View Single Post
  #34   Report Post  
Old 15-07-2003, 02:12 AM
Bob White
 
Posts: n/a
Default Do Theories Have to be Testable to be Scientific?


"Jeff Utz" wrote in message
...
For a theory to be scientific, at least two criteria have to be met:

1) It has to be based on reliable data.

2) It has to be falsifiable.



You seem to be confusing theory with working hypothesis concerning the
theory, old boy. It is not the theory ("ETs may in reality exist" for
example) that has to be falsifiable, it is the null hypothesis that does.

Nobody ever has to prove a theory is false. The burden of proof is always on
the affirmative, and can never be sifted to the negation. The null
hypothesis stands forever unless knocked down by logically satisfactory
evidence of the hypothetical thing.


Null : of, being, or relating to zero
www.m-w.com
(as in, "There are no ETs.")


---
Testing the Null Hypothesis
by John Marcus, MD
email

http://www.setileague.org/editor/null.htm

SETI is perhaps the most highly interdisciplinary of sciences,
encompassing not only astronomy, biology, engineering and physics, but
also psychology, metaphysics, probability, and belief. But it is, first
and foremost, a science, one to which we hope to apply the scientific
method.

[...]

The Scientific Method for the Argus search is this:

There are no ET's. (null hypothesis).

.... [W]e now design an experiment (Project Argus, for example) to try to
prove that statement wrong, recognizing that it takes only one clear,
unambiguous counter-example to reject the null hypothesis. ...

---