Umlaut vs. diaresis

[<mari.voipio@iki.fi>]

[This message includes lots of linguistics and quite little computing and
ConTeXt and there are lots of special characters [esp. umlauts] in the
text. You've been warned.]

On Thu, 9 May 2002, Hans Hagen wrote:
> Subject: Re: Font problems (special characters, URW)
>
> At 12:49 PM 4/26/2002 +0300, mari.voipio@iki.fi wrote:
> >I agree - all my ä:s (\"a), ö:s (\"o) and Å:s (\aa) typed directly with a
> >Finnish keyboard look kind of funny in the ConTeXt output, because the
> >umlauts/rings are so high up.  Also to my eye \aumlaut and \oumlaut are a
> >lot closer to what I'm used to.
>
> is this language dependent? If so, we can remap \adiaeresis cum suis onto
> \aumlaut in finish; keep in mind that the adiearesis may use glyphs while
> umlauts use composed characters

Yes, I would say it can be language dependent. Different languages use "a
with dots" and "o with dots" for different reasons (to express varying
linguistic phenomena) and that might affect their looks, too - but I'm not
sure about this, I'm talking here from a professional linguist's point of
view, my typesetting skills are small.

What I can tell you is how the Nordic languages view the above mentioned
characters and use them:

Most Nordic languages use umlauts/diaeresis: Swedish has å, ä, ö, Finnish
uses mostly ä and ö and occasionally å (Swedish being the second official
language of Finland), Danish and Norwegian have å (in placenames still
sometimes written "aa"). In these languages we cannot talk about the
"real" umlaut, i.e. pronunciation of a vowel changing because of other
vowels or as marker of a verb's time etc.  Finnish _never_ had anything
like that and the Scandinavian languages lost the connection hundreds of
years ago. This can among others be seen in the way we sort text, a and ä
are separate characters: Finnish/Swedish alphabet goes
a,b,c,...,x,y,z,å,ä,ö and Norwegian/Danish alphabet goes
a,b,c,...,x,y,z,æ,ø,å/aa (i.e. Aabenraa and Åbenrå are sorted similarly).

Thus, in our minds, ä and ö are "glyphs", we don't think about them as
composed characters. Icelandic is a trickier case, because it does have a
real umlaut a->ö, but I think they also think of "ö" as a glyph, at least
they sort it separately in dictionaries and there are words where ö
appears without the umlaut phenomen (like computer=tölva).

The result of the above is that we usually place the dots (diaeresis) or
rings quite close to the rest of the character as they are parts of the
same glyph. In ConTeXt the closest presentation of this seems to be, funny
enough, \aumlaut and \oumlaut while there isn't any fix to å (a ring); /aa
(or å directly from my Finnish keyboard) really looks odd in the ConTeXt
output as the ring is very thin and quite high up, looks a bit like it is
about to fly away.  :-)

However, to me this is a minor problem as long as I get the letters with
diaeresis to show up at all and *that* has been solved couple of weeks
back already with \enableregime[il1]. The exact position of the dots over
a and o and the ring over a belongs to the "I'd like to see this fixed
[before I start writing my master's in Swedish]"  category rather that
"must fix".

As I said earlier, I can easily provide pictures to illustrate the
problem, explaining it in words feels a bit inadequate. I also apologise
to anyone whose email program didn't like my special characters, I'm aware
that they may look very odd in some clients.

Mari Voipio
(in my other life student of Scandinavian languages at Helsinki
University; in this life just a common technical writer wanting to use
ConTeXt for my manuals in umpteen languages, Swedish included...)