Neurons and Memory

Jeebus

How does the brain knows when and where to store memories—memories such as a car accident? And how does it know whether to place the memories in long-term or short-term memory?

Plus doesn't one hypothesis called: long-term potentiation (or LTP) have any basis on how the brain is stimulated by memory? And how do neurons store this information? Like, lets say there is a cell, named Cell A and cell A is repeated multiple times there becomes a connection between the other 'neighbor' Cell B and that cell is stimulated by the other "neighbor" to Cell B and is stimulated through inputs? So they combine to form a memory cell? Is that how the process works?

Another God

I don't know too much about the brain, but I am pretty sure that things don't go either to short term or long term memory. Everything you experience which you pay any sort of attention to goes to short term memory first, and then those things which you try to remember, the things which you think about again while it is still in short term memory, these things will be placed into long term memory.

It's more like a seive process, sifting out the useless information, then sifting out the information that has no relevence for the future, leaving you only with stuff which seems important to remember.

nigeld

While we know quite a lot about the brain, including the fact that there are changes such as LTP which clearly are linked to the recording of memory, I think it would be fair to say that we really do not understand how memory works.
At the moment, I think this area comes into the 'The more we know, the more we realise we don't know' category.

Point_Particle

I agree with nigeld and the statement that "we really do not understand how memory works". However, I would be willing to take this assertion a couple of steps further and claim that we really don't understand how the brain works. We know that there are over one hundred billion of these functionally simple, mechanical cellular automata, whose collective functionality is such that we recognize the behavior of the system as a whole as being intelligent. Memory and the capacity to accumulate and retain information are certainly defining characteristics of our conception of intelligence, but how this data is stored in the conduction states of groups of neurons is still beyond the capabilities of modern cognitive science to model.

zoobyshoe

Nigeld and Point-Particle are pretty much correct about our current understanding of memory: we don't understand it.

There is one very important "location" in the brain for short term memory and that is the pair of organs called the hippocampi (hippocampus=singular). If both of these are destroyed the person loses all ability to form new short and long term memories. Their old long term memories will remain intact, but nothing they experience after the destruction of the hippocampi can be remembered. If only one hippocampus is destroyed, memory, both long and short term is only slightly diminished.

The conventional understanding is that all memories are held first in short term memory for about ten minutes before being added to long term memory. Nothing can go directly into long term memory. The destruction of the ability to hold short term memories, therefore, also destroys the ability to form long term memories.

No one has been able to discover where memory is stored, or even if it is stored in the same sense we think of information being stored in a computer. It is quite possible that short term memory isn't even "stored' in the hippocampi, but that these organs are merely "governing" its storage somewhere else for the ten minutes or so it is at hand.

Point_Particle

The problem with the (non-biological) computer analogy is that it can really only be carried so far. Computers store information as sequences of ones and zeroes, either the capacitor is charged or it is not. So far as I understand it, biological computers, while similar in the sense that they are formed of a complex interaction of a multitude of binary state components, are far more dynamic and unpredictable, at every level, than your typical silicon microprocessor. While it is true that at any given moment, a single neuron is either conducting or it isn't, the algorithms that govern these conduction states vary with time. This is not true with non biological computers. Zoobyshoe's "ten minutes" seems like, with all due respect, a drastic simplification of neurological mechanics. It is redolent of the right/left brain dichotomy in which one side governs logical/mechanical processes and one governs artistic or creative processes. The division between the "right" and "left" components of the brain is only useful for making it easier to talk about different modes of thought. That is, it's a semantic argument, and not a scientific one. I feel the same way about long/short term memory. We don't understand the mechanisms governing the transcription of empirically accumulated information into neural code, so I don't see how we can talk about an absolute difference between short and long term memory, if there even is a discernable difference.

zoobyshoe

For evidence of the existence of short term and long term memory you might want to research concussions, Korsakov's Syndrome, and short and long term memory. In addition to Korsakov's Syndrome there are at least a couple other conditions that prevent short term memory from becoming permanent. Neurologist Oliver Sacks has written two pieces about this. One The Lost Mariner is in his famous book The Man Who Mistook His Wife For A Hat, and the other appears as a side story in his book The Island Of The Colorblind. In both cases the men in question have a memory span of about ten minutes. Everything farther back in the past than that just goes into oblivion for them. The other common evidence for a short term memory of about tem minutes come from people who are knocked unconscious by blows to the head. They can never remember what happened because their memory is blanked out from about ten minutes before the blow to when it happened.

The existence of short and long term memory is an accepted tenet of neurology, and isn't an oversimplified version of anything.

(Redolent is a cool word, isn't it?)

nigeld

Each Sunday for the last year I have played snooker with someone who is increasingly suffering from Early-onset Alzheimer's. It has proved to be a remarkably scientific way to monitor its progress, and how his short term/long term and other cogitative functions are affected. Although it has of course been extremely sad to see him suffer so.
The first sign of problems was with the scoring, and then with keeping track of whose turn it is (which are areas that all of us seem to problems with on occasion) Next it was with whether he should be going for a red or a colour. More recently, it has been which ball to hit with the cue. However, Pete’s stroke play is still totally unaffected, as is his judgement as to how best to take a ball to a pocket. What is also interesting is that when Pete is not under stress his memory improves dramatically and he can go a whole game without fault

zoobyshoe

Whenever I read about someone diagnosed with Alzheimers who apparently goes in and out of the memory problems it makes me wonder how carefully they were diagnosed.

There is a kind of simple partial seizure that mimics Alzheimers called a jamais vu. Sometimes this merely subtracts the sence of familiarity from a situation such that the person feels the people around him are imposters because they don't feel right. Sometimes it subtracts their memory of one element of a situation. One woman having this kind of seizure complained to the bank that her ATM card wasn't working. It turned out she had lost all memory of the need to punch in her code to make it work. In severe form the person cannot recognize people they've known all their lives. I have read more than one account of people getting lost on their way home during this kind of seizure, and recalling after it was over that they stood and looked down their own street, or even at their own house, and rejected it as not theirs, wandering away to try and find the right street and house.

Stress is the #1 seizure trigger in people with chronic seizures. It raised a red flag in my mind when you said your friend's memory improved so dramatically when he was not under stress. "Early-Onset" was another red flag.

It is quite possible for a person to experience status epilepticus, non-stop seizure activity just about indefinitely if there's no intervention.

I first became aware of the jamais vu in reading the book Seized by Eve La Plante. One of the people she writes about "Charlie" suffers exclusively from this kind of seizure. It first came on suddenly when he was sawing wood. He went into a bewildered Alzheimers-like state and did not know where he was. He wandered around the woods near his home in Vermont for hours untill his daughter got worried and went looking for him. When he saw her he said "Who are you?". She got him to a doctor who wisely thought to give him an EEG right away, and discovered his brain was seething with seizure activity. A simple partial seizure: no loss of consciousness. That story in particular has made me worried about any diagnosis of Alzheimers in which the person seems to have inexplicable periods of lucidity and intact memory.

With all due respect, all the personal anecdotes I have heard from people in Great Britain regarding their neurologists, indicate that they are the poorest informed about seizures, which manifest in an enormous variety of ways from patient to patient, depending on what part of the brain the seizure activity is limited to.

Point_Particle

zoobyshoe, all I meant was that I think rigidly quantifying the difference between short and long term memory might be premature. In all likelihood, the discrepancy is far more subtle. I've heard about the man who mistook his wife for a hat. Also I've read about a man who was in a motorcycle accident, suffered neurological damage and was forever thereafter unable to identify birds! These observations made of human memory and recognition definitely support the idea of modular and domain specific neural hardware!

zoobyshoe

I am not aware than anyone has "rigidly" quantified the difference. "About ten minutes" is a pretty loose and imprecise quantification.

What discrepancy?

That's nice. The specific chapter that might be informative to you is called The Lost Mariner which is about a man with Korsakov's Syndrome. This prevents his short term memory from getting transfered into his long term memory. He always can only remember the past ten minutes or so. Korsakov's Syndrome is the result of years of heavy drinking during which the alcoholic neglects a proper diet. The long term lack of one of the B vitamins in specific is known to be what causes the damage to memory functions. It is Korsakov's Syndrome that gave rise to the popular myth that alcohol kills brain cells. It doesn't, but severe alcoholics who chronically sacrifice eating in favor of drink can end up starving themselves of this important B vitamin for years untill their memory is permanently damaged. I brought your attention to the story simply because it is a widely available case report about Korsakov's Syndrome, a pathology which is dependent on the existence of the short-term/long term memory for its symptoms.

An unusually specific form of agnosia. There's probably more to the story.

The existence of "modular and domain specific neural hardware" has been known about since Broca discovered the important language center that exists in the lateral left frontal lobe in 1861, now called "Broca's Area" in his honor. Since then neurologists have learned what most parts of the brain do, and the "modular" nature of the functions is not in any doubt. No one is at work trying to support this idea: it is old news. You might be interested in The Human Brain by Issac Asimov, which is a readable breakdown of the functions of the different areas of the brain. (It's a bit old and could use alot of updating, but is a place to start.) Almost any book about the brain will have alot of discussion about area-specific functions.

Zooby

Point_Particle

I disagree that there is no contention surrounding the organization of information processing subunits in the human brain. John Tooby and Leda Cosmides (1992) talk extensively about the faults of the Standard Social Science Model, a model of behavior which holds wide spread popularity and constitutes an attempt at drawing a distinction between the standard and natural sciences. The SSSM predicts that the human brain is divided into a few domain general plastic networks, and connections are set and personality is formed through experience and interaction with culture. I am in no way a proponent of the SSSM, and I bring this up only to make the point that the book on integrated domain specific neural modules is not closed.

Do you propose that all information processed by an individual's neural hardware is stored in memory? If not, then what serves to filter out the information that will not be maintained? If we define an "individual" as being all the collective properties of both brain and body, it is a fairly safe assumption to make that the individual does not have complete control over what information gets recorded in memory. If it is true then we should expect there to be certain kinds of data that are more easily incorporated as memories than others. Since computational resources are finite, active memory filtration has clear adaptive potential, and these evolutionary vestiges should still be active in neural mechanics and by extension, behavior, today. Could there be memory without filtration? Also, the evolution of behavioral heuristics governing attention would predict that humans remember certain forms of information more readily than others; have there been any experiments attempting to elucidate these forms of information?

zoobyshoe

Who made such an assertion? What I said was that the existence of (I did not say organization of) domain specific neural hardware is accepted. No one is trying to prove it. Everyone agrees, for example, that motor functions are initiated in the precentral gyrus at the rear of the frontal lobes, and then coordinated in the cerebellum and brainstem.

There is long term memory and short term memory. Without rigidly quantifying anything, it is accepted that short term memory goes from the present back about ten minutes. The existence of these two inter-related, but separate, kinds of memory consistently creates observable effects, which I have already mentioned.

The issue of what is included in memory and why is a whole different issue.