My reply to Mat's 5/3/02 post.

Glen M. Sizemore gmsizemore2 at
Sun May 5 07:20:38 EST 2002

Sorry about having to start a new thread, but Mat's posts do not appear on
my server reliably, and I can't post at the one place where I am able to
retrieve them.

> Mat: There is no reason to suppose that particular behaviours map >
isomorpically to any particular neural circuit.

> > GS: I never said they did. I am talking about delineating behavioral >
processes that underlie behaviors that appear very different. Somehow >
neurophysiology must be consistent with this, but the notion of a >
"particular neural circuit" is simplistic, though, ultimately, the notion of
> circuitry is not so bad - its just that the circuitry exists at multiple >
levels. >

Mat: You would be doing nothing more than what you complain of in
psychology - building elaborate constructions of 'underlying' behavioral
processes and then trying to see how the brain implements them while
glossing over the essential and flawed assumption that brain actually does
this stuff you observe in the way you think.

GS: You're out of you alleged mind, Mat. Your phrase "...building elaborate
constructions..." is particularly inappropriate to the experimental analysis
of behavior. The EAB is a primarily inductive, phenomenological discipline
that has produced phenomenological laws of behavior that are of enormous
reliability and generality. Such laws are facts about behavior. Any
physiology MUST be consistent with them. So, for example, if one repeatedly
extinguishes and re-conditions a particular response class, the rate of
extinction and reacquisition occurs more quickly until a steady (or at least
stable) state is reached. This is a fact about behavior - you might say that
it is not an important fact (I would disagree) but it is a fact nonetheless.
If one has a theory about "what the brain is doing" and it is inconsistent
with this fact, then the theory is incorrect or incomplete. Now, of course,
any fact out there in psychology of any sort must be reckoned with but there
are legitimate criticisms of the nature of the facts of much of psychology,
not the least of which is that they show the reliability and generality (if
they do this at all) of estimations of population parameters. In any event,
what you are suggesting is the equivalent of trying to come up with a
statistical mechanics that refuses to take notice of the phenomenological
laws of chemistry and thermodynamics. At some point, your abilities must
include explaining why the order that exists in chemistry qua chemistry does

Mat: Your previous example was rats pressing levers or pedals to get food
and you suggested that it was the 'getting food' bit was the behaviour and
not the pulling or pressing. Well gee whiz how profound, but still not very
useful. I presume you would then construct some sort of theory about
instincts and try and find how the brain gives rise to them, rather than
seeing what the brain does first and building a theory on that. You've got
it backwards. Trying to impose a behavioural theory on the brain will more
than likely blind you to what it really does.

GS: What utter nonsense. I have already dealt with it above. The
experimental analysis of behavior is silent on what the brain does in order
for the observed facts about behavior to be true. But true they are,
nonetheless, and a theory of the brain that cannot encompass them is either
wrong or incomplete. Since operant conditioning arrived on the scene very
early in the history of life, I would think that the facts uncovered in the
EAB (which do have rather good species generality) would be seen as
particularly crucial testing grounds for any neurobiological theory of
brain/behavior relations. If one's theories about brain function cannot
encompass the simple facts of operant conditioning, extinction, stimulus
control etc. that apply to a host of organisms (and not just all
vertebrates) then one would have to view the theories with suspicion.

Mat: Just because you think from observations the hunger instinct underlies
certain behaviours does not mean it actually does.

GS: You presume way too much, and you are totally missing the point, as
should be clear to you..

> > GS: I repeat, anything one says about the brain and behavior MUST be >
consistent with behavioral facts

Mat: Why? becuase behavioural facts are correct?!

GS: Absolutely.

Mat: Why should the brain be doing it the way you think?

GS: The facts are facts about behavior. They are not facts about "how the
brain does what it does to make behavioral facts the reality that they are."

Mat: And if it isn't who's wrong?

GS: The facts cannot be wrong. The only thing that can be wrong is theories
about how the brain's activity gives rise to a world in which there are
certain behavioral facts.

Mat: You or the brain? (!)

GS: The only thing that can be wrong is theories about how the brain's
activity gives rise to a world in which there are certain behavioral facts.
The behavioral facts cannot be wrong - just as the laws of chemistry are not
made wrong by a successful statistical mechanics. But a successful
statistical mechanics cannot be had if it doesn't try to explain the
macroscopic regularities of chemistry and thermodynamics, and a successful
understanding of the brain's function implies that it explains exactly the
behavioral regularities that I am talking about.



More information about the Neur-sci mailing list