Science? (was Re: The Single Parent Question)

[Robert Whiting]

On Thu, 5 Jul 2001 proto-language <proto-language at email.msn.com>
wrote:

<snip>

>From: <JoatSimeon at aol.com>
>Sent: Saturday, June 30, 2001 10:58 PM

>>In a message dated 6/30/01 8:35:12 PM Mountain Daylight Time,
>>X99Lynx at aol.com writes:

>>>This won't help you here.  I won't ask if you've ever argued a
>>>science case in Federal Court or ever did plasma analysis or
>>>worked on neural systems or did any high-order economic analysis.
>>>You've claimed an extremely high level of certainty with regard
>>>to the reconstruction of proto-languages. I'm looking at the
>>>scientific validity of that claim.

>[JS]

>> -- no, you aren't, because you don't have the knowledge-base
>>to do so.

>[PCR]

>I do not subscribe to much of what Steve writes either but I do
>believe he is asking pertinent question.

But he is not asking a question.  He is making a claim.  He is
claiming that he is competent to judge the scientific validity
of comparative linguistics without having to know anything about
comparative linguistics either with respect to the data or to the
means of evaluating and interpreting the data.

>This type of snobbish response is not at all helpful.

If it could convince Steve that he is trying to play ice hockey
with the equipment used in baseball, it would be very helpful
indeed.  But since it won't, it is not particularly helpful
except to people who aren't Steve.  But is no more unhelpful than
hubristic pronouncements that the conclusions reached by a
discipline are incorrect without knowing how those conclusions
were reached.

What is unhelpful is to claim that these conclusions are wrong,
not because they are not supported by the data, but because they
were not reached by the methods that are used in other
disciplines.  The only reasons that can be accepted for the
conclusions being wrong is that they are not supported by the
data or that there is another conclusion that is better supported
by the data.  To do this requires familiarity with both the data
and the methodology by which the conclusions were reached.
Pointing out this simple fact is neither snobbish nor unhelpful.

>Steve is not suggesting for one moment that he has the training
>in IE linguistics of many on this list.

Of course he isn't.  What he is suggesting is that he doesn't
need that training in order to accomplish the same thing that
people with this training have and that despite their training
and his lack of it, he has done it right and they have done it
wrong.  In short, he is suggesting that all that training in
comparative linguistics is a waste of time, that anyone who can
tie his own shoes can do comparative linguistics.

>He is questioning the methodology used based on his familiarity
>with it in other disciplines.

But he hasn't mentioned any disciplines that use the same
methodology.  He is choosing disciplines that don't use the same
methodology or that that have different types of data from
comparative linguistics.  He is trying to use the tools that he
is familiar with to do a job that they weren't designed for.  He
is trying to drill a hole with a crosscut saw because he only has
or knows how to use a crosscut saw.  Furthermore, he is telling
people who know how to use a brace and bit that they can't use it
for drilling holes because he doesn't know how to use it.  It is
simply a non sequitur.  He is comparing things that are not
comparable.  If he wants to compare comparative linguistics with
a comparable science, let him talk about dactylography -- about
how one goes about comparing fingerprints to see if they are the
same.

His claim is that one can't do science without accurate
measurements.  He is apparently unaware that the definition of
science says nothing about making measurements.  In effect, he is
saying that the only way to tell the difference between a square
and a triangle is to measure the perimeter and compute the area
and then calculate the ratio of one to the other, or, if one has
a different kind of measuring device, by measuring the interior
angles and adding them up.  When people tell him that actually
you can tell the difference between a square and a triangle by
comparing the shapes, he says "shapes? -- you can't measure
shapes! -- what kind of science is that?"  When he says that the
conclusions of comparative linguistics can't be believed because
they aren't produced with the methodology used in physics or
chemistry or molecular biology while at the same time making it
obvious that he has no idea what the methodology used in
comparative linguistics is, he is simply paraphrasing the old
dodge about lack of qualifications not being a bar to making
decisions.  He is saying "I don't know anything about comparative
linguistics, but I know what I like."

In short, he is not being objective about comparative linguists;
apparently he can't be because he doesn't seem to understand what
the objects of comparative linguistics are.  He confuses
terminology at the most basic levels, using terms like language
and language family interchangeably.  His only valid conclusion is
that the objects of comparative linguistics are not the objects
of physics.  What he can't seem to grasp is that that does not
make the manipulation of the objects of comparative linguistics
any less scientific than the manipulation of the objects of
physics.  Sometimes it is possible to say (scientifically)
whether two objects are the same or different without having to
(or even being able to) measure them.

>Linguistics should be done with the same, generally recognized
>scientific methodology as other sciences --- whenever that is
>possible.

It is recognizing when it is possible that is the trick.

Besides, Linguistics *is* done with the same scientific
methodology that is applicable to other sciences:  Data are
analyzed to arrive at theories that have explanatory power, that
have a test for inadequacy, and that can be confirmed or
disconfirmed by additional data.  That is scientific methodology.
Scientific methodology says nothing about a requirement for
quantification or accurate measurements of physical properties.
Differences in how this methodology are applied are usually
dictated by the nature of the data or the goals of the
discipline.  Where the data are easily quantifiable and/or
measurable, quantification and measurement may become an
integral part of the methodology, but where not, not.  Within the
parameters of the disciplines, it is as objective a statement to
say that Spanish and French are genetically related languages or
that Greek 'hippos' and Latin 'equus' are cognates as it is to
say that the base of natural logarithms is 2.71828+.  But one
cannot tell how objective a statement about genetically related
languages or cognates is if one does not know what genetically
related languages or cognates are.

Please don't mistake what I am saying here.  I am not saying that
methodology can't be criticized.  You have heard me say often
enough that Meritt Ruhlen's methodology is inadequate to discover
what he claims that it does.  What I am saying is that
methodology must be judged in relation to the data it is used on
and the conclusions it reaches based on that data.  Methodology
cannot be faulted because it is not the same as the methodology
used in another discipline where the data types or the goals of
the discipline may be different.  And it cannot be faulted
because one does not like the conclusions that it produces.  In
the final analysis conclusions are believable only because of the
evidence and the argumentation that produce them.  If the
evidence doesn't warrant the conclusions or if the argumentation
is full of gaps and inconsistencies, that is what makes the
conclusions suspect, not the fact that they weren't produced with
equations using universal constants or by a megabuck machine with
micrometric precision.

>I find this totally unobjectionable.

Unobjectionable in principle, but objectionable in malpractice.

Bob Whiting
whiting at cc.helsinki.fi