Chats
From Lettvin
Contents |
Older
20090427
You have grasped the essence of what I wrote. Your changes are subtle, but appropriate. There are few concerns I have towards your changes.
One is the separation of sensing from actuation. It was my intention to prevent this separation. Reflex is entirely a closed though subtle and convoluted loop. The experts separate sensing from processing from actuation and this is a point of divergence I respectfully have with them.
There may be a better style with which to deliver the message. Your full-out criticism is not just welcome, it is needed. I know I write with greater density than most people and that my style is actually repulsive to some.
I would love to see how you would utterly rewrite it. Do you think there is room for total rewrite to achieve a more human feel?
If you know Cajal's excellent book "Advice for a young investigator" http://books.google.com/books?id=DxDFBLQei8MC&dq=Advice+to+the+young+investigator&printsec=frontcover&source=bn&hl=en#PPA3,M1 you will understand how well I think these ideas can be written.
How Cajal succeeds at powerful criticism while remaining a respectful and generous gentleman is a marvel for me that inspires my writing efforts. If I could only write like him, I could be happy.
20090430
"A neuropil (a volume in which the tree-like parts of neurons interdigitate) is one view of a layer. Another view is the volume in which the cell bodies of the neurons live. Signals are injected into a layer and undergo topological transform, then come out as a new set of modified signals. These layers can be stacked quite deep." What I need to know is YOUR view or the view that you are using. Another alternative is for you to say that both, either, or neither are correct or what you have in mind. The problem here is that you are trying to be as clear as you can but you are using language that, because of its density seems either ambiguous or inconclusive. (Please note the "seems".) Part of the problem with the chunks I've read so far is that you are using academic rules of rhetoric to attempt to explain, if you will forgive the pun, your clear vision of a mechanism. The difficulty is that you are far-sighted and I am near-sighted and between us we will have to develop the concept in the ground between our areas of perception and expertise. Don't make me get Sharon to smack you upside the head for some answers bro.
20090504
10:44 AM David: So shut up already and talk to me of the sensorium. I must admit that the definition that I've been most intrigued by seems slightly different than the one you have given me. Here's how I have written it up.
The sensorium is an ecology, which includes the apparatus of sensation, the environment they sense and the learned and innate systems for attention, interpretation and action. The sensorium gathers, represents, and react to changes of information.
The perceptual organs and processors are distributed across the body, in relation to the physical environment. Any single perceptual modality may include or overlap multiple sensory structures, as well as other modes of perception, and the sum of their relations and the ratio of mixture and importance comprise a sensorium. The perception, understanding, and reasoning of an organism is dependent on the particular experience of the world delivered by changing ratios of sense.
10:46 AM me: Perhaps I have abused the correct definition of sensorium. By it I meant "that which is to be sensed" excluding the mechanism of sensation.
10:47 AM Environment is both ambiguous in its own right and further clouded by modern usage.
10:48 AM "Sensible objects" is closer to the right meaning.
10:49 AM David: That's what I thought you meant, but I wanted to be sure. I have to admit that the idea of the closed loop of change - sense - action - change - sense - etc. is a delightfully Moebius way of thinking. The idea of the sensorium being the circuit rather than a component.
10:50 AM ... and I don't want that loop to seem sequential
10:51 AM but more as a kind of Taoist sensing is change is action is change is sensing
me: Here is a conundrum that will help you grasp the subtlety of this. A patient had her head fixed in a frame and her eye muscles paralyzed with a drug. Her vision faded to a uniform gray sense. When objects were moved in front of her, most of the scene was restored. When she was told to look to one side, she saw the world as shifting to the other side.
10:53 AM You are correct about the loop. Remember that the job of the brain is to eliminate all input that can be predicted. This cannot be done without a complete loop back out into the world. "Holding something steady" is a case in point.
10:54 AM David: I understand, but as I said ... there is a delightful prayer wheel concept in the idea that one is constantly creating one's own sensorium merely by existing.
10:55 AM me: To predict the world, one must embody the world; at least in the sense that any sensible thing that could be of consequence must be modeled and sent to the senses as a counter-signal to cancel the signal as it comes in.
10:57 AM I get the pleasure of relating to ancient philosophies and extracting value therefrom. However, I would caution that the analogy is fraught with dangers from at least two fronts. First, hammering the ideas into a shape that fits the philosophy may damage their properties. Second, important audience members can be lost by appearing religious.
10:58 AM I leave it to Herta von Dechend and Giorgio DeSantillana to make myth fit science.
11:01 AM I will coin a new phrase: "Cancellation is knowledge."
11:03 AM David: I'm not trying to bring ancient philosophies in just enjoying the fact that my image of this loop is similar to the extrapolation of the Taoist expression that 'one cannot step into the same river twice', to 'one cannot step into the same river once', since the very act of stepping into it changes it. So something sensed is thereby changed.
11:04 AM Don't worry, none of this maundering will appear in the edits ... it's just a tangent that exerted some diversive force
11:05 AM me: Let me content you that prediction is trivial in a nervous system by describing how easy it is to locate a moving boat in a lake.
11:09 AM Let us say that you live on a slow river and have three neighbors with boats, a cigarette boat (big and fast), a Chris Craft pleasure boat, and a fishing skiff. From just the waves that come in you want to know who went by, and in which direction. You also want to know where that boat is as soon as the waves get to you.
11:10 AM The river is deep, but has a wide shallow bank.
11:11 AM You carve channels in this shallow bank, and waves that enter these channels move more slowly because of the shallowness.
11:16 AM If you carve one channel at right angles to the river, a wave will propagate down the channel over a certain amount of time. A channel carved at a different angle to the same spot will cause the wave to take longer to propagate. Several channels at different angles leading to the same spot will reliably ADD the waves that arrive at the channel openings in a particular time sequence. This addition will cause the water to spout for a wave from a boat making a wake with a particular angle to the shore and traveling at a particular velocity.
No other wave will do it.
11:17 AM So identification of the boat is not difficult, since each boat has a characteristic wave.
Just carve different channel sets to identify each neighbor's boat and look for which one spouts.
11:19 AM Now, predicting where the boat is is easy. There are two approaches. First one could carve two channel sets for each neighbor, one for left-going and the other for right-going. Second is to actually look along the original wake.
11:24 AM A wake is merely a set of historical data that has high reliability of being along a straight line. The straight line intersects with the boat at the current time. So looking down the wake is sufficient to find the boat. The nervous system has the capacity to detect wakes of moving features and identify the sensor that is about to change at the time it is to change. The nervous system can then present a counter-signal to prevent a signal to be generated (cancelling the incoming signal). Should no signal be seen, the model was correct, and the knowledge in the model is correct. Should a negative or positive signal be seen, the model was incorrect, and this signal propagates deeper into the nervous system for further attempts at modelling.
That is all for the moment.
I gotta go ASCII 0x10 0x10. BRB
11:25 AM David: More than enough ... let me ponder it a while.
6 minutes
11:31 AM me: Before you go off, do you see how prediction is "causal"? It is not mysterious or guesswork or heuristics. It is purely mechanical. All you need is to set up the right channels, allow waves to enter the channels, to be able to read out identity and predict position.
11:32 AM The beauty of the nervous system is that it has simple rules for making, growing, shrinking, and destroying channels to achieve appropriate identification and prediction.
11:35 AM David: Yep, i got that.
11:36 AM me: Wow! I didn't think it would be that easy to communicate. I guess we're not in Alabama.
20090518
12:36 PM me: Fraternal sibling; are you ready for a new paradigm for shared documentation development?
11 minutes
12:48 PM me: Is you is or is you ain't my chat bro?
12:49 PM David: Sorry kemosabe, I was busy with some administrative stuff that I could not interrupt
me: Ah so. Need more time?
David: No. All set. But I may have to get a coffee refill in the near future.
12:50 PM me: Do you mediawiki?
12:51 PM David: Not yet ... but I have a feeling that I may learn soon.
12:52 PM me: I happen to LOVE mediawiki beyond all other wiki engines. It lets me do math formulae and has a richness of markup that others seem to lack.
David: Where do I get the core to get started or get in?
12:53 PM me: I will give you the wiki we will share and URLs to documentation that will get you started with the markup. Gimme a few minutes.
12:54 PM David: OK getting coffee brb
15 minutes
1:09 PM David: I'm back.
1:10 PM me: Before I give you the link to our wiki I want to apologize for the extremely poor configuration. The text and math are usable, but the "skins" have failed for some reason, and I have been unable to get any help to repair it.
1:11 PM http://www.mediawiki.org/wiki/Help:Formatting
http://www.mediawiki.org/wiki/Markup_spec
1:12 PM http://en.wikipedia.org/wiki/Wikipedia:Cheatsheet
1:17 PM David: Ok gottem ... give me a few minutes to look them over and get my bearings.
me: The major contribution I have made to it is in the Analogies page under my name.
1:18 PM Mediawiki keeps a running history of changes, so we don't have to worry about spotting revisions.
10 minutes
1:28 PM me: I will continue to seek help for finishing the configuration and making it look right.
David: What is the problem you're having with the skins?
1:29 PM me: Compare the following two pages.
http://lettvin.info/wiki/index.php/Main_Page
http://lettvin.net/wiki/index.php/Main_Page
1:30 PM You will see the difference that plagues me.
David: Yes.
1:31 PM me: I have spent altogether too much time trying to fix it.
1:32 PM David: Have I sufficient access to tweak?
1:33 PM me: Do you have experience enough with linux and mediawiki installation to have confidence doing so?
I will give you the access if you do.
1:34 PM David: Linux perhaps. Don't bother giving me the access for the moment.
me: It is probably a VERY simple thing like "soft-link this directory to that one".
5 minutes
1:40 PM me: I FIXED IT!
OMG!
1:41 PM David: Yes you did ... now all you need is a logo.
1:42 PM me: I got plenty. Gimme a moment.
27 minutes
2:09 PM me: Logo in place. Maybe a bit obscure, but it has lots of meaning.
18 minutes
2:28 PM me: What do you think of the wiki so far?
2:30 PM David: Slightly different from what I've been used to but pretty simple. I just dropped one of the blocks of text we were working on in Google docs into a test page in my section.
2:32 PM me: I have taken a cursory look, and approve of this mode of collaboration.
2:34 PM One approach I am considering is for you to edit my pages directly, marking sections you have rewritten in italics or color and referring it to the rewritten material in your section.
I avoid using color and italics in general.
David: It's good for me, especially if you feel more secure with it. I will move the texts from Gdocs over and delete them.
2:35 PM me: I feel VERY secure with it. I have been documenting math for half a decade using mediawiki. I love it.
2:36 PM Just so you know, renaming pages is not a strength in mediawiki. You can start a new page and transfer the material, but renaming is just not an option (as far as I know).
2:37 PM David: Yeah I figured that out <g>
14 minutes
2:51 PM David: Instances of our work in Gdocs has been deleted and purged
have
25 minutes
3:16 PM me: It looked like you were about to ask a question.
I see the word "have".
3:18 PM David: edit of the text immediately above it should have been have not has. Editorial pickiness I s'pose.
3:21 PM me: Oh. I get it. Contextual ambiguity clouded my judgement.
You were correcting yourself.
4:38 PM David: When you talk of layers, are you using some construct of your own, or the layers of the LGN or the areas of the visual cortex?
4:42 PM or should I (as I suspect) be looking at Jeff Hawkins work on the Memory Prediction Framework?
17 minutes
4:59 PM me: Context please. Which use of layers are you discussing?
5:01 PM In the "Analogies" context this is literally atomic layers 1 Angstrom deep each.
5:02 PM My usual other use of "layers" is the 5 layers of the retina.
5:03 PM Each layer is a microns deep, and is usually a synonym for neuropil.
David: The context is that which you spoke of as having to do with neuropils. /forgive my ignorance as I get up to speed.
me: There are 5 layers of cell bodies in the retina.
5:04 PM There are four layers of interdigitation between those 5 cell body layers. Most neuroscientists would consider the retina as a single neuropil. I do not.
5:06 PM I must leave for the evening. I await what you are currently typing.
David: So when you talk of "The brain senses signal activity occurring in the sensorium, the environment that generates differentials in the intervening medium that an animal's sensors can detect, then over one or more layers, projects these changes into a discrete spatial topology within a volume." Are you talking about the retina or some layering in the cortex?
5:07 PM I CAN wait til tomorrow for an answer if you need to go
5:08 PM me: The retina senses, transduces, and modifies the signals considerably. The first layers of the colliculus and LGN begin to organize spatially. Deeper layers reconstruct 3D sensorium objects.
5:09 PM The topologies are responsible for the spatial organization and reconstruction.
TTFN
David: Sayonara
20090521
11:09 PM me: I added some more to the wiki under BBK
23 minutes
11:32 PM David: I'll check it out.
11:33 PM me: Hey! You're on!
David: Yep
11:36 PM me: I'm adding to it, so expect hiccups.
11:38 PM David: That's okay. I'm busy drawing diagrams and flowcharts to try to get my mind around things. And I will have an odd question for you sometime as soon as I make sure that it's not infantile.
11:39 PM I gots pages and pages of pretty pitchers and diagrams.
me: Infantile is often better than informed.
11:41 PM David: Well ... I just want to check some sources other than my memory of something that Papa said a long time ago. If I remember it rightly there'll probably be some face to face hand-waving and pencil sharpening needed
11:42 PM BTW do you use Firefox?
me: Try me anyway. Delays scare me.
Yes.
11:43 PM David: Do you have Greasemonkey installed by any chance?
me: Not sure. How would I test it?
12 minutes
11:56 PM me: I will have to restart firefox to use it. It sounds good.
David: Okay ... so here's the question.
5 minutes
12:02 AM David: You have a brief mention of primal sketches, and it occurred to me, in relation to that, that many years ago Papa mentioned that the reason we are able to detect edges is because of the jitter of the eye. In other words, the nearly constant tiny jerks from side to side. So I had this odd thought about this jittering for perceiving an edge and its potential relationship to the information that you've been giving me. It would seem to me that this jitter would force changes down the neural pipe since certain receptors would always be sensing a change.
me: yes and no.
12:03 AM You are in a dark cave for five days. Someone fires a single strobe flash lasting 1 millionth of a second. You see the scene in clear detail.
David: It then occurred to me that perhaps this jitter was part of the perceptual mechanism that let us concentrate or focus on an unmoving, unchanging point for long periods of time as in looking throug a microscope or examining a minicule area.
12:04 AM me: The critical part of the data is "this is different from here to there" or "this is different from before to now". Either way works.
David: that should have read minuscule
12:05 AM me: Now let me shutdown firefox and restart. BRB
12:06 AM Gee. The restart made me lose the last several interchanges.
12:08 AM David: Try going to the Google talk icon in the systray.
12:09 AM me: I have no idea what you meant.
David: Lower right ... icon looks like an envelope in a cartoon balloon
me: Where?
12:11 AM David: In the System Tray in the lower right corner of your screen ... or are you not using Windows and putting me on.
me: I am using linux.
12:12 AM I often go weeks without using windows
12:15 AM I see that greasmonkey/wikied uses screen real-estate to give me menus of things to use. Both nice and not. Every time I move my mouse I lose my thought. Every time I type a thing, I keep my thought.
12:16 AM I will try it out. It seems to give some advantages and I can simply go ahead and use my knowledge of the markup to ignore the menus.
12:17 AM David: A note about Google talk. You do know that these conversations are stored in the chat section of your gmail interface and you can print them out. That's what I do ... for as you know every one of your pearls of wisdom is precious to me.
12:18 AM It is also useful for blackmail.
me: If you have archiving turned on, you are correct. If not, then not. The transition to greasemonkey invalidated a small section of our chat.
12:19 AM David: I don't know why that would have an effect.
12:20 AM me: Perhaps the shutdown during a software installation caused a purge of unsaved data.
12:21 AM David: That should only have happened if yo ... you were using Firefox for reading the chat weren't you.
me: Indeed, the originals of those chats are gone from the original chat logs.
12:22 AM BTW I have developed a new philosophy of life that I call "Way of the Rod". This is not a joke.
David: Well since you're pretty much the only person I chat with, I archive and print.
12:23 AM Oh go on spoil the child.
me: A rod is a retinal photoreceptor, oh joking one.
Are you interested?
David: Yes my bro and you are a conehead.
12:24 AM Of course I'm interested
me: I vaguely know the reference, but I never saw any such movies or show episodes.
Oh. Okay. I get it.
I will now proceed without humor.
12:25 AM A rod has two portions called inner and outer segments.
12:26 AM The outer segment contains the pigment used to capture photons.
Once a photon is captured, the pigment breaks and is no longer a pigment.
12:27 AM In a very short time in bright light, all the pigment would be used up.
David: Visual purple right?
12:28 AM me: (visual purple is but one of many pigments, but for rods I think you are right)
So there is a layer of cells beyond the cones in a layer called the pigment epithelium.
12:29 AM This layer detects the pigment poor condition of a rod.
It then bites off the end of the rod in a process called phagocytosis.
This shortens the rod.
12:30 AM The shortening of the rod is detected by the inner segment.
The inner segment produces more pigment that it stuffs into the end of the rod away from the pigment epithelium.
12:31 AM The inner segment is nearer to the photon source, so the densest pigment is always found away from the pigment epithelium.
In total darkness, the rod grows to a maximum length and then stops.
12:32 AM In extremely bright light, phagocytosis is fast.
The rod shortens in bright light as the inner segment attempts to keep up with phagocytosis.
This means that there is less pigment available at any given time in bright light.
12:33 AM And the maximum pigment is achieved in total darkness.
And so, now, here is the "way of the rod".
12:34 AM As those around you nibble away at what you value, slow them down a bit if you want, but pay more attention to making more value away from where they are nibbling.
12:37 AM David: I hope you realize how hard it was for me to avoid asking if phagocytosis was something like an ongoing bris.
me: http://webvision.med.utah.edu/movies/phago4.mov
You ARE a joker. I am too, but this is actually a very important philosophy for me.
12:39 AM David: I actually knew something about that. I researched it since I have hypersensitivity to bright light. It's not something that I can call pain but the instinctive avoidance reaction is the same as you would see for a pain reaction
me: Let me put this another way. To joke in the face of a serious effort is something I always expect from our father.
Thank you for turning serious.
David: ... and I understand and am not disparaging the philosophy.
12:40 AM me: Are there existing philosophies that profit from this kind of thinking explicitly?
12:41 AM A corollary is that the nibblers are always eating the dead or useless stuff.
12:42 AM David: Yes, but usually not with scientific metaphor. Discussions of phagocytosis is not the stuff of everyday life for most people.
12:44 AM me: Independent of scientific metaphor, is there an explicit well-founded general philosophy of "ignore the nibblers, nay even feed them your leftovers, as we go out and get more than they take"?
12:45 AM Is this the Al Capone philosophy?
Or Dubya?
Or politicians in general?
12:46 AM Or is there a more populist-friendly version?
David: Well you could make a case for the pig latin Illegitimati non te carborundum.
12:47 AM me: That doesn't fit. It means "prevent the nibblers altogether". I thought of that.
David: But I'm gettin to the end of my consciousness and I can't think of others off the top of my head. How about I ponder it overnight.
me: I am suggesting a decided effort to feed the nibblers, but on less than you take for yourself.
12:48 AM Sleep well. Curl into the arms of Morpheus and purr.
David: Not to go all medieval on your ass, but how about render unto Caesar etc.
12:49 AM me: That is the nibblers' point of view. Not the nibblee.
12:50 AM David: 'til the morrow bro.
me: Al Capone lost some battles, but won more. Difference is, he had hard feelings over his losses.
TTFN
20090523
11:21 PM me: bromine
11:23 PM David: C'est moi.
11:24 PM me: Have you been likened to an antacid before?
11:27 PM David: NNNnnnnnoooooo...... not that I know of
11:28 PM me: Bromine... Bromide... You see? You are not the only one to emit obscure belly laugh material.
11:31 PM David: Oh dearI guess I should be glad that you didn't extend it to Bostonian aristocracy Indian caste or males of a certain bovine strain
11:32 PM me: I concede. Blammo! Bravo! You got me.
11:33 PM David: It's the ADD ... I have little to do with it. It just uses me as it will.
11:35 PM me: And for my past thin skin I apologize. I think you might understand that my equivalent condition was used as a weakness exploitable by the ever vigilant wormtongue.
11:36 PM So. I may not understand a single word you say, but I'll help you drink your wine.
11:38 PM David: You could put the wine in a jar annelid on it.
11:39 PM me: BTW I'm thinking of starting a new company and calling it Aneuran.
David: Or if it's tequila ... annelid in it.
11:41 PM ... Aneuran? In honor of the founder of the British National Health Service?
11:42 PM me: Lost me on that one.
Anuran is a frog.
11:44 PM David: Aneuran Bevan was an important British politician
11:45 PM me: I was thinking of the subtitle "Frog's brain" or some such.
David: ... but I see that I've misspelled his name
11:47 PM For years I've used a dummy business name. Rana Pipiens Software, with its president ande CEO Harold Batrachian
It should be Aneurin Bevan.
11:48 PM He was Welsh.
me: So you like it maybe?
11:49 PM David: I think you have to be nerveless to appreciate being aneuron.
11:50 PM me: Getting serious, I believe single-celled animals have nervous systems.
11:51 PM David: You mentioned that before. I don't doubt that they react to stimuli, but I wonder how complex the "system" could be?
11:52 PM me: Extremely. Do you want examples?
David: http://en.wikipedia.org/wiki/Aneurin_Bevan
Fire when ready Gridley.
11:53 PM me: I may not be able to provide on-line images to back me up, but I have Cajal's excellent two volume "Histology of the Vertebrates".
11:54 PM Vertebrates are eukaryotes, as are plants, jellyfish, worms, and pretty much all life as we know it (excluding prokaryots bacteria and archaea).
11:55 PM Eukaryotes are distinguished by having long chain proteins inside their cell membranes.
11:56 PM These long chains may be of several species. Some are made of actin. Some are made of Neurofibril. There are other polymer chains comprising many monomers chained together, usually unbranching).
11:57 PM These polymers often, as one chain, cross from one membrane protein to another membrane protein.
11:58 PM You might know, already, that membrane proteins are the cellular gateways for compounds to be admitted into and expelled from the cell.
11:59 PM These membrane proteins are responsible for CHOOSING whether to admit/expel a thing or not.
12:00 AM The polymer chains cross through the cytoskeleton, passing near each other and, should the nucleus CHOOSE to, near the nucleus as well.
12:01 AM Neurofibril is my particular favorite, so I will focus on it for now.
12:02 AM Neurofibril is capable of supporting a propagating bolus of ions that travels along its length in one direction.
12:03 AM David: Ah ... like that swirl of energy down the nerve.
me: This propagating bolus traverses the cell from one membrane protein to another at the other end, unless it is intercepted.
Yes.
12:04 AM The bolus, as it travels down the polymer, may encounter another polymer close enough that the energetics of jumping from one polymer to another does not favor either strand, giving a 50% chance of the bolus leaving its original polymer and propagating, instead, down the other polymer.
12:05 AM The nucleus moves around within the cell, much like a spider in its web.
As it moves, it may CHOOSE to move two strands closer together or further apart.
12:06 AM By changing this distance, the nucleus can cause signals to be rerouted from one place to another.
David: You use the word choose ... I'm not sure what you mean by that.
12:07 AM me: The routing can have convergences and divergences that can effectively operate in the manner in which I believe the nervous system does.
There are proteins on the surface of the nucleus.
12:08 AM These proteins are moved about on the surface of the nucleus in a manner not unlike spiders moving their legs to walk about their webs.
12:09 AM The proteins can grab polymer, pull the polymer from here to there, glue the polymer in place, or unglue it, and then release the polymer, to walk to some other site.
The operation is very skilled appearing.
Does that help you understand what I mean by choose?
12:10 AM David: That's an extraordinary amount of active processing ... I wasn't aware of the complexity of it.
12:11 AM me: I have books on symbiosis that talk about nucleus as an invader. Also mitochondria.
There are now known viruses that look like nuclei without cells.
David: It's interesting to use the word choose ... which I relate to consciousness, but since I can't think of an alternative ...
12:12 AM me: I believe it to be conscious.
David: I figured that. I just wasn't able to correlate it.
12:13 AM me: I have no way of drawing a line between this and that such that I can say that across this fine line on this side is conscious and on that side it is not.
David: I understand. I'm not arguing against it, just trying to put it into a familiar context.
12:14 AM me: You can teach unicellular critters to tolerate and swim into bad tasting water to get to sugar, when untrained critters will not.
12:15 AM David: I know that ... I just hadn't extrapolated as far as you.
12:16 AM me: This "behavior" is not possible without an information processing system.
A "learning" mechanism is required.
12:17 AM There is no option but to assume that the mechanisms available are doing the work.
David: You're making my brain hurt so good.
12:18 AM me: It is also not possible to make an information processing system that operates on the surface of the cell (membrane) that would process information in the required way.
I do not deny that some of the processing occurs there.
12:20 AM The membrane proteins are interlinked and sometimes are long and loop through the membrance many times in many shapes. But the operations I expect to see require divergence/convergence operations in 3D organizations.
12:22 AM So, as I said before, unicellular critters have nervous systems.
12:23 AM David: If you'll forgive an obscure reference, I reminded of the poem :Big fleas have little fleas on their backs to bite 'em. Little fleas have lesser fleas and so ad infinitum.
12:24 AM I wonder how deep the incidence of this type of system goes?
me: No ad infinitum, but yes.
12:25 AM No deeper than DNA.
12:30 AM The DNA strands coil and uncoil, hiding and exposing lengths. Proteins are constructed and fragmented as the engine for doing so is instructed by the DNA. This coiling and uncoiling is done within the context of proximity to the nuclear membrane. The spider walk is achieved by DNA instruction based on signals arriving at the nucleus from the cell membrane. Signal can take the form of ion bolus (fast) or resource for the engine to use (energy and raw materials) or outgoing signals and excreta.
12:31 AM Think of how a hagfish can tie itself in a knot while sliming you such that there is no hanging on. The DNA does what it wishes, and everything else follows.
The nucleus is a mighty and naughty pirate that shares its piracy with the mitochondria.
12:32 AM Poor cell is left gasping and struggling with the oars as the beat is given by the pirates.
Lost in antiquity is the nature of the eukaryotic cell before the pirates came.
12:33 AM David: This is a lot to process even for an advanced colony such as me.
me: Is my reasoning ringing true and clear?
12:35 AM David: It seems to be. I'm not looking to verify or argue. This is new territory for me.
12:36 AM Can I assume that this is general (outside my little pocket of the universe) knowledge?
12:37 AM me: Do you see why I am not equal to writing this into a cogent and clear account? I shall be pilloried by science when my views are exposed. I hope to make the Brain Building Kit so successful for illustrating the principles that my statement that "these were the best analogies I could draw from the primary references" can be taken as reasonable.
12:38 AM No you cannot.
I have had to reason these out a-priori as necessary to account for that which the available research fails to expose.
12:39 AM David: Ahhh. Give me a moment ... and let me put the kettle on BRB
me: I proceed from mathematical necessity.
OK
10 minutes
12:50 AM David: Back. I had to brew myself some soporific herbal tea ... been having trouble sleeping.
me: Sorry to hear it.
12:51 AM I take it you are somewhat surprised at the material I've been presenting.
12:53 AM David: So tell me ... this unicellular nervous system ... is there any way of experimentally confirming this activity? I understand the reason that there must be SOMETHING going on (why didn't I think of it before) but is there a way to prove that the mechanism is as you imagine?
12:55 AM me: You sound like our father. I am insulated from such concerns to a great extent because a properly run experiment takes years to mount and I have 1. no time to do it, 2. no skill to do it, 3. no resources to do it. and much much more in the way.
12:56 AM I take the stand that no-one else has succeeded in making a model that imitates what is observed. I will succeed in making a model that imitates what is observed. Whether this is how nature works is up to some experimentalist to prove.
12:59 AM David: I wasn't nay-saying, merely trying to, once again, find context for the ideas. They are new to me certainly but they seem viable so I was just trying to find out if there was any source or support material or any derivative or contrary views.
1:01 AM BTW ... back when we started this conversation Deni told me to send you and Sharon her love.
1:03 AM me: I have pieced this together from a variety of sources. Some are decades old Russian monographs, others by folks like Lynn Margulies, others by the recent work on water around Gerald Pollack, some around the century old drawings of Cajal, century-and-a-half old work by duBois-Reymond, and much much more.
I will try to remember to tell the sleeping Sharon of Deni's love.
1:04 AM There is an amazing book on Symbiosis that seems to not have the popularity I think it deserves.
1:06 AM David: Sherlock Holmes once said "...when you have eliminated the impossible, whatever remains, however improbable, must be the truth."
1:07 AM me: That is one of my guiding principles.
David: I'm not going out on a limb here, but so far as I know you are the only person to have voiced a theory about unicellular intelligence.
me: It has its origins in the stentor story.
1:08 AM I have not encountered such a theory before either.
1:13 AM David: I took a short jaunt in the googlesphere and find that there's not much at all other than the weaselly "seem to react". But I did find this http://machineslikeus.com/news/primitive-single-celled-microbe-super-communicator
1:17 AM me: Wow! I came up with this about 4 years ago, and here is a paper from 1 year ago sufficiently similar to give me comfort. I LOVE when other people think the same things as me.
1:21 AM David: Check this out from a book in 2006. It's not much but ... http://books.google.com/books?id=IH9N4SKWTokC&pg=PA6&lpg=PA6&dq=single-celled+animal+intelligence&source=bl&ots=Teym6hiQ_o&sig=u3LaP6ys_yOBdGf5fqwKaex9JmY&hl=en&ei=q9YYSq29C-nAtweKkOnbDA&sa=X&oi=book_result&ct=result&resnum=6#PPA6,M1
1:23 AM me: What they call "associative memory" is one of the principal properties of Project/Collect 3D manifolds with which I work.
1:24 AM David: I just started skimming this but it also looks useful http://www.astrobio.net/news/article2111.html
1:26 AM me: Yes! I had a conversation with her about 5 years ago where I said these things to her, unbeknownst that she already was thinking along similar lines.
It came together 4 years ago for me when I started coming up with mechanism.
1:27 AM David: ... and this one relates to the mention of slimemold intelligence above. http://www.epsrcham.org.uk/Papers_files/p246.pdf
1:28 AM me: Thank you. I guess I am in better company than I realized.
1:29 AM But these are more observational than modeling.
As usual, I want to make something that imitates.
1:31 AM David: The stuff is scattered about. Let me see if I can bring my unwieldy brain to find more fuel for your fires. I understand that these may not be directly usable for the concept you're using, but you never know when a stray observation or anomaly can provide insight. I'll look around some more tomorrow. I should probably start thinking about sleep.
1:32 AM me: Okay. I have so much stuff in my head, I would like to focus more on getting it out than in justifying it quite yet. I fear my age is slowing my brain and I want this stuff to come out before I lose it.
1:34 AM David: You keep going. I'll rejoin the wakeful world tomorrow and review what you said above and see where it takes me.
me: okay
David: 'Night bro
me: gnyt
20090524
me: Do you have time for a lengthy trip down the axon?
Sent at 3:14 PM on Sunday
David: Sure.
Sent at 3:17 PM on Sunday
me: Neurofibrils are deeply buried in the actin cytoskeleton that extends through the length of the neuron from end-to-end (said that before).
The Neurofibril chain has charges placed at intervals.
Sent at 3:19 PM on Sunday
me: Negative charges that is. Between these negative charges are paired charge atoms organized into the molecular structure of the Neurofibril monomers. As with actin, water tends to "stack" over these paired charges (polar association) and form a gel.
Sent at 3:21 PM on Sunday
me: The unpaired negative charge forces local water to stand up like iron filings on a magnet, forming a significantly harder mesh than the organized water.
Sent at 3:22 PM on Sunday
me: However, nature does not like unpaired charges. Usually some ion, most likely K+ will associate with the unpaired Neurofibril charge, cancelling the local electric field (somewhat) leading to an apparently fully polar surface to the Neurofibril, whereupon all water stacks until these stacks compete with stacks from other polar surfaces. All this I covered before (I think).
Now come the dynamics.
Sent at 3:25 PM on Sunday
me: When the actin gel is liquified by the loss of Ca++ or the introduction of heat or mechanical force, Na+ rushes down the gel to where the Neurofibril strands are coursing by on their way through the cell.
When the Na+ reaches the Neurofibril, it breaks even the radially hardened water around the charge sites.
Sent at 3:27 PM on Sunday
me: The Na+ cancels the local charge and water begins to stack. The gel reforms by stacking water over the canceled Neurofibril charge. The Na+ is trapped by its shell of hydration under this stacked water which is highly regular coursing along the Neurofibril.
A K+ ion is not prevented from crossing through this gel because it will exchange water from its shell of hydration with organized water with very little energy gap.
A K+ ion associates much more closely with the Neurofibril than Na+ and the Na+ is kicked off the negative site, but has nowhere to go.
Sent at 3:31 PM on Sunday
me: If there is a neighboring free site along the Neurofibril, the Na+ can be pushed along under the organized water to a new Neurofibril site.
Sent at 3:33 PM on Sunday
me: When the Na+ rushes into the actin gel, it cancels the electric field of the actin polymer and the K+ that was in high supply diffuses out, leaving K+ in low supply.
Sent at 3:34 PM on Sunday
me: So, when the Na+ rushes in, the Neurofibril will be subjected to a change in proportion of Na+ to K+. This causes it to lose charge cancelling K+ as the Na+ increases. So the unstable Na+ Neurofibril association is quasi-stable until the actin begins to gel again, kicking out the Na+ and K+ begins to form its unnaturally high concentration again.
This unnaturally high K+ is what kicks the Na+ off the Neurofibril site, but with nowhere to go, it must move longitudinally.
It is prevented from moving backwards because there is always slightly less Na+ Neurofibril charge canceling in front of it than there is K+ Neurofibril charge canceling in back.
Sent at 3:37 PM on Sunday
me: Thus a bolus of Na+ slips like a sleeve down the Neurofibril polymer chain in whatever direction it is predisposed to go due to the paired positive/negative excitation I described in an earlier chat/email.
Sent at 3:38 PM on Sunday
me: The dendrite is given a paired opposing excitation, a bolus of Na+ is formed, it is pulled down the Neurofibril by electric field due to disparate association forces, it propagates through the cell body (not all neurons are like that, i.e. spinal cord has cell bodies off to one side), hits the denser gel of the axon hillock and is stopped dead.
However, this stopping transfers energy to the gel which softens.
Sent at 3:41 PM on Sunday
me: If multiple Neurofibrils deliver a bolus of Na+ to the hillock, the softening reaches a critical point and subsequent boli are no longer stopped. For a brief period of time, Na+ passes through the hillock and proceeds down the axon.
After they have passed the hillock reforms its overdense gel and stops subsequent boli.
I do not believe in "pulse generation" by the hillock. I believe in passive yielding to coincident boli.
The boli that make it through continue down the axon. One Neurofibril is available for each side of every bifurcation of the axon.
The Neurofibril carries its bolus sleeve all the way down to the synaptic bouton where the Na+ is released into the synapse, changing its ionic population.
Sent at 3:45 PM on Sunday
me: The synapse responds by increasing its rate of neurotransmitter release, relative to its reuptake rate.
The neurotransmitter released into the cleft is held in the cleft by walls surrounding the cleft.
While in the cleft, neurotransmitter causes target cell proteins to either bind or release Ca++. Binding causes excitation, releasing causes inhibition.
Sent at 3:48 PM on Sunday
me: Either way, the target cell produces Nitric Oxide which diffuses rapidly in all directions, increasing the rate of reuptake of neurotransmitter, effectively shutting down the signal generated by the bolus by negative feedback.
Sent at 3:49 PM on Sunday
me: The synapse slowly recovers by coming to an equilibrium balance of neurotransmitter in the cleft versus neurotransmitter kept within the bouton in vesicles.
Sent at 3:50 PM on Sunday
me: So, for an "irritable" cell (one that generates a pulse), the target cell always experiences a transient because its own feedback shuts off arriving signal immediately after arrival.
I will pause and take a breath.
What do you think so far?
Sent at 3:52 PM on Sunday
me: BTW, as usual, this is entirely fictional designed to fill in the gaps where the literature fails to explain what is going on.
David: I'm generally following but I'll have to back up and work my way through it. I think I know where this is going too, so I'm sitting here with a grin, but I'm not going to jump ahead. I want to watch it develop.
Sent at 3:55 PM on Sunday
me: I believe I have "closed the loop" from excitation/inhibition pairing as Ca++ binding/unbinding to the generation of the bolus to its propagation through the cell to its coincident aggregation at the hillock to the propagation down the axon to its effect on the relative rate of output/input of neurotransmitter affecting the Ca++ binding/unbinding. Is there any key part left out?
Sent at 3:58 PM on Sunday
David: I don't think so, but I'm working on it.
Sent at 4:05 PM on Sunday
me: Remember what I said last night. The bolus may jump from Neurofibril strand to Neurofibril strand at the behest of the nucleus which arranges probabilities to whatever it fancies.
Sent at 4:06 PM on Sunday
David: Let me deal with that last idea later.
Sent at 4:08 PM on Sunday
me: I will shut up for now. Later I will tell you how probability clouds expressed as arbors leads to model formation and prediction.
Sent at 4:10 PM on Sunday
David: That's a good idea. Right now let me take the current stuff on and roll it around a bit.
me: I sure can talk some shit, huh?
Sent at 4:14 PM on Sunday
JDL 20090525 NOTE: ACCIDENTALLY USED Tubulin where Neurofibril was intended. It was corrected in above chat.
