Re: [lojban] Transliterations survey

[Nick NICHOLAS <nicholas@uci.edu>]

Before I come back to this issue, a little anecdote of my own on Modern
Greek transliteration.

Words Greek borrowed from Turkish used diphthongs for front rounded
vowels, and killed the rounding on I, Cowan-style. Thus:

Karago"z (hero of shadow puppet shows, cultural equivalent of Punch & Judy
or Mickey Mouse) -- Karagiozis

Ku"tahi (name) -- Kioutaxis

HanIm 'lady' -- Xanoumissa 'harem girl'

Words Greek borrowed from French and German used monophtongs, killing the
rounding. These were borrowed later and in written form, of course; and
are usually
spelled in such a way as to try and remind the reader of the original
anyway:

Duesseldorf -- Diseldorf (spelled Ntysselntorf)

Goethe -- Gete (spelled Gkaite, because ai is the only digraph that sounds
like /e/)

calorifeur 'heater' -- kalorifer (spelled kalorifer, because French has
the e in there already anyway.)

Greek has a palatal l (though not word-finally); however, it prefers to
render the French palatal l as soft-gamma [IPA j-curl, here J], the front
allophone of gamma [here G]:

Maquillage 'makeup' -- [maciJaz] /makiGiaz/

Mille-feuille 'a kind of sweet' -- [milfeiJ] /milfeiG/

Most non-Greeks hear its J as IPA j ([Janis] /Gianis/ > Yannis), which may
make things a little clearer. (So: these are roughly 'makiyaz', 'milfeiy'
in English.)

In Greek, palatals are front allophones of velars. (Allophones are
positional variants of sounds, depending on what sounds surround them. For
example, the n in 'in the meadow' and 'in chastity' are quite different,
because the tongue moves to the same place as 'th' and 'ch', respectively.
Most of you did not know that, and assumed they were both just n's.
According to linguistics, you're right: they're different allophones of
the same phoneme.)

This means that a Greek
will say [cipros], and mean /kipros/ 'Cyprus'. ([IPA c] is the palatal
voiceless stop, which Albanian writes as __q__.) It also means an Albanian
is perfectly within his rights to borrow the word as _Qipro_. But a Greek
will be confused if they see anything but __Kipro__. And though the 'g' in
Vangelis is exactly the same sound as the 'gy' in Magyaro*r*szag
([vaFelis], [mAFarorsag], where F is IPA upside-down F), be assured that
Greeks think of that sound as a g -- the same sound, that is, as in
Vangos [vagos], the pet form of Vangelis.

In Greek dialects, in fact, palatals go to palato-alveolars. So an
Athenian will say not [kenedi] but [cenedi], and a Cretan villager will
say [tSenedi] (Lojban tcenedi.) But as far as they're concerned, they're
still saying the underlying form /kenedi/.

This to me implies that Cyprus should be in Lojban kipros, not kiipros,
tiipros, or even worse (using the local pronunciation) djipros. Similarly,
I would have to advocate minxen rather than mincen: the ich-laut is a
variant of the ach-laut, and x will make Germans think ch; c will make
them think sch.

One more thing is that canonical does tend to imply (to me) canonical
according to the local hegemony, rather than the local form. That means
minxen rather than minge, niu,orlynz. rather than nolinz., and (much more
contentiously) timicuara (Rumanian) rather than temecvar (Hungarian). I
know full well this is not going to be looked on favourably.

OK, talk amongst yourselves for a while... :-)

-- 
==  ==  ==  ==  ==  ==  ==  ==  ==  ==  ==  ==  ==  ==  ==  ==  ==  ==  ==
Nick Nicholas, Breathing  {le'o ko na rivbi fi'inai palci je tolvri danlu}
nicholas@uci.edu                   -- Miguel Cervantes tr. Jorge LLambias