[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: "running" elephants - locomotary analoges



HP Hutchinson wrote:
<The "you need a time machine to test your hypothesis" applies to anything
done in historical science, and is not a deeply intellectual criticism.  We
all know it but also know that various indirect methods are useful as
tests.>

I appreciate the cogent discussion of when computer simulations might be
useful and when not.
However, this last statement could use some clarification because it's not
entirely logical.  That is, if hypotheses in the historical sciences cannot
be tested, then there is no way to know that indirect methods are useful as
tests.

Just speculating, if a particular type of knee makes jumping impossible, and
if an animal is known to have that type of knee, then that animal almost
certainly didn't jump.  That type of statement is based on general
principle, and on available observation (what kind of knee the animal has).
Using an example from the discussion, a conclusion that pterosaurs above a
certain size couldn't fly would be based on general principles, but the
material of which the bones were made is unknown.  Because part of the
material necessary to draw a conclusion is wrong, the result is wrong.
So, computer simulations are useful because of their ability to combine a
large number of principles and observations, but may still come to
inaccurate conclusions because either the general principle is inapplicable
or because the observations are inaccurate.
Which is an elaborate way of saying garbage in, garbage out.
A problem will arise when there exists a set of self-consistent speculations
which happen to be factually wrong (if observations could be made).  In that
case, a whole range of different approaches to a problem using subsets of
those speculations will produce consistent results, apparently confirming
the speculations.  However, all that will actually be confirmed is that the
erroneous speculations are logically consistent.
Consistent garbage in, consistent garbage out.
This leaves any computer program or set of computer programs subject to the
same limitations as a person sitting comfortably in an armchair, sipping a
beer, and agglomerating ideas.
Seems obvious, but you'd be surprised at how many people think having a
computer program which processes massive amounts of information adds
credibility.