Why do we Dream?

I just finished reading this Digg posting about why we dream, and decided to post my own theory.

The theory is based on a simple premise; that individual brain cells accumulate waste products while they are awake and thinking, and that they need to rest and stop thinking in order to purge the waste. The ability for a cell to operate normally decreases as waste accumulates. A typical cell requires 4-6 hours a day to purge its waste. Many other cells in the body need to rest after periods of activity, so this premise seems reasonable.

Sleep and dreams are an inevitable and emergent consequence of this premise; here's why:

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 1.

When one or more subsets of your neurons go into operational mode during the night, they form a "partial 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 parts 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 part 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 often 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 constraining of the following 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 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, dream abstraction should be proportional to the percentage of your brain that is asleep.

In summary:

• Each neuron needs to rest several hours a day in order to dump its waste products.
• A neuron cannot participate in thinking while it is resting.
• A neuron is more likely to enter its rest phase when it is not being stimulated.
• Most neurons tend to rest at night.
• As neurons become rested, subsets of your mind begin to work again.
• Dreams are nothing more than the workings of a partial mind.
• Dreams are often abstractions because a partial mind cannot completely resolve constraints.

Comments

[SBG/P]

I don't necessarily sleep to dream, but rather dream to change the world.

[eon]

Not desiring to add comment to this page; just wanted to say your theory is fascinating and sounds very plausible, assuming the science of cell waste products and necessary purging is correct. If Ignignokt01's comment on the Digg page you referenced is also true (about released hallucinogens), then that implies that some hallucinogens are natural waste products of cell metabolism. Also fascinating!

If some folks have managed to adapt to a mere 4-5 hours of sleep per night, perhaps it is because their bodies have developed the capability to coordinate cell purging to all happen at once, instead of the 2/3-to-1/3 ratio you propose.

And that ratio as an explanation of the oddities of our dreams also makes a startling amount of sense. (As does the hallucinogen theory.)

The only hole seems to be an explanation for why a cell, once "purged" and awake, would go back to sleep again. You mention probabilities, but there'd still have to be a mechanism giving rise to the probabilistic behavior of the cells. I think you can find an explanation for that, though, as all the rest seems to fit so well.

All in all, the most logical explanation I've seen to date of both sleep and dreaming.

Wonder what experiments could prove or disprove this theory...

[hysterix]

This theory, and the subsequent comment are both fascinating and makes an incredible amount of sense.

kudos

[Bill Cornelius]

Hi Grahm: The percentage of wakeing/sleeping cells can't create lasting memories without brain cell consensus (rather than a majority), it seems to be necessary that they don't otherwise we'd be overwhelmed with conflicting responses to partial descriptions. So here's a test: there has to be a mechanism that knows when a consensus is achieved. The "ah-hah" moment is some kind of hormonal reward for the "correct conclusion". I don't know what kind of synapses are involved (like the state of wakefulness has a lot of hormone stimulants), but I'd bet they're there. Something that makes cells line up & fire in cascade, if any are missing, the message doesn't go through.

[Melanie Stephan]

Why do we dream? Why is life like a black box?

There is one thing that we all share in common in our dreams. Every night there is one thing we all do. It is the same for everyone. If you can answer that question you will have a better understanding of dreams.

Do you know the one same thing we all have in our dreams? All of us do it.

Dreams are your God skills?

Gods Messanger, Melanie

[Melanie Stephan]

God has a message for you. It is about the meaning of 'First is Last and Last is First'. The message is this:

In the morning I go to Heaven. In the afternoon I live my life. In the evening I die, death.

What does this mean? It means that Birth is Last and Birth is First.

Gods Messanger, Melanie