Re: [asa] historical versus experimental sciences

From: David Campbell <>
Date: Mon Aug 10 2009 - 13:00:30 EDT

>> Sure, Behe would grant that if you know (a) that two creatures are
>> evolutionarily related, then you could predict (b) that they will have many
>> of the same base sequences, and therefore (c) that they will generate many
>> of the same proteins.  But the first step is gratuitous.  If you know they
>> have the same base sequences, the information that they are evolutionary
>> relatives is irrelevant to predicting the proteins.  You can get to (c) from
>> (b).  You don't need (a).

But we rarely know b. And a tells us which creatures are likely to be
the best match. For example, if you want to find something similar to
a human protein but a little bit different, evolution tells us to look
at primates. We have no genetic data whatsoever on millions of
species of organisms, but we can be confident, based on evolution,
that primates are the best place to look for something chemically as
well as physically similar to us. Rodent are pretty close and
generally much easier to keep in the lab, so practical considerations
might favor mice (or, for that matter, even more distant organisms).

More fundamentally, the question is what you want to do. Of course
you don't need (a) to get from (b) to (c). (a) helps you get to (b),
though one could try to get there empirically. (a) is therefore not
totally irrelevant, gratuitous, etc. If you want to focus on b to c
type problems, you can, but this does not make (a) wrong, imaginary,
or useless.

> Maybe it is possible to have a strong hunch that (a) is true without having all the details of (b), <

Yes, many organisms still have not been investigated with molecular
techniques, and the vast majority of those that have been studied have
only a handful of genes sequenced, at most. Overall, morphological
data still guides a lot of our understanding of the overall
evolutionary picture. There is no a priori non-evolutionary reason to
expect a general molecular similarity between morphologically similar
organisms above and beyond any functional constraints.

>> If a physician knows that a patient has cancer of the liver, and if (let's
>> say) it happens that cancer of the liver, once begun, always proceeds in
>> certain ways, and can only be arrested by certain means, does the physician
>> need to know whether the person got the liver cancer from an inherited
>> genetic defect, from working in a certain chemical factory, or from fallout
>> from Chernobyl?

In fact, this isn't a good example, either. In the case of an
inherited genetic defect, the patient's entire liver consists of cells
that are prone to becoming cancerous. While the immediate treatment
may be the same, longer-term efforts at preventing recurrence will be
different. On the other hand, the chemical factory or Chernobyl
causes would suggest higher risk of other problems such as genetically
unrelated cancers.

Dr. David Campbell
425 Scientific Collections
University of Alabama
"I think of my happy condition, surrounded by acres of clams"
To unsubscribe, send a message to with
"unsubscribe asa" (no quotes) as the body of the message.
Received on Mon Aug 10 13:01:25 2009

This archive was generated by hypermail 2.1.8 : Mon Aug 10 2009 - 13:01:25 EDT