Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!mips!pacbell.com!news.arc.nasa.gov!uhccux!uhunix1.uhcc.Hawaii.Edu!davidh
From: davidh@uhunix1.uhcc.Hawaii.Edu (David A. Helweg)
Newsgroups: sci.bio
Subject: Re: BATS revisited
Summary: lab vs field performance
Message-ID: <13592@uhccux.uhcc.Hawaii.Edu>
Date: 23 Jun 91 00:11:43 GMT
References: <1574@gagme.chi.il.us>
Sender: news@uhccux.uhcc.Hawaii.Edu
Organization: University of Hawaii
Lines: 48


> 
> Article 4646 (45 more) in sci.bio:
> From: troach@netcom.COM (Tom Roach)
> Subject: Bat's sonar revisited
> 
> fairly good presumption that it uses sound, and "chirped" sound at that, to 
> dothis....   For a bat, or anything else, to be able to distinguish a 10
> nanosecond differential in time of arrival of a sound wave is difficult
> for me to believe.
> >Anybody out there with more information, or better yet, an explanation
> to discount/explain anything I have said above?

The "10-nanosecond" paper that I referred to is written by J. Simmons et al.,
and appears in Journal of Acoustical Society of America in 1990.  I don't have
the exact reference at my fingertips and there is almost certainly more than
one article by Simmons' lab for that year.  I'll post the exact reference when
I can find it.

One important point to keep in mind here is that Simmons collects his data in
the lab, using "acoustic jitter" methodology.  This requires a bat lying on a
platform to discriminate between an artificial echo that is generated after a
*fixed* delay and an artificial echo that is returned after a stochastic delay
with a mean delay equal to that of the fixed delay.  Thus, the echoes are
"jittered" in time-of-arrival.  The bats Simmons used were capable of discrim-
inating between stationary and jittering echoes with a jitter envelope of 10
nsecs.

As you have guessed, this setup is anything but ecologically realistic.  The
bats crawl to the correct side (the jittered side) of the platform and get a
mealworm reward.  They are not even hearing true echoes, rather, they are 
presented with a delayed copy of their outgoing cry.  But the important point 
is that the animals seem to be able to detect those incredibly miniscule
portions of time.

This is truly a puzzle, in terms of neural coding.  Certainly individual
neurons have a margin of error in their firing rates and neurotransmitter
release of about a millisecond.  But it is a far stretch of the imagination
to think of even a population of neurons that cascade onto another layer (etc)
that could achieve the kind of resolution Simmons reports.  I'm certainly not
saying that it isn't possible, but after working in the field of animal
cognition for some time, I would first look to see if the electronics and
behavioral paradigm weren't providing some kind of cue to the bats....

As for ecological validity, do humans need the neural system that allows us
to invent email for survival? (I'm sure I'll get it for that statement ;-))

dah