Chronology | Current Month | Current Thread | Current Date |
[Year List] [Month List (current year)] | [Date Index] [Thread Index] | [Thread Prev] [Thread Next] | [Date Prev] [Date Next] |
The language "extrapolated from observations" is interesting to me,
especially when later we read "I can't...see how that can be tested"
McDermott is making a distinction between these two ideas, as if the
first indicates the possibility and the second would nail it down.
Am I reading that correctly.
If so, I want to suggest that the second does not exist in the sense
I think he means it. There is no way to prove something absolutely
true. Look at Duhem-Quine or Hanson's Patterns of Discovery. The
central idea is that any experiment built to test a theory has built
into it the assumptions of the theory, so the test is contaminated.
The best we can do is to design an experiment which will yield a
positive result if our ideas are valid. If the experiment works out,
then it supports the validity of our idea but in no way proves it.
The more experimental results consistent with our ideas, the more
confidence we have, but proof never comes.
In this sense, observations consistent with big bang or evolution led
support to the ideas, but do not prove. There is always the
possibility of some other mechanism explaining the same event.
I know this gives ammo to the ID folks...thats why its such a tough
problem, and so easily distorted by them.
cheers,
joe