Venezia (was: Place names in foreign languages; Peking/Beijing; B urma/Myanmar)

From: Marco.Cimarosti@icl.com
Date: Mon Oct 11 1999 - 14:21:59 EDT


Wait, Sarasvati! Please don't kick me off the list yet :-)
With this post I am bringing the thread back to Unicode technical issues...

Karl Pentzlin said:
>Anybody out there who requests Shakespeare's play
>"The Merchant of Venice" to be renamed "The
>Merchant of Venézia"?

Pierpaolo Bernardi replied:
>FYI: in Venezia is not necessary to put an accent.
>However, if you nevertheless want to put the accent,
>it should be a grave accent: Venèzia.

Actually both Karl and Pierpaolo are right, somehow:

- "Venezia" -
is how Venice would normally be spelled within normal Italian text.

- "Venèzia" (with grave accent on the 2nd e) -
is how the name would be spelled on a dictionary or encyclopedia. The grave
accent is added for two reasons: to show which syllable has the stress (the
2nd one, in this example) and to show the standard (= Florentine)
pronunciation of the e vowel ("open", in this example). If the e would have
had a "closed" pronunciation, the acute accent would have been used instead.

- "Venézia" (with an acute accent on the 2nd e) -
Is how the name is spelled on the popular road maps by the Touring Club
Italiano. This important publisher use in their publication a single accent
mark because they chose to only indicate the *position* of the accent, but
not the opening of the "e" and "o" vowels. The reason why T.C.I. made this
choice is easy to imagine: there is a lot of disputing in Italy about how
these vowels should be pronounced, especially in minor place names, so the
editor wanted to avoid arguments and blunders.

(Incidentally, "Venézia" is how the name is actually pronounced by most
Venetian people: with a closed "e"; but this is a different story, and would
bring us back to the vicious language/dialect discussion).

But I promised that I was bringing the thread back to Unicode. In fact, this
discussion about Italian accents reminded me of a good argument in favor of
combining accent marks vs. pre-composed characters.

The reason is this: there are cases when some typographical attributes of
the combining mark need to be different from those of the base letter.

For instance, you could need to have a *light* accent on a *bold* letter.
This may seem strange, but it is actually used on several Italian
dictionaries and grammar books. Namely the very popular dictionary by Nicola
Zingarelli (published by Zanichelli) uses this graphical trick to
distinguish required accents (that should actually be written in normal
spelling) from commodity marks, that are there just to show the correct
pronunciation.

On Zingarelli's dictionary, an entry like "città", would be all in bold
face, including the grave accent. This is to indicate which that accent
*must* be written: writing "citta" or "cittá" would be a spelling error.

On the other hand, in an entry like "Venèzia" the accent would not be bold.
This is to indicate which that accent is there just to teach you the correct
pronunciation, but it *should not* be written in a current text.

Such an usage is not limited to Italian, on the contrary it is a common
usage of grammar literature to print negligible diacritic marks with a
different color, font or weight. In an Arabic grammar that I used, the
letters were in black but the vowel signs (that are not actually written in
current text) were in blue. I am almost sure that I have also seen a Latin
grammar where the breve and macron signs (for short and long vowels)had a
different colors from the other text. I think that red cantillation marks on
black letters are also used in Hebrew religious texts to signify that vowel
marks were a later addition to the text.

Of course, one could say that color or bold attributes are not present in
plain text, and that they belong to "higher-level protocols" that are not in
the scope of Unicode.

Yet, these higher-level protocols are superimposed on an underlying text
that is encoded in Unicode. If an accented letter is represented by an
indivisible unit, how is the higher-level protocol supposed to give two
different attributes to the same element?

Regards. Marco



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:53 EDT