what's next? : news and views from graham glass
Updated: 11/26/2002; 12:36:18 AM.

 

Subscribe to "what's next?" in Radio UserLand.

Click to see the XML version of this web page.

Click here to send an email to the editor of this weblog.

 
 

Tuesday, November 19, 2002

why do animals sleep? this is a question that i pondered about 15 years ago and came up with an answer that i've never bothered to document until today. it's based on a simple premise that i seem to remember reading a while back, but have never really cross-checked. the premise is that individual neurons accumulate waste products during normal operation that they need to "dump" during a resting period in order to continue to function properly. the more waste that's accumulated, the more abnormal the cell operation becomes and the longer it takes to dump. averaged over time, a cell might spend 2/3 of its time working within tolerance and spend 1/3 of its time dumping waste products.

OK, let's say that you buy this premise. how does this lead to sleep and dreams?

well, if every individual neuron independently decided when it would go from normal mode into rest mode, then at any one point in time, 2/3 of your brain cells would be operational and 1/3 would be at rest. now let's assume that a particular cell is more likely to stay in its operational mode if adjacent cells are also in operational mode. this would make evolutionary sense, because if you're focused on a task and 1/3 of your cells are asleep, it would presumably make the task harder. since there is less stimulus at nighttime and typically less tasks to accomplish, a brain would quickly enter into an emergent rhythm in which during the daytime, *most* cells would probablistically be awake and at nighttime *most* cells would be at rest, dumping waste. note that these are only probabilities, and there's still a finite chance that during the daytime portions of your brain would be at rest, and a finite chance that during the night some portions of your brain will awake. in addition, the probability of portions waking up during the night rises as the cells dump their waste products and increase their chance of entering waking mode.

as the majority of cells enter the saturation phase of waste, you'd expect to see a fairly rapid collapse into sleep mode, since when a cell goes into rest phase, it stops stimulating its neighbors and makes it more likely that they'll also go into rest phase.

during the night, most cells would initially go into rest phase, dumping waste products. but as the night progresses and cells being to "lighten their load", the probability that cells will transition into operational mode begins to rise. mathematically, there's a good chance that various subsets of a brain will spontaneously go from sleep mode to operational mode and back again throughout the night, with the probablility of the entire brain going into operation mode by the morning rising to one.

when one or more subsets of your neurons go into operational mode during the night, they form a mind which is missing the parts that are still sleeping. a dream is the thoughts and memories experienced by this subset of your mind, and its content will vary based on which bits are awake.

for example, say you are partially awake and you hear an alarm clock. an awake portion of your mind might associate the alarm with a police siren, and another awake portion of your mind might extrapolate this to the possibility that you are in a car chase. the bit of your brain that would normally suppress this possiblility with the higher likelyhood that it's simply your alarm clock happens to be asleep, and so the waking part of your brain rightly operates with the highest probability deduction that is available to its parts, and you dream of being in a car chase.

dreams are many times abstractions of an underlying concrete concept/problem in ones life. this makes sense, because if a set of, say, 10 parts of your brain represent and constrain a complete concept under normal operation, and then you send 3 parts to sleep, the remaining 7 parts would tend to represent an abstraction of the original concept (as you remove constraints one by one). for example, the concept of a chair might be represented by a simultaneous contraining of the mini-facts: 1) 4 legs, 2) has a back, 3) wooden, 4) on the floor, 5) seat, 6) 5 letters, 7) begins with "C". if you remove the constraint (3), then you might also end up with a camel. if you remove constraint (3), (6) and (7) then you could also end up with a table or elephant. in general, the higher the ratio of asleep-bits to awake-bits, the more abstract your dreams should be.


1:28:51 AM    

© Copyright 2002 graham glass.



Click here to visit the Radio UserLand website.

 


November 2002
Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
Oct   Dec