Re: Refutations and the Middle J (was Re: Another Alxasaurus query & others)

["John V Jackson" <jjackson@interalpha.co.uk>]

--Original Message-- From: Dinogeorge@aol.com Cc: dinosaur@usc.edu
Date: Wednesday, February 10, 1999 01:53 AM



>In a message dated 2/9/99 7:13:39 PM EST, th81@umail.umd.edu writes:
>
><< Point C-G, "refute" is typically used in the scientific circles in which
I
> am familiar (functional morphology, sed/strat, systematics, evolutionary
> biology, etc.) as the antonym of "support" (and thus a synonym of
"reject").
> New evidence comes around and may either support a previous hypothesis,
> refute/reject it, or be equivocal. >>
>
>"Refute" is by no means an antonym of "support," despite above-mentioned
>popular usage. Just because there is no support available for a hypothesis
>does not mean that the hypothesis is thereby refuted. Refutation removes a
>hypothesis from competition; mere lack of support does not.
>


I tend to agree, though if a sizeable proportion of the population change
their language, I suppose it's a fait accompli.  I think it started with
politician types saying "That accusation has been refuted" etc.  A definite
change maybe, but to no particular purpose - or benefit.



Dwight Stewart said:

>
>    Normally, we use refute/support in describing how data relates to
>a hypothesis.
> A given data set may tend to support or refute a given hypothesis,
>but it does not
> necessarily settle the issue.  New data may do the opposite.
>Absolutes are
> indeed rare in science, due to the very nature of scientific
>investigation.  But,
> this is not a negative thing, as scientific knowledge isn't dogma
>(or, shouldn't be)
> and is therefore a "work in progress".
>    Honestly, in Physics anyway, some of our theories experience a
>"Lazurus
> Syndrome". :-)  One would suspect that the other sciences experience
>similar
> "ressurections"?
>
> JMHO;

(AGAA!)

Of course, since it's really difficult to say anything for certain, maybe
nothing can ever be disproven.  But I still feel the concept should remain.