Grimm's Law and Predictability (ex Re: The Neolithic Hypothesis)

[Robert Whiting]

On Tue, 27 Apr 1999 X99Lynx at aol.com wrote:

>In a message dated 4/26/99 3:32:41 PM, whiting at cc.helsinki.fi wrote:

><<Predictability and repeatability are
>indeed the hallmarks of a sound scientific theory.>>

>I'm happy we agree about something, more or less.

Actually, I suspect that we would or could agree about a great
deal if you could curb the tendency to jump on the first conclusion
that comes along without looking to see where it is going or even
which direction it is headed in.  This method will always get you
somewhere, but it will seldom get you where you really want to go.
When the evidence seems to disagree with your assumptions you
should question both the evidence and your assumptions.

><<But they are the results of hypothesis testing, not the basis for
>hypothesis formulation and testing. >>

>Scientifically valid hypotheses (be they right or wrong)
>"predict" results and this is the basis of experimentation.  The
>term "predictability" may be somewhat arbitrary, but once again
>it is the common term of art.

But scientifically invalid hypotheses can also "predict" results
that may be the basis of experimentation.  And such hypotheses
are sometimes apparently confirmed by experimentation.  But
eventually it (usually) gets figured out that the results that
were "predicted" by the invalid hypothesis were also predicted
by another hypothesis that no one had even thought up yet.

You are quite correct that this is a term of art.  These people
are playing a game called Vetenskap.  The game is played by
bringing your scientific research to the attention of a small
committee based in Sweden.  First prize in Vetenskap is an
all-expense-paid trip to Stockholm to meet the King of Sweden.
Oh, and while you are there, you get to pick up a medal and a
certificate and about a million dollars in cash.  There is no
second prize.  Now since, as we have already agreed,
predictability is an important aspect of good science, people
who are playing Vetenskap want to make it clear that their
scientific work predicts so they make extensive use of terms
like "our hypothesis predicts" to express the idea that "our
hypothesis accounts for."  Since good science predicts, they
want to make sure that everyone knows that their hypothesis
predicts and therefore is good science.  Not everyone who uses
this terminology is playing Vetenskap at the top level, but there
is still tough competition for research funds, so, as you say,
"hypothesis predicts" is the common term of art.

But we seem to have undergone quite a role reversal here since
the days when you rather stridently asserted that

     There should be NO question that "usage" overwhelmingly says
     that the "definition" of a word is the "dictionary
     definition."

in response to my observation that

     Contrary to popular opinion, dictionaries do not define
     words (Academies do that, or try to) but only record
     usage.

Now you are claiming that the "dictionary definition" is
irrelevant and usage determines meaning entirely, while I am
constantly dragging out the dictionary to show that a word
is not defined the way that you use it.

<snip of three quotations obviously obtained from a web search
for the phrase "hypothesis predicts">

>What is most relevant about the three quotes above (as well as
>thousands of others) is that they all include the phrase
>"hypothesis predicts."  As a matter of language the two terms
>very often come together in this way in scientific usage.  And
>this reflects clearly an understanding that the "predictions" of
>the hypothesis are what is being tested.

But this is not the issue.  It is already agreed that hypotheses
can make predictions and these predictions can be tested by
experiment so you are wasting your time demonstrating something
that doesn't need to be demonstrated while everyone else is
saying "ho hum" and "so what?"  No one questions that some
hypotheses make predictions or that you will find a lot of
references on the web by searching for "hypothesis predicts."
You will also find a lot of references by searching for "bus
plunges" but this does not mean that it is a necessary
characteristic of busses to plunge or that anything that plunges
is a bus.

So one could paraphrase your little homily about "hypothesis
predicts" to apply to "bus plunges":

     As a matter of language the two terms very often come
     together in this way in journalistic usage.  And this
     reflects clearly an understanding that the "plungings"
     of the busses are what is being described.

And this does not prove anything more about the definition of
"bus" or "plunge" than the original does about the definition of
"hypothesis" or "predict."  The simple fact is that "predict" is
not part of the definition of "hypothesis" and "hypothesis" is
not part of the definition of "predict."  "Bird" and "fly" are
often linked as subject and predicate, but this does not mean
that all birds fly or that anything that flies is a bird.

So your contention

    The basic idea is that a hypothesis or premise ought to
    predict observable results.  Otherwise it cannot be tested.

is not only a non-sequitur, it is a non-sequitur based on a false
premise.  Some hypotheses do not make predictions with observable
results.  Hypotheses that do not make predictions with observable
results can still be tested for scientific validity.

Obviously you only know of one kind of hypothesis (the kind that
predicts) and you only know of one way to test a hypothesis (by
experimentation to verify predictions).  And this limited
perspective keeps you from being confused about what you think a
hypothesis is (something that predicts) and what predictability
is (something that can by hypothesized about [with a hypothesis
being, of course, something that predicts]).  This is so circular
that I'm surprised that you didn't meet yourself on the way back.

>Whether that is philosophically right or wrong is another
>matter.

I know this will come as a shock to you, but scientific method is
a philosophical concept.  It belongs to a branch of philosophy
called epistemology which deals with knowledge (scientia) and its
limits.  Basically epistemology tries to clarify what we know and
how we know it.  So whether scientific method or some particular
aspect of it is "philosophically right or wrong" can hardly be
considered "another matter."  Scientific method *is* a philosophy,
and whether a method is scientific or not *is* a philosophical
matter.

Bob Whiting
whiting at cc.helsinki.fi