"maxwell" <mmmaxwell at hotmail.com>:
>Not necessarily random. Let's stay with your idea-- of dreams as replay.
>By definition, this is non-random activity.
>Resolve this contradiction, please.
No, I was saying that dreams are not playback. I think dreaming is when
random synaptic connections are generated. This is to prepare for mental
activities the next day. According to the "clonal selection theory", new
memory is formed by eliminating some of these random synapses.
If this hypothesis is correct, it would be different from the theory of
"memory consolidation". One way to (inconclusively) settle this is
to see whether memory capacity is limited by the amount of REM sleep
during the previous night (I predict yes) and that the amount of REM
sleep _after_ the presentation of materials would be largely irrelevant.
Using the "study for exam" example, i would suggest the amount of REM
sleep _before_ studying should be most important.