Re: Teacher check

[Robert Zatorre <robert.zatorre@xxxxxxxxx>]

At 13:50 02/03/05 +0100, Martin Braun wrote:

> Music Perception was not the right source
> then. Rats hear fine at least down
> to 250 Hz and at least down to 70 dB SPL at this frequency:
> http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=893752
>
> http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=8188553

In fact, the audiograms I referred to in the Music Perception
paper were taken from the sources you cite. Attached is the figure from
Heffner et al, hearing Research, 1994, which also plots the earlier one
form Masterton plus two others. As you can see, rats can hear
lower-frequency sounds, but only at high SPLs. Some strains don't hear
much below 1kHz. We don't know what strain of rats was used for the
Japanese study you cite, but if music was played to them at around 65 dB
SPL, they would not have heard much below 500Hz. Compared to humans,
their entire freq range is shifted. Their most sensitive region is
between 8 and 32 kHz. So what were they hearing over those speakers in
that freq range, I wonder? (we don't know cause we don't hear that high)
The main point here is simply that the assumption that playing music to
them results in a similar input to their ears as to ours has to be
qualified by their very different peripheral hearing apparatus. 

> More importantly, even if their
> limit was 500 Hz, the rats would easily hear
> all notes in the Mozart piece. For a long time people could not
> hear
> frequencies below 300 Hz on the phone. But on the same phones they
> could
> easily follow a male voice down to a pitch corresponding to 50 Hz.
> (Perhaps
> the undergraduate that "flunked" your course can tell you the
> magic ;-))

I flunked him before we got to the section on missing fundamental.
;) 

But how does this work in the rat? what is the "existence
region" for them? How are the unresolved partials that they would be
hearing from all those low notes be processed? Is their cochlea identical
to ours? Should we just assume that what we hear is what they hear?
Perhaps this has all been worked out and someone on the list knows about
it. But it needs to be established, not assumed, wouldn't you
think?

> > The
> > second point, which I would expect an undergraduate to be able to
> > point out (or else flunk my course), is that the cited study used
> > no
> > control group other than no treatment. So the specificity of the
> > conclusions is, shall we say, a bit suspect.
>
> Nobody has claimed any specificity in this case.

Well, I was responding to the phrase "...we have data
showing similar signs of appreciation of Mozart's music in rats as in
humans..." which sounds rather specific to me.

So let's be explicit about the issues. We have no reason to dispute the
actual results observed; But we do have reason to question the
conclusion. What are the other factors that might have yielded the
observed result? This is what I would expect all of us to be thinking
about. Have we examined and rejected other, more parsimonious
explanations of the phenomenon? 

For example, based on the description in that paper, the rats in the
experimental group were taken to a special closed cage where they were
exposed to the music. Perhaps the characteristics of that cage and/or
room/environment were such that it was calmer (light, temperature,
absence of other animals in the room, absence of other noises in the room
etc) and led to the observed changes. Note that in that paper there were
no differences between the times that the animal was exposed to music,
and the time period after exposure ceased (a full two hours). This could
be, of course, because they were so happy with Mozart that they stayed
happy for that period of time (although you'd think that if they liked it
so much they should have become stressed when it was turned off); or it
could, of course, be that something else was making them calm and relaxed
and the cessation of the music had no effect. I would vote for the latter
until the former is proven.

I am far from an expert on animal studies. I'm sure someone who knows
about these manipulations would have far more to say about the
experimental design. For instance, people often use cross-over designs
(within-subject) to try to disentangle all these difficult variables. My
general point is that there are good reasons why one needs to be careful
about control groups and control conditions. Without them, you cannot
rule out nonspecific effects.

> The result was that the
> rats strongly responded to the presented music. The result was not that
> they
> responded differently to this music than to other
> music.

We in fact do not know that they responded to *music* per se.
Perhaps any sound would have resulted in similar effects. No other
stimulus was tested, so we simply don't know. Perhaps stimulation of any
type in any modality would have done it. It could be a generalized effect
of being stimulated as opposed to not being stimulated. Maybe being
manipulated by the physical act of being transferred from one cage to
another is a factor. A well-controlled study would have examined all
these possibilites and rejected them one by one. 

> Sorry for the discomfort of facts
> and reason.

Discomfort? Facts and reason, together with music by the way, lower my
blood pressure, I'm sure!! It's their absence that makes me
stressed.

Thanks for the good debate.

Cheers

Robert

Attachment: Heffner et al 1994 rat audiogram.jpg
Description: JPEG image

-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Robert J. Zatorre, Ph.D.
Montreal Neurological Institute
3801 University St.
Montreal, QC Canada H3A 2B4
phone: 1-514-398-8903
fax: 1-514-398-1338
web site: www.zlab.mcgill.ca