If books are common, telephones must be rare (was Re: Rate of Change)

[Robert Whiting]

On Wed, 27 Jun 2001 X99Lynx at aol.com wrote:

>I wrote:
><<If you said "pad" and "philadelphia" have no surface
>similarity, that might work.  But "pita" and "father" clearly do
>have some prima facie structural similarities.>>

Especially if you come from certain areas of New York City where
<th> is realized as [t] and have a non-rhotic phonology that
realizes both final -a and -er as [@] (schwa).

>In a message dated 6/26/2001 11:40:28 PM, JoatSimeon at aol.com writes:

>><< -- well, you're wrong there.
>>Consider the following words, all cognates and all meaning
>>"horse" and all derived from *ekuos
>>jor
>>eoh
>>'sp
>>yuk >>

>No I think I'm right.

No surprise here.

>Consider the following words, not apparently cognates and ALL
>meaning "horse:"

>caballos
>ko:n
>so qui li
>marka
>umma
>nag
>lo:
>zaldi
>mustang
>hippos

>The words in my list make the words in your list all look pretty
>close, in comparison.  As I said, it is a matter of degree.

This is a fairly spectacular non sequitur even by your usual
standards.  The part of your original claim that you have wisely
snipped (without indication) from your original post was:

  I would suspect that "found" cognates that truly have no
  recognizable resemblances are rare.

This is what JoatSimeon at aol.com was replying to and now you claim
that the fact that unrelated words often have no recognizable
resemblances proves that cognates that have no recognizable
resemblances are rare.  As I say, a non sequitur of truly epic
proportions.

It is much the same as saying that looking at alligators and
butterflies proves that related species that are not similar are
rare.

>And that's the problem with subjective evaluations.  They lack
>objective measure.

No, that's the problem with letting jockeys bet on horse races;
or a judge bet on the outcome of a case he is hearing.  They can
decide who is going to win.

But the only subjective evaluation is yours.  Your list of words
has no standard by which it can be measured.  All the words in
the first list were once the same word (that's what "cognate"
means.  None of the words in your list were.  Therefore it is
possible to measure the distance between the words in the first
list and their ancestor but it is not possible to measure the
difference between the words in your list because there is
nothing that they are all related to.  But you are claiming that
because your list is subjective, then the first list must be
also.  Just another non sequitur.

Whenever you have two interrelated binary choices there are four
possible outcomes.  If the binary choices are related and
unrelated and similar and dissimilar then you can have:

  1  related and similar
  2  related and dissimilar
  3  unrelated and similar
  4  unrelated and dissimilar

You are saying that because 4 is common, then, a priori, 2 must
be rare.  This is practically the definition of a non sequitur.

>That's why a scientific approach calls for at least some
>precision in measures of variance.  As in thermometers, radiation
>spectrums and scientifically valid rates of change.

And the final non sequitur (for this posting).  Science is not
the ability to measure things accurately.  Science is the ability
to prove claims through evidence and argumentation.  Science is
the ability to explain phenomena or processes in a rational and
consistent way based on evidence.  Science requires evidence, not
accurate measurement.  Accurate measurement is not part of the
definition of science.  It may be essential in some sciences
where it is part of the evidence, but it doesn't enter into some
others.  Again, the trick is to know the difference.

Bob Whiting
whiting at cc.helsinki.fi