bishop

[Robert Whiting]

[Sorry for all the quoted material (and even doubly quoted).  It was
the only way I could see to tie the whole thing together.]

On Fri, 15 Jun 2001, Eduard Selleslagh wrote:

<snip>

> 4. We also have cattus > gat(t)o, and in a very different context, but
> under the same constraints I believe, Konstantinopolis > Istambul, a
> far too complex name for the invading Turks, who didn't understand a
> word of Greek. They just kept the two syllables that caught most
> attention STAN-POL, plus voicing and adding an epenthetic vowel to
> make it pronouncible to them. (Like 'e-special' in Spanish). I wonder
> if they ever realized its true meaning (Christian emperor
> Constantine's city), because otherwise they would probably have
> changed it. Note that the reconstruction as a derivation from 'eis
> te:n polin' is a 19th c. linguist's fable based on the mistaken idea
> that the same rules apply when a word jumps between two unrelated (or
> perceived to be so) languages, as during the historical evolution of
> the same language.

Then, on Tue, 26 Jun 2001 petegray <petegray at btinternet.com> wrote:

> Why do you claim, Ed, that eis te:n polin > istambul involves rules
> jumping languages?  Wasn't the contemporary Greek pronunciation
> /i:sta:mbulin/?  It would simply be taken over as a meaningless
> phonetic string.

And on Mon, 2 Jul 2001 12:51:46 +0200 Eduard Selleslagh <edsel at glo.be>
replied:

> I never said that: I said Konstantinópolis > Istambul was the
> consequence of jumping languages, and that 'eis te:n polin' was a 19th
> century invention (reconstruction, without any attestation) by
> linguists that thought that when a word is loaned by another, very
> different, language, the same rules as in language evolution should
> apply.

> In Byzantine (and New) Greek 'eis te:n pólin' would have been (would
> be) pronounced 'istimbóli(n)', with two i's and o instead of u. The
> Turkish accent is on 'a' (Istámbul > Ger. Stambul), which shows that
> the I is prothetic, while the derivation from 'eis te:n pólin' would
> have had the accent on the 'u' (Istambúl). Even less than half a
> century ago, many Classicists studying ancient Greek hardly ever
> bothered about the notoriously capricious (to westerners, that is)
> Greek accent. To the Turks, all the syllables of Konstantinopolis
> apparently formed a meaningless string, as I noted before.

And on Mon, 2 Jul 2001 Leo A. Connolly <connolly at memphis.edu> wrote:

> petegray wrote, in response to Ed Selleslagh:

>> Why do you claim, Ed, that eis te:n polin > istambul involves rules
>> jumping > languages?  Wasn't the contemporary Greek pronunciation
>> /i:sta:mbulin/?  It would simply be taken over as a meaningless
>> phonetic string.

> <eis te:n polin> should already have been something like [is ti(m)
> bolin], as in current Greek.  The vowels would be all wrong.  And why
> on earth would the Greeks have given something meaning 'to the city'
> as the name of their capital?  Ed is right: that etymology is a fable,
> but it's so bad I'd be reluctant to blame it on a linguist.  It's in a
> class with _tip_ 'gratuity' < T(o) I(nsure) P(rompt) S(ervice), or
> **** < F(or) U(nlawful) C(arnal) K(nowledge).  It's amazing how many
> people accept such things, but linguists shouldn't.

And finally, on Fri, 6 Jul 2001 colkitto at sprint.ca (Robert Orr) wrote:

> Pete Gray:

>> Why do you claim, Ed, that eis te:n polin > istambul involves rules
>> jumping languages?  Wasn't the contemporary Greek pronunciation
>> /i:sta:mbulin/?  It would simply be taken over as a meaningless
>> phonetic string.

> why is this a problematic explanation?

Because it is unlikely to have happened.  Not that it is impossible as a
theoretical concept, but the outcome is wrong.

Failure to perceive the semantic content of a string of foreign phonemes
in place names (especially if it isn't known to the local inhabitants) is
such a commonplace that that it is the most common (and likely)
explanation of why place names generally aren't translated into other
languages, simply transcribed and phonetically adjusted.

Pentti Aalto once told me a story about a lake in Karelia which the locals
called simply 'järvi' "the lake".  Then the Russians came and asked "what
is the name of the lake?"  They were told, and so the name in Russian
became Ozero Järvi ("Lake Lake").  Then the Germans came and asked what
the name of the lake was and were told and so the German name of the lake
became Ozerojärvisee ("Lake Lake Lake").  I don't know if the story is
true or not, but Pentti Aalto told a good story.  However, if one looks at
the old British survey maps of Iraq, one finds a number of oft repeated
names of tells such as Tell Mabarif ("I don't know") and Tell Shu'ismo
("what's its name").  The surveyors asked the locals for the names of the
tells and simply recorded their answers.

And I rather think that that is how the Turks came to call the city
Istanbul.  But not by wreaking havoc on a putative 'eis te:n pólin'.

> Leo A. Connolly:

>> <eis te:n polin> should already have been something like [is ti(m)
>> bolin], as in current Greek.  The vowels would be all wrong.  And why
>> on earth would the Greeks have given something meaning 'to the city'
>> as the name of their capital?

> The capital was known as "i polis" the city for an extended period.

But "i polis" doesn't look very much like Stambul (or even Istanbul) at all.

> The study of toponyms is littered with such cases, cf. Scottish Gaelic
> Gallaibh, Cataibh.

My Gaelic isn't very good.  Do these mean "into the city"?

>> Ed is right: that etymology is a fable, but it's so bad I'd be
>> reluctant to blame it on a linguist.

> It's actually rather a good one.   There's an article in Language:

> Tiersma, Peter. 1982. "Local and General Markedness". L 58: 832-849.

> whch deals precisely with such cases, where placenames or nouns
> denoting location locative or directonal forms become unmarked

Yes, and the ancient Assyrians always referred to Assur as 'a:lum'
"the city", but nobody else did.

There is simply no need to derive Istanbul from any form of Greek 'polis'
other than its original one: Konstantinopolis.  Anything else is an
Occam's Razor violation.

It is extremely common cross-linguistically for place names, especially
long ones (over about three syllables) to become shortened by processes
that have nothing to do with regular sound changes in the language
involved.  It is a sort of way of accommodating a natural tendency
(frequently used forms tend to be easy to articulate).  It is also in
keeping with a general tendency to associate the unmarked form with the
most common function.  Such processes can involve the simplification of
clusters, the elision of one or more syllables or syllable segments from
the beginning, the end, or out of the middle or a combination of any of
these.  When we observe such developments as:

  Bethlehem       -->  Bedlam
  Alabama         -->  'Bama
  Birmingham      -->  Brum
  Kürfürstendamm  -->  Ku'damm
  Worcester       -->  Wooster
  Magdalene       -->  Maudlin (pronunciation only)
  Mississippi     -->  Missisip
  Missouri        -->  Mizzou
  Lancaster       -->  Lancs
  Lugdunum        -->  Lyon
  Neapolis        -->  Napoli
  Philadelphia    -->  Philly
  Indianapolis    -->  Indy
  etc,, etc., etc.

it makes it easy to see how Istanbul could have gotten the way it is.

Regardless of how these transformations may have come about, if San
Francisco can become Frisco, I can see no reason to doubt that
Konstantinopolis could have become StaNbul or StaNboli (N for [m] or [n]
depending phonotactics of the language involved) in the common speech of
the time (however it may have been written).  Thus German has Stambul and
Turkish has Istanbul with the normal prothetic vowel used to resolve
initial consonantal clusters (cf. Smyrna --> Izmir, etc.).

My conclusion would be that Konstantinopolis > Stambul/Istanbul has
nothing to do with jumping languages or with any putative derivation
from 'eis te:n pólin'.  And I would doubt that it was the Turks who
took the two most prominent syllables of an unanalyzable phonetic
string for the name of the city.  I expect that that happened long
before the Turks got there.  The Turks probably just took the current
unmarked form, Stamboli (or the like) and adapted it to their own
phonology.

>> It's in a class with _tip_ 'gratuity' < T(o) I(nsure) P(rompt)
>> S(ervice), or **** < F(or) U(nlawful) C(arnal) K(nowledge).

> No,it isn't.

Not quite.  The difference is that these examples are really just jokes,
a form of word play.  I once wrote several joke papers for a friend (a
shortschrift, so to speak) that abounded in entities such as the Sam
Houston Institute of Technology and the Free University of Central
Kentucky.  On the other hand, 'eis te:n pólin' as the derivation of
Istanbul looks more like what I would classify as a (semi-)learnéd folk
etymology.  Done by someone who could recognize the similarity of the
Greek words to Istanbul, but who was unaware of the phonological
difficulties, and especially unaware of the fact that there is no need to
account for the prothetic vowel of Istanbul with a Greek form.  That is
the one thing about Istanbul that is unmistakably Turkish.  And if you
don't need this, then the whole thing just falls apart.  I would tend
to blame it on a philologist familiar with Greek but not Turkish rather
than a linguist.

>> It's amazing how many people accept such things, but linguists
>> shouldn't.

> Yes they should.  It's part of the warp and woof of linguistics, much
> more so than various obscurantist mathematical formulations.

Linguists should take cognizance of them, but not be taken in by them.
Word play can provide important insights into how speakers view the
underlying structure of their language.

Finally, I would like to comment on a remark by Ed:

> I wonder if they ever realized its true meaning (Christian emperor
> Constantine's city), because otherwise they would probably have
> changed it.

Istanbul was often referred to by the Turks as Islambul.  One particularly
finds this on Ottoman coins struck there.  Obviously the -bul element
was recognized as "city."  The Turks weren't stupid even if they didn't
speak an Indo-European language. :)

Bob Whiting
whiting at cc.helsinki.fi