[From Bill Powers (960518.0700 MDT)]
Avery Andrews 960518 --
I can't resist the speculation that the inhibitory effects of
bitter-tasting food isn't the effect of a toxin-avoidance control
system. Modelling this system is a bit of a challenge, since
it shouldn't supress appetite in general, but inhibit eating
particular items or kinds of things that are discovered to be
unpalatable.
Think of the whole feeding system as a set of control systems with
different goals. One is concerned with calories, another with avoiding
toxins, another with supplying missing ingredients (salt, vitamins),
another with getting liquids, another (maybe) with weight control, and
so on. All the control processes operate at the same time, but they have
to time-share the apparatus for foraging and ingesting. Whichever error
is the largest determines the use of the lower systems. There isn't any
direct inhibition or competition; it's just that the goals may conflict
in terms of usage of the motor machinery. This is Bruce Abbott's
explanation of the suppressant effect of quinine, and I agree. It's not
"appetite" in general that is suppressed; just the ingestion of the
quinine. Of course if the quinine is contained in food, then the eating
of food will be reduced by the quinine-rejection system.
···
-----------------------------------------------------------------------
Chris Kitzke (960517.1705 EST) --
Sounds like you are trying to fit data to a theory after the data
is taken. Would it be better to conduct an experiment with the
theory first, then see if the data fits, rather than looking at
data and seeing what theory explains it?
I think it's generally agreed that the safest approach is to look at the
facts first, then try to develop a theory to explain them. The problem
with doing it the other way is that you might not be able to avoid
trying to prove that your wonderful theory is true, and this leads to
selecting the facts that are in your favor and ignoring those that don't
fit. You may have noticed that even on this enlightened Net list, some
people support their theories by giving positive examples that show
they're right, but without looking very hard for counterexamples that
show they're wrong.
Actually both approaches are used; when a theory is concocted to explain
a set of facts, the next step is to see what this theory would say about
the effects of new conditions, like introducing disturbances or changing
the environment. Now the theory is being used to make a _prediction,_
and the experiment (where you look for new data) tests the prediction.
After a while you can't tell which is the chicken and which is the egg.
You go back and forth between modifying the theory on the basis of old
data, where it doesn't work quite right, and testing it against new
data, hoping that you've fixed the problem.
Bruce's post was a little confusing, because like a good scientist he
was presenting the other side of the case as fairly and convincingly as
he could. It was only at the end that he brought up the problems with
the critique of control theory and showed where the holes in the
argument were.
I saw another example to go with Bruce Gregory's nice illustration of
real science from astrophysics. On PBS there was a show about a bunch of
archeologists looking for the graves of Butch Cassidy and the Sundance
Kid, in Bolivia (I think). They found a gravesite that was said to be
the right one, and dug it up, and found all sorts of convincing evidence
including a skeleton of the right age, sex, and size, with a suggestion
of an injury to a leg bone and an apparent gunshot wound to the forehead
that matched wounds that Sundance was said to have received. The
evidence mounted up and up, and everyone was getting all excited, until
an old picture showed up right near the end of the show. On it the grave
marker could be read, and it showed the name of the person buried there,
which was not either of the Terrible Duo. The whole team of scientists
sort of went flat, and one of them was heard to mutter "Shit!" right on
Public Television. But not one of them made a move to discard this
embarrassing bit of evidence. They just started talking about looking
into some other possible gravesites.
I guess the moral of the story is that if you're a really honest
scientist, it doesn't matter which comes first, theory or data. But you
have to be willing to endure a lot of disappointment and admit that
you've wasted a lot of work, maybe years, before you can trust yourself
to do the right thing when the crunch comes.
-----------------------------------------------------------------------
Best to all,
Bill P.