Richard Norman <rsnorman at mediaone.net> wrote:
> "Andrew Gyles" <syzygium at alphalink.com.au> wrote in message
> news:9219i9$eui$1 at nnrp1.deja.com...> > In article <1em22oz.1ck5cit1l7pviaN%harry at dherwin.org>,
> > harry at dherwin.org (Harry Erwin, Ph. D.) wrote:
> > > Theophilus Samuels <theophilus.samuels at btinternet.com> wrote:
> > >
> > > > Elaborate please.
> > >
> > > What about Simmons's evidence for 10-100 nanosecond resolution of
> > range
> > > in bat biosonar? The neurons involved are comparable to typical
> > > mammalian neurons in their recovery times and rates of spiking. That's
> > > pretty good evidence that the key cue is the exact timing of the
> > action
> > > potentials involved, not their presence or absence.
> > >
> > > What about Levy's work on variable timing of action potential
> > generation
> > > in the hippocampus? During replay, the neurons spike much more quickly
> > > than they did during the initial exposure to the stimuli.
> > >
> >
> > > --
> > > Harry Erwin, PhD, <mailto:harry at dherwin.org>,
> > > Senior Lecturer in Computing at the University of Sunderland,
> > > Computational Neuroscientist (modeling bat behavior) and
> > > Senior SW Analyst and Security Engineer.
> > >
> >
> > I am not a specialist and I do not dispute what you say here. Indeed I
> > find attractive the idea that information is encoded in the exact
> > timing of the spikes. Is there any information yet on what 'reads' the
> > time intervals between spikes and 'decodes' them to give back the
> > original information to the receiving neuron (or neurons)? At the least
> > a 'stopwatch' would be needed, would it not?
> >
> > Andrew Gyles
>> Am I correct in believing that timing short intervals is really
> a "differential comparison" question?
>> That is, rather than have an absolute timing sense requiring a
> stopwatch, the neuron is able to detect whether cell A fired
> before cell B or vice versa.
Yes.
>> And 10 nsec seems rather a bit much. See, for example
>> Pollak GD
> Some comments on the proposed perception of phase
> and nanosecond time disparities by echolocating bats.
> J Comp Physiol [A] 1993 May;172(5):523-31
Other labs are able to replicate Simmons at the high end, but only
Simmons seems to believe the low end. I've modeled the process from the
hair cells to the cochlear nucleus, and I believe the high end results
may be the result of hyperacuity. The number of neurons involved at each
stage is fairly small (on the order of 20000), but large enough. I may
give the modeling problem to a doctoral candidate to work on.
--
Harry Erwin, PhD, <mailto:harry at dherwin.org>,
Senior Lecturer in Computing at the University of Sunderland,
Computational Neuroscientist (modeling bat behavior) and
Senior SW Analyst and Security Engineer.