Re: CHAT Stambul (was: A new version of Genesis)

On 14 Jun 2004 Roger Mills <rfmilly@MS...> wrote:
> I too have seen the _eis ten polis [istinpolis]_ explanation favored over
> "corruption" of Constantinopolis, but have my doubts.  For one thing, why
> should [i] change to [a]?
and
On 14 Jun 2004 Nik Taylor <yonjuuni@EA...> wrote:
> I don't question the "City" part of the story, I question the
> preposition "in".  I could easily see "The City" becoming analyzed
> as a name, but the incorporation of a preposition seems a bit
> far-fetched to me.

  (IMHO rather _eis tEn polin_ [istinpolin]. "E" is for Green eta.)

  According to my sources, the first mention of this etymon was in
a 10 century Arabic source as "Istan Bulin", that is in two words
and with the ending "-in".

  Every element of the changes have reasons:

1. The lexicalization of a prepositional phrase into a single word:
The etymon was created by Turkish-speaking inhabitants of Asia
Minor (or of the Balkans as well). It was not a prepositional
phrase for them, rather a frequent single denomitation heard from
Greek speakers. E.g. we have examples among geographical names in
Carpathian Basin, mainly in Hungarian-Slav relation, where
prepositional or suffixed names became lexemes in another language.
Or the opposite: where non-morphematic parts of the names were
treated as a morpheme (suffix, article, preposition etc.) in the
other tongue, therefore they were removed.
  (Comments: In Hungarian we still use Hungarian definite article
"a(z)", e.g., before journal names like "The Times", "Le Figaro",
"La Stampa". If you know Has^ek's brilliant work "The Good Soldier
S^vejk", you can recall that the name of his favorite pub was "U
Kalichu". You can still visit this pub in Prague and you can drink
some beer in "U Kalichu". Note that dispite the fact the "U
Kalichu" is a prepositional phrase meaning "to the bowl", we treat
it as a single name in Hungarian and in English. Etc.)

2. Change [i] to [a]: The Turkish idioms in Asia Minor know vowel
harmony. The form "Istin_Bulin" was anti-harmonic, therefore
several processes to avoid anti-harmonic structure were evoked. The
first of them was the change of [i] > [a] due to the regressive
harmonic assimilative effect of [u]. The middle stage of this
process could have been the sound [1] (as the back pair of the form
[i]) and the [1] ~ [a] alternation is not a rare phenomenon in
Turkish languages.

3. The omission of ending "-in": As we can see, this happened after
the [i] > [a] change and not earlier than the 10-11th century. I
suppose this omission was a result of two processes: (a) "-in" was
also anti-harmonic therefore it had an instable position; (b)
frequently used geographical names are often shortened, therefore
the anti-harmonism of "-in" was resolved not by a vowel change but
by its omission. (A third possible factor: Turkish people was
present in the Balkans living together with Slavic people. In the
Slav languages, ending "-in" is quite frequent in geographical
names and this "-in" is often used to Slavonize place names, cf.
western part of Hungarian capital "Buda" and its Slav name "Budin"
< "Buda" + "-in". Its probably that these zero ~ "-in" alternations
served as a pattern.)

4. Why is there an anti-harmonic initial [i], though? According to
my source this [i] is not primary, that is not the continuant of
the "ei" in "eis tEn polin". This "ei" was omitted* already in the
Byzantine Greek, cf. present-day Greek compound "stEn" [stin] <
classical "eis tEn". This [i] is a Turkish prosthetic vowel to
avoid initial consonant cluster [st]. During the borrowing process,
these prosthetic vowels are not part of the underlying form, they
are inserted automatically and unconsciously: it's typical that the
speaker does not know that he/she added a plus vowel. This means
that the prosthetic vowel is not part of the word and it's not a
subject of vowel harmony rules. Note that we can find often a front
prosthetic vowel on back words, this is true also for the Hungarian
(I will give Hungarian examples because I know better this
language. In this field Hungarian shares Turkish typology). Thus
we have Hungarian "iskola" from the Latin "schola", archaic
"ispota'ly" from "[ho]spitale", "istra'ng" from German "Strang"
'trace (on horses)' etc. This is because the harmonic boundary
between front prosthetic vowel and the back remaining can help to
mark the original word boundary (note that compound words can be
anti-harmonic: the harmonic boundary here marks a strong
morphematic boundary).
  (* The above arguments can be true even if the [i] of Istanbul
comes form the "ei" of Greek "eis". This primary [i] could have
been re-interpreted as a prosthetic vowel.)