The Arrow of Time

"Research is what I do when I don't know what to do."

I like sleep. I like it for all those moments when you are somewhere between asleep and awake, and for all the snippets of dreams that you can recall if you wake up at just the right moment.

At some point, for a project long forgotten, I became familiar with the Crick-Mitchison theory of REM sleep. The theory is so old (1983) that it's hard to find a coherent account of it on the internet. The idea is that dreams are a kind of purging of useless memories and/or memory connections, essentially a type of reverse-learning.

I remember this theory because I see can distinct patterns in the material that appears in my own dreams. I'm generalizing what follows from a small set of data (my own experiences) that has not been compared to other existing data sets of a similar nature (if any exist), but the ideas are interesting nonetheless... What I see is a certain class of partially-processed perceptions and thoughts - things that were tagged with X amount of attention at one time, but which, for one reason or another, were never fully processed or integrated with existing knowledge. It's as if the processing was unexpectedly interrupted - 'I'll think about it later. Right now, this new thing is more important.' I find in my dreams images, thoughts, and ideas that can be traced to a similar interrupted and never-completed chain of processing.

What makes the whole thing so fascinating (to me) is that these images and sounds and ideas are woven seamlessly together into a dream in what largely appears to be the same temporal order in which they were acquired. This may not seem all that stunning, until one thinks about the fact that not only is the dream-purge mechanism identifying these specific fragments of conscious experience which seem to have undergone a similar level of partial processing, but it is also utilizing or preserving the temporal order in which those thought events occurred... Now we're into a whole new set of questions.

How does memory code time? (Relational time, not rate of stimulus presentation.)

Why do dream -generation/purge mechanisms 'sort' the 'data' they are purging according to this relational temporal coding? Is it a way to be energy efficient? If so, according to what dynamics? Is it done to preserve or strengthen the remaining temporal relationships among preserved memories? If so, how do these dream mechanisms recognize that temporal ordering? How is it coded?

Will I ever get a chance to hole up in a library and try to research this problem more thoroughly? ;) (Sometimes I think this blog is simply going to be a collection of all the questions and ideas that I don't have the time to explore fully...)

Understanding how the brain preserves and codes temporal relationships may 1) yield additional understanding on relational spatial-relationship encoding, and 2) enable us to disrupt those mechanisms and study any effects they may have on a person's ability to navigate a trajectory through the multiverse. (Oh yeah, I'm going somewhere with all of this. ;)