unicode Siouan

Koontz John E John.Koontz at colorado.edu
Wed Jan 19 00:28:54 UTC 2005


On Tue, 18 Jan 2005, R. Rankin wrote:
> John Koontz's Siouan Doulos is based on the SIL Doulos font and is very
> good.

To clarify matters, what I somewhat ambitiously called Standard Siouan is
not really a type face, but a character set - a specified list of
character assignments to the 256 cells in an 8-bit character scheme.
Standard Siouan (abbreviated SS, I'm afraid) provides the characters
needed to represent typical modern Siouanist practice including scholarly
usage for journal publication, the Colorado Lakota Project, Ken Miner's
Winnebago alphabet, etc.  I supplemented this with three additional sets
to provide the characters used historically by (1) Dakotanists (mainly
Riggs and Buechel) and (2) the BAE, plus (3) a small supplementary group
of characters peculiar to modern phonetic transcriptions of Siouan
languages.  So there are four character sets, though the main one used is
the SS set.

I used SIL's font tools to produce renderings of these sets in three type
faces, which are (a) Doulos (SIL's proportional serif face), (b) Sophia
(SIL's proportional sans-serif face), and (c) Manuscript (SIL's monospaced
serif face).  SIL uses its own faces to avoid having to license industry
standards like Times, Helvetica, and Courier; or Dutch, Swiss, and ???;
or New Times, Arial, and New Courier (depending on your vendor).  Each of
the three SIL faces includes a standard upright version, an oblique or
Italic version, and bold versions of each of these.  So, with 3 faces and
4 variants of each face, there 12 variants for each of the 4 character
sets, or 48 fonts in all, one of which is Standard Siouan Doulos upright.

Well, actually, there are 60 fonts.  Because Doulos et al. aren't Times et
al. they don't actually match Times et al. in things like height,
baseline, descender length, etc.  Realizing this would be awkward I also
supply the standard ANSI (ISO something or Windows) character set in each
of the faces and variants.  As close to that as the SIL tools allow,
anyway.  This way, as long as you're willing to use Doulos, Sophia, or
Manuscript as a body text, you can get a body text that matches any Siouan
language materials embedded in it.  In fact, unless you need to use some
of the usual Windows "upper range" (128-255)  characters for something,
you can usually just use the Standard Siouan set in the body, too.  If you
don't, things can look fairly odd.

There are several problems with using the Standard Siouan et al. character
sets.  One is that these sets are only used - at the moment - by
Siouanists and you would be better off in the long run using a Unicode
scheme.  Just few years ago that was a pipe dream, but we are almost - not
quite - at the point where (a) you can get a complete Unicode-font in
every environment where you'd want it, in some face or other, and (b) that
environment supports entering and rendering things like vowel-ogonek-acute
for all vowels and not just most of them.

The Windows extended character sets are also a pretty good match for
Siouan now, and they may map to 16-bit Unicode (or some encoding of it)
now, too.  However, you will find problems with things like some
vowel-ogonek-acute combinations, all vowel-ogonek-grave (etc.)
combinations, some raised letters (as letters in their own right and not
superscripted regular letters), some consonant-hacek combinations
(j-hacek), and maybe some of the tailed-n letters (enye, eng?).  Oh yes,
also glottal stop.  This list may not be precisely correct, as I tend to
forget the details from one time to another, but I think it serves as a
quick list of the potential problem points you need to look at before
trying to use the extended Windows set on short notice.  If you don't need
these characters you have nothing to worry about.

Apart from the disadvantage of using the non-standard Standard Siouan
character set, there is also, of course, a disadvantage to using Doulos et
al. as faces. This disadvantage applies even with SIL's Unicode version of
Doulos.  The issue is that a publisher might well insist of Times or some
other face or set of faces to the exclusion of Doulos.  This is comparable
to the problem that Mathematicians encounter using Knuth's Computer Modern
faces with TeX.  American publishers don't like modern faces.  They like
ancient or intermediate faces (like Times), and they tend to want to pick
the fonts themselves, so they may also dislike home-brewed ancient faces
like Doulos, etc. - anything other than their usual choice.

Note:  A modern face - I think Garamond is a good example - has a strong
contrast between thick and thin strokes.  Ancient faces like Times use
about the same weight for all strokes.

The solution with TeX is that there are vendors that sell Times versions
of the TeX character set, e.g., the MathTimes set.  What Siouanists need
is Unicode tools that handle rendering of all the simple and compound
characters that Siouanists need, and do it in standard faces, if not all
faces.

We haven't achieved either one, I think, but I understand that we are
getting close to the full Siouanist character set.

I think at least one Mac adaptation of my fonts succeeded for a student of
Dick Carter, but I have never tried to keep track of the Mac adaptations
or how they were achieved, so I can't help there.



More information about the Siouan mailing list