Re: Patriarchy: Re: What Matriarchy?
Gerold Firl (geroldf@sdd.hp.com)
21 Aug 1996 19:10:06 GMT
In article <woody-1608961324210001@192.0.2.1>, woody@alumni.caltech.edu (William Edward Woody) writes:
|> In article <4uqla0$300@news.sdd.hp.com>, geroldf@sdd.hp.com (Gerold Firl) wrote:
|> > When you say that science assumes that no one fiddles with the
|> > experimental results, I'm not sure if you are referring to supernatural
|> > fiddling (a trickster god who messes with the data) or regular old
|> > human fiddling. Science does recognize that both error and fraud are
|> > part of life, which is one of the reasons for the emphasis on
|> > replication. Unless a finding can be replicated, it's treated with
|> > caution.
|> The history of the Milikan oil-drop experiment is very apropos here.
|> It's something they teach every first year student here at Caltech.
|> (Mostly because he was a pretty famous 'tech prof for whom the
|> library is named.)
|>
|> He measured the mass of the electron. The problem was, he threw out
|> experimental data which suggested the mass was half of what he thought
|> it should be. And he came up with a number about twice the accepted
|> mass we believe the electron has today.
|>
|> But since he was such an influential guy, and such a prominant scientist,
|> everyone who went to reproduce his findings also threw out the
|> experimental data which suggested that Milikan was wrong. But as
|> time went by, later scientists threw out less and less data. (You see,
|> one of the things any good scientists does during an experiment
|> is to throw out 'obviously incorrect' data. What makes the data
|> 'obviously incorrect?' The underlying assumptions the scientists
|> has about the subject before conducting the experiments.)
This example seems to support the view that the scientific emphasis on
replication will catch human error and fraud. Not only were milikan's
mistakes rectified by subsequent researchers, but now his experience is
taught to undergrads as an example of why you should go strictly by the
data, not your preconceptions. When those caltech students are doing
experiments on their own, and are tempted to throw out some suspicious
looking data, they will think of milikan. (and remember, sometimes you
simply get faulty data, and it *should* be thrown out; the difficulty
is in knowing the difference between faulty data and a faulty
hypothesis.)
|> > As far as supernatural fiddling is concerned, there doesn't seem to be
|> > any need for that hypothesis.
|> Why?
Because there is insufficient evidence for it. If the first law of
thermodynamics was inoperative every easter, that would be a strong
indication that the natural needs to make room for the supernatural.
That doesn't happen, however.
|> And acknowledge something so many people in the world fail to do:
|> that I could be wrong, and my being wrong could reach up one day and
|> bite me in the ass.
You sound like a scientist.
--
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Disclaimer claims dat de claims claimed in dis are de claims of meself,
me, and me alone, so sue us god. I won't tell Bill & Dave if you won't.
=-=-=-=-=-=-=-=-=-=-=-=-=-=---- Gerold Firl @ ..hplabs!hp-sdd!geroldf
|