they are much more likely to have their message heard-imagine people clapping in unison versus applauding randomly-so who fires in step with whom may be much more important than what any individual neuron is saying."
'synchronization' is an artifact of TD E/I-minimization.
"By recording from a few dozen neurons simultaneously in a rat performing an olfactory discrimination task, we hope to discern whether synchrony between spikes carries information about the identity of the stimulus that is relevant to the behavior."
Suggestion to the Experimenters:
The 'synchrony' observed in the locust derives in the fact that only a small collection of olfactory-detectable molecules have significance to the locust. So, of course, since these molecules are, themselves, energydynamic templates, any mechanism that detects them will respond stereotypically.
So, when working with mice, present odors that are non-stereotypical. You'll still see the 'synchrony' artifact of TD E/I-minimization, but since, in this mammal, olfaction is much more capable than is olfaction in the locust, you'll see that the 'synchrony' artifacts of TD E/I-minimization that can be induced are limitless - and that means that it's not 'synchronization', but an energydynamics convergence.
The former is necessarily-limited.
The latter is infinitely-adaptable.
So, get the longest list of smelly stuff that your endurance allows, and you'll distinguish between the two alternatives.
The limitless-convergence is limitless-convergence, not 'synchronization'. You'll still see the 'synchronization' artifact, but, after you satisfy yourselves that it is in fact an artifact, toss it in the Artifacts Bin.
The same-stuff applies to brain function in general, and it's easy to see why the limitless-convergence dynamics are more functional. It's commensurate with optimized adaptability, whereas the 'synchrony' dynamics would not be. If the 'synchrony' dynamics were, then they'd be the limitless-convergence dynamics, and not the 'synchrony' dynamics :-]
"It seems plausible that something like canonical circuits exist, given the regularity of anatomical structures observed, the finite number of genes which can control the development of circuits, and so on."
This problem has been Resolved in AoK, Ap6, all along. Genes don't determine the circuitry. Energydynamics driving the genes does, and as close as anything gets to being "cannonical", in terms of its microcircuitry, is that all of the circuitry of the ['normal'] nervous converges upon physical embodiments of TD E/I-minimization. For high-level discussion of the all-permeating energydynamics, see the discussions of cortical column and stripe formation in AoK, Ap6. There are no 'cannonical' microcircuits within the brain be-cause, as is discussed in AoK, Ap5, neural glia are intimately involved in neural energydynamics, and, since the neural glia are both conformation-variable and replacable, since they are involved in neural energydynamics, there can exist no 'cannonical' microcircuitry. [Again, this is commensurate with optimized adaptability. For a discussion of why such matters, see AoK, Ap1.]
"[...] the question of how the various brain regions coordinate their functions and activity"
"Are there basic rules about how all that spike traffic is routed and rerouted to where it is needed when an organism acts intentionally in its environment?"
"How are all those representations put together?"
"A single concerted action appropriate to the situation must be engaged and all others suppressed."
"All these problems point to the centrality of mechanisms for flexibly integrating information across many different areas of the brain, for putting together many kinds of information into a common framework."
Supersystem configuration, AoK, Ap5.
"With so many messages coming and going between the different parts of the brain, how is all the traffic safely routed to the right place at the right time?"
TD E/I-minimization convergence within supersystem configuration mechanisms.
"oscillations" and "rythms"
TD E/I-minimization convergence within supersystem configuration mechanisms. See, for instance, the discussion of hippocampal "ratchet-pawling" in AoK, Ap5. As I've discussed in long-former posts here in b.n, the 'oscillations' and 'rythms' occur as topologically-distributed ratchet-pawling over-shooting during TD E/I-minimization.
The 'oscillation' ia just an artifact of the infinitely-variable TD E/I-minimization energydynamics. What it embodies is the TD network's localized 'recognition' of TD E/I-increase, and the "pawling" [latching] -response latency of the network with respect to such.
Billions of little topologically-distributed "whoops-a-daisies", when TD E/I(up) occurs locally. See the discussion in AoK, Ap5, and the material in the refs cited in AoK, Ap5, for details.
"By restricting different kinds of neural activity to specific phases of this cycle, related kinds of information could be grouped together while preventing crosstalk with other types of information."
This's what happens, but the circuitry doesn't 'know' anything about the information with respect to which it is configuring itself. All it 'knows' is that TD E/I(down0 'is good' and that TD E/I(up) 'is bad', and it configures itself accordingly. Although, at this 'level', it's 'blind' to information-content, it's extraordinarily-functional be-cause, as is discussed in AoK, Ap3 through Ap8, if/when TD E/I-minimization occurs 'inappropriately' the host organism's experiential external environment imposes TD E/I(up) back upon the nervous system, and a new convergence upon TD E/I-minimization ensues.
"periodic clock cycle"
No such thing exists within nervous systems.
Yeah, there's stuff that has the appearance of constituting 'periodic clock cycle' stuff, but it's all just artifactual with respect to TD E/I-minimization.
If 'periodic clock cycle' stuff actually existed within nervous systems [instead of what's there being an artifact of TD E/I-minimization], the existence of the 'periodic clock cycle' stuff would 'dictate' the 'limits' of adaptability in a way that's analogous to the way computers' clocks [MGhz] dictate information-processing per unit 'time' - it's why a fast contemporaneous computer can be programmed more-flexibly [with greater adabtability] than can an 8088-CPU machine. Using 'least-common-denominator' instruction sets, one can put analogous programs into both machines, but one will only grow-old waiting for the slow machine to complete a program run. Now, imagine that your Life depends upon the adaptability-differential inherent. You'll want to use the fast machine.
Toward the same end, evolutionary dynamics transcended 'periodic clock cycle' stuff, period.
Brains do it, in a no-holds-barred, converge-upon-TD E/I-minimization-via-all-possible-means way that transcends 'periodic clock cycle' stuff, which is commensurate with the infinitely-varying problem-sets that confronts nervous systems, and simultaneously optimizes adaptability.
I'll gladly accept any [non-trivial] formally-published, replicable, experimental results as a Test of what I assert in my comments in this post, thereby demonstrating that it all stands Proven. [Of course, I'll have to trace things through the larger chain of experimental evidence. But I've already done that [decades ago]. These days, be-cause of the on-going efforts of my Colleagues in Neuroscience, it's just easier to do such.]
I have to stand for such a Test in-person. [Room and board will help 'cause I've still got to maintain my pile-of-sheet-metal.]
K. P. Collins
[P. S. It's 'hilarious'. My machine's been 'grabbed' all the while I've been preparing this post.]
Sorry, be-cause of a malicious attack, I've had to instruct my ISP to block all email :-[ So, if you want to contact me, please do it in the bionet.neuroscience newsgroup by posting a message.
-------------- next part --------------
An HTML attachment was scrubbed...