Do Sheep Dream Of Electric Androids?

Masem writes: "The NYTimes (free reg req) has a story on work that is being done to determine whether animals dream. The evidence for brain activity and REM is there, but researchers have now found that with 4 test mice which have run a maze several times, their brain activity while they sleep strongly mimiced that of when they are running that particular maze, as opposed to brain patterns when doing other tasks including running a different type of maze. It suggests that the biological function that controls dreaming is related to memory somehow."

Comments

Look at my nick.

[electricmonk]

That would be flattering... but somewhat odd...

I've never seen sheep in person.

Re:Is this that big of a deal?

[ramb]

Well yeah it is. I think the point of the study (and similar work by Buzsaki at Rutgers) is that not only does network replay itself, but it does it at a faster timescale. It's this replay on fast-forward that allows connections between individual units of the network to become strengthened (Hebbian theory if you follow this stuff) by calcium flow through the NMDA receptor. The time scales of real time playback are too long to engage the consolidation-like processes that we know about.