the Comparative Method and Semantics

[X99Lynx at aol.com]

In a message dated 9/1/99 11:10:25 PM, kurisuto at unagi.cis.upenn.edu wrote:

<<to be judged cognate, the words have to 1) be phonologically derivable by
regular sound changes from the proto-language, and 2) have meanings which can
have plausibly developed from some meaning in the proto-language.>>

If we just start with the definition of cognation we can be a bit more tidy
about this, I think.  If cognates are words with common ancestry, then the
methods of finding that ancestry can be compared.

Note that you use "phonologically derivable"  That is the kind of cognation
we are talking about here.  Phonological.  Words that are "semantically
derivable" are not cognates in this methodology.  (E.g., Miguel Carrasquer
Vidal once pointed out on this list that "czerwony" (Pol. "red") <<probably
derived from the root *kwrmi- ~ *krwi- "worm", from the red dye made out of
worms>> but that the sound laws did not permit calling the two words
cognates.  This at-first improbable etymology can be mapped out, but the
phonology will not certify it.)

(Larry Trask has a neat section on etymology in his textbook, where he
observes that other textbook on historical linguistics don't mention it much.
 It is really worth reading because it, because I think it solidifies the
impression that when it comes to what we call "semantics", every word has its
own history.)

So, really, the "meaning" of a word only makes phonological ancestry more or
less likely, it doesn't define it.

But the "exceptionless sound laws" support the idea that if two words are a
phonologically the same, they have a high probability of common ancestry.  Of
course there are random exceptions, but these are better accounted for by
measuring statistical uncertainty (the old bell-shaped curve) than by making
conclusions about probable etymology.  That is a much bigger ballpark and the
fences can be out of our reach.

<<For example, suppose we had a word in Language A meaning 'river', and a
word in Language B meaning 'leather'.  Suppose that the phonological form
of the words is such that they _could_ be derived from a single word in
the proto-language via regular sound changes which have already been
established.  Nevertheless, we would almost certainly not judge these
words to be cognate, because it is almost inconceivable that there could
have been any semantic developments which would produce the meanings
'river' and 'leather' from any imaginable original meaning in the
proto-language.>>

I cannot tell you how unjustified that conclusion is.  Not only can I imagine
plenty of "semantic developments" that would work.  I can also give you about
500 examples of documented developments that are much wilder than 'river' and
'leather'.  A historian with a good knowledge of ancient material processes,
articles of trade and traffic would not find the connection very challenging,
I'm afraid.  The example I gave connecting
"gay" as liturgical joy and "gay" as a description of status in the work
place is much better.  And tracking the etymology over hill and dale tells us
the two words come from the same parent.  And they ARE phonologically cognate.

Unless you have the historical background to definitely eliminate the
connection between 'river' and 'leather', your off-the-cuff impression is
just not enough to settle the matter.  Without text and context, there is
just no way of knowing if you are right or wrong.  In fact, from the
historical point of view, your assertion is really unprovable.

What's the consequence of you making that judgment?  Possibly a bad analysis
of a key cognation in two languages and a bad history of phonology between
the two languages.

The comparative method using phonological analysis is a powerful tool because
of all you said about the sound laws.  Like carbon-14 dating, it carries a
high degree of statistical value with it.  Applying it to 'semantics' (except
to get started or as a boost of confidence in results) is like using carbon
dating to measure the stylistic difference on ancient pottery.  It's a
different variable altogether.

Someone wrote here on the list something like the meaning of a word is the
response you get.  There's a fundamental truth there.  We only have half the
information we need when we have the word.  We do not necessarily know the
effect on the listener - who may well in a local context have heard "leather"
and thought "river."

<<2) have meanings which can have plausibly developed from some meaning in
the proto-language.>>

And again this is backwards.  We can only guess at meanings in the
proto-language by looking at meanings in later language.  If our guess at the
proto-meaning does not fit a clear phonological match, then it may be that
our proto-meaning is wrong.  Especially because it is a guess entirely based
on later meanings.   Reconstruction should not force us to conclude that
actual history is wrong, when it is clear that we can only reconstruct words
based on actual history.

Regards,
Steve Long