Wow, nice explanation! I actually felt like I understood that, without even having to read the paper.

Good post. I wondered about the significance of timing in neural processes a bit last year but lacked adequate background knowledge. I've gotten re-interested in it since reading Terry Sejnowski's Edge response this year:

http://www.edge.org/q2008/ q08_8.....html#sejnowski

I just wonder why this only getting attention recently. It always reminded me of timbre in acoustics -- spike timing just seemed like too handy a way to encode useful information for the brain not to take advantage of it somehow.

I just wonder why this only getting attention recently.

It might be because its PITA to understand all these electronic analogies for a biologist or psychologist. At least that's why it's taken me a while to get into it..

Also the ability to record from many neurons at once in live behaving animals is relatively recent (maybe 10-20 years). The other techniques are fairly technical as well and may have required greater precision micromanipulators. Aaaand recording from neurons generates tons and tons of data that has to be stored somewhere and that is getting easier and easier..

some guesses anyway

Reaction times of 0.15 seconds are common. That it the total time required for each neural stage to process the information and pass it on to the next stage. Detecting and responding involves neural circuits transversing the entire brain and passing through many neural layers. My guess is that internal processing within a layer and transmission to the next layer takes on the order of .01 seconds. With 15 stages you get 0.15 second RT. "Firing rate" signaling would then have to be several times faster than 100 Hz.

Perhaps, the spatial resonance frequency differences in the dendritic arbol act to add and subtract frequencies so that the signal transmission rate between layers can be much faster than the dendritic resonance frequencies? (Much as a fast information stream can ride upon a much slower frequency radio wave.)

(Another possibility is that "learned behaviors" tested in reaction time experiments bypass slow hippocampus neural circuits.)

While 3-20 Hz seems too slow to transmit new information, it does seem likely that such frequencies could serve to synchronize spatially separated brain regions.