Auditory time acuity

[Pierre Divenyi <marva4!EarLab!pierre@xxxxxxxxxxx>]

Jont's statement about temporal acuity decreasing at "deeper layers" should
incite fans and modelers of the cortex to come forward with some suggestions
(if I were Neil's age, I would say solutions). Here are the facts: temporal
acuity appears to be OK (sort of, with a DL of about 6 % for discrimination
of envelope discontinuities), or at least workable, as long as the envelope
is that of a waveform in one frequency channel (ie, about 1 CB wide). Across-
frequency integration of temporal information wreaks havoc, ie increases the
DL sometimes by a whole order of magnitude. Note that at what some of us
think to be lower levels (CMR, suppression of false echoes), across-channel
integration of temporal information seems to be possible without the
detrimental outcome. Well, does not this smell of an integrator of activity
somewhere in the cortex? As a psychophysicist (although some would protest),
I would like to venture to say that, from the statistics of envelope
discrimination within and across channels, one could guesstimate the
properties of the integrator -- its waiting time distribution and, from that,
the statistics of the driving process (Mal Teich, are you there?). I know
that this is a pie in the sky, but maybe even our MEG colleagues could
kick in some information to describe, and perhaps even locate that
elusive integrator. Mind you, that little bastard is like Big Brother
watching -- a little man near the top of the totem pole.

Our resident experts may want to relate these ramblings to Neil's
Time Range #3 (auditory time, .5-100 Hz).

        Pierre