Sorites Paradox and Ahistorical Accretion

H. Mark Hubey HubeyH at
Wed Mar 10 15:29:10 UTC 1999

----------------------------Original message----------------------------
There are certain things that really get me confused. These
certain things are things done by linguists apparently
much after (or before) thought. ONe of them is something
that looks to me as if it is ahistorical accretion.

I will demonstrate via an example. The word "yoke" is said
to be IE and to derive from *PIE. But there does not seem
to be any other meaning attached to it except "yoke". From
the recent centuries (especially in the sciences and technology)
whenever new words are coined out of nowhere they seem to be
compounded words from Greco-Latin especially created for a specific
purpose by scientists. Examples are words like entropy, enthalpy,
probably energy, synchronous, etc. I find it hard to believe that
this was done 6,000 years ago. So the word "yoke" must come from
another word which was pressed into service in the new setting.
What meaning could/should it have had? Pull? Hitch? Tie? Bind?
Put up front? Tie up front?

Which brings up another problem. I seem to see two ways of
reasoning used in the literature. If a word can be found
which has not etymology in a language it is thought to be
original in that language? That seems contradictory. After all,
if we found an etymology in another language, clearly then we'd
have to lean in that direction. So then if an etymology cannot
be found, then the word should be left neutral, pending something.

But what?

How about looking into other languages?

Which brings up another problem. It is called a sorites paradox
in logic. I can illustrate it using some kind of logic related
to fuzzy logic. Suppose we could assign numbers to how similar
objects are to each other. We could do this by using distinctive
features. For example, an orange is closer to an apple than it
is to a chair. No doubt about that. How about similarity of
an orange to a lemon vs its similarity to a grape? We can make
up some relevant distinctive features. Just for the purposes of
this example, I will make up a few; size, color, skin texture,
taste (sweet, sour etc).. Orange and lemon are closer in size, and
closer in skin texture, and even in color, although not in taste.
So if we were weighting each feature equally, we'd judge orange and
lemon closer to each other.

Now let's try this on something like this: "how close is a qumquat
to a quince vs qumquat to apple?"

If we do not know anything about what a quince looks and tastes
like and we don't know anything about what a qumquat looks like
we cannot answer the question. Suppose we knew 2 out of 3. Let's
make it, apple, orange and quince. (For the purposes of this example
pretend you don't know anything about quinces). We might then
look at the apple and orange and decide that there are a lot of
things they have in common, and might make ourselves believe that
they are similar. But we all recognize that this is wrong.

But how then does the linguist who only knows IE languages
proceed to look at some language X and determine if it is
IE without knowing anything about AA, Altaic, Uralic or
Nilo-Saharan or any other language that could have been in
that region, including Elamite, Sumerian, or Hurrian or Adgyg-Abxaz,
or Vainax, or Kartvelian?

This is like looking at a morphing video of Eddie Murphy being
turned into Bill Clinton and deciding that they are the same
person. That in fact is what is done when an agglutinating
language at time t0 is said to be descendant of, say, an
inflecting language.

Now back to yoke. Is it accidental that the word for yoking/hitching
is <cek> in Kipchak-Karacay-Balkar, and that <chek> means
"to pull" in Turkish, and that <yUk/cUk> means "load" as in
a load put on an animal or a cart/wagon. Is it derived from
IE yoke?  This is just an example to illustrate a point. Do the
naysayers to Nostratic take these (not the example) into
consideration? Or do they use the rule that if they have seen
a word in writing some language, then it belongs to that language?

Best Regards,
hubeyh at =-=-=-=

More information about the Histling mailing list