Memory Processing & Dreams [was Re: Sleep]

Robert J. Bradbury (
Sat, 30 Oct 1999 20:31:34 -0700 (PDT)

On Sat, 30 Oct 1999 wrote:

> It's strange, though, that we don't consciously perceive any difference
> in our "knowledge base" or our memories after sleep.

Well, after I sleep I'm more "clear-headed". I think you hit it on the head with "consciously". If the purpose of the information is to improve your "reflex reactions" (physically, emotionally, etc.) in "survival" situations, then you definately do not want the memory integration (and decisions) to be thought about consciously.

Gee, today I saw my brother responded by throwing his spear too slowly to stop the oncoming saber-toothed tiger. Guess I better "think" about how fast I should throw my spear the next time a saber-toothed tiger is coming at me... While we may review events and come to conscious descisions about their meaning or how to respond to them, "consciousness" is an overlay on reflex reactions and if you had to think about those reactions the time-delay could hurt you.

Stuff which is critical to our survival had to be dealt with by nature before consciousness developed. Perhaps that is why most species sleep. Also perhaps for energy conservation.

> If it is late in the evening, I have memories of the day's events, and
> also memories of the previous day's events. The former are somewhat
> sharper and more detailed, typically, which seems normal because they
> are more recent.

Probably because they are sitting in the short term memory bank. If we follow Calvin's interpretation probably lots of hexagonal neural meme-ettes are replaying over and over (at a subconscious level). If you go to sleep and process those into long term memory, the next day you wake up with many fewer of those meme-ettes running.
> But the previous days' events have had the opportunity to be "processed"
> by sleep, to be "integrated" or "incorporated" while today's events
> have not. By this theory, there ought to be some power which yesterday's
> events have which today's do not,

Only at a subsconscious level if my argument above makes sense. It may also be true that you process today's events into the rapid response net in relatively close to real time, but the sleep period is for conflict resolution (thats what gives the sociopaths trouble) and or properly weighting things that may be in relative conflict.

> or some fundamental difference in my access to those events from
> my access to today's.
Most daily events aren't "life threatening" enough for there to much weight on them. But if nature evolved a mechanism to do this type of memorization and priority "weighting" it isn't going to stop operating on days when nothing goes wrong.

It is pretty clear that we are much better at remembering events where there is a high emotional content. So the adrenelin must be playing a role in the quality and/or quantity and/or weightings that get attached.

> But I don't perceive
> any such difference. Whether I learned something yesterday, or I learned
> it today, I seem to have (roughly) equal access to the learning.

Go over to your stove, turn on the burner for 5 minutes, then put your hand on it for 10 seconds. After a month goes by and your hand has recovered, I'd make a bet that you will remember the events leading up to putting your hand on the stove *much* better than you remember the events of the previous day.

Whether you dream about it I'm unsure. We don't know enough about how such events are processed. But I can remember recurring dreams of childhood involving things I was afraid of. I don't think these go away as adults, we are simply afraid of much less.

> What we do see is that people go temporarily crazy when deprived of (REM)
> sleep. They hallucinate and behave in a psychotic manner.

To some degree, I'd want to check the references on this. I'm almost sure one thing that occurs is they experience "mini-sleep" periods (where they go into REM mode for a few seconds). I can imagine a situation where short term memory fills up and it activates both your "anxiety circuit" and does frequent dumps of the overflowing data buffer. The mini-dumps could be interpreted as dreaming while awake or "hallucinations".

> This does
> not seem to fit in any obvious way with this "consolidation" theory.
> Sleep deprivation does not primarly cause memory problems, it causes
> perceptual and behavioral problems.

Getting stuff out isn't the problem, the problem is the backup on getting stuff in. If this is part of a basic survival process, I imagine nature will have stacked the deck to make you a little "upset" when it isn't able to operate properly.

BTW, the "short term memory" I'm discussing is not the 6-9 item memory for things like phone numbers, it is the data buffer for the experiences of the days events (that presumably get partially or wholly wiped every time you go to sleep).