Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!rutgers!apple!motcsd!hpda!hpcuhb!hpindda!kmont From: kmont@hpindda.HP.COM (Kevin Montgomery) Newsgroups: comp.cog-eng Subject: Re: "Cognitive queueing" (was: Typing versus Handwriting) Message-ID: <3500008@hpindda.HP.COM> Date: 27 Jun 89 22:07:11 GMT References: <252@sundown.ACA.MCC.COM> Organization: HP Information Networks, Cupertino, CA Lines: 28 Warner@s3dawn.UUCP (Ken Warner) writes: >Maybe this is reflective of how much of who-knows-what neuro-transmitter is >available. Does sleep allow for the replentishment of neuro-transmitters? >Maybe we have all the information we need to come to a solution but don't have >the gas to put it together and present it to ourselves. After sleep we have >the juice--all the pieces are there and kerplunk, it fits together. I think the "aha" phenomenon has more to do with interference (ie. noise). Putting on my connectionist hat, when trying to recall something after laboriously concentrating on it could produce the condition where known units similar to the desired concept are saturated, as is the desired unit. However, determining the most activated unit is difficult due to the large numbers of highly saturated related units. After "clearing one's mind" of these things by sleeping, etc, the activation levels decrease to less-than- saturated levels and finding the more activated would be easier (depending on the nature of the activation depletion function) when presented later with only the relevant stimuli. (the "allow the system to reset and repeat only inputs relevant to problem" idea) This would also help to explain why recall is better after studying and sleeping, as opposed to studying, followed by an unrelated task. The unrelated task could potentially use some of the same primitives as the recall stimulus and would therefore continue to activate and interfere with the recall. Sleep however, would allow activations to return to "ground state" and recall from cues would be presumably better. just speculation, kevin