After reading all the discussion regarding bird evolution, I
realize that we scientists do a poor job of telling people how science
actually works.

Science involves making hypotheses or theories, which have
predictions which can be tested. An hypothesis which doesn't create
testable predictions is not a scientific theory. We put the predictions
to a test, and if the result agrees with the prediction, we say that the
theory is supported; if it fails, we abandon the theory or change it. It
is easy to test a theory which predicts things in the present time, such
as the theory of gravity, but it is much harder to test things which have
occurred in the past.

How do we know where Columbus first landed in the Americas? There
is no way to test this directly, but historians have made many predictions
which can be tested (e.g. does the topography of Watling Island fit
Columbus' descriptions). The same is true for evolution that has occurred
in the past. We can and have made many predictions. For example, we can
predict that in rocks of a certain age, we will find certain types of
fossils (e.g. dinosaurs) but not others (e.g. man). We can predict that
DNA sequences will be far more similar between man and apes than between
man and lobsters. Most of our predictions have proven to be correct, and
in those cases where they have failed, we have changed the theory to
accomodate the new information. It is because the predictions regarding
evolution have proven to be correct so consistently that scientists accept
the theory of evolution. We can also make predictions on the basis of the
creation science theories, but so far, in most cases these predictions
have proven to be incorrect. That is why we do not accept them as
scientifically valid theories.

Is evolution in the past a "proven fact"? This is a value
judgement, rather than a scientific judgement. If the predictions are
correct sufficiently often, we tend to say that the theory if "proven".
But that does not make it correct, and if new data which contradict it
appear, the theory must be changed. But it all depends on testable
predictions!

************************
Robert Cleland

Professor, Botany Dept. Box 355325
Director, Biology Program, Box 355320
Univ. of Washington
Seattle, WA 98195-5325
Phone (206) 543-6105; FAX (206) 685-1728