Re: The Geejay

From: Asmus Freytag (
Date: Fri Jan 04 2008 - 13:44:37 CST

  • Next message: Asmus Freytag: "Re: chairless hamza"

    On 1/4/2008 1:08 AM, André Szabolcs Szelp wrote:
    > Of course, if the glyph's use can be demonstrated over several books of several different publishing houses over a nontrivial span of time, it should be; as a historic character.
    Given the publishing house in question, I would assume that the
    particular publication was in wide use at the time. That may affect its
    status as a historic character.
    > Coming back to your "compiled little consideration", while keeping my abovementioned reservations, I'd say it's definitely a caseless phonetic character (this can also clearly be seen from the examples you provide).
    > While it's true, that used with other letters the shape and size may be slightly irritating to the eye (mânGer)*, however I'd like to remind you that the encoding of this character -- if at all -- is for representing historic text and not to reintroduce a new phonetic character for contemporary usage! A casing pair on this letter would be the introduction of a non-existent (not 'unencoded'!) character to Unicode.
    Correct. Also, passing aesthetic judgment on the notation is, as you
    note, beside the point.
    > Also note, that those at Langenscheid back then did not care about the aesthetic aspect of one phonetic character not matching the others. An antiqua-based ligature definitely stands out more from fraktur than a antiqua glyph with ascender and descender in an antiqua font.
    > They did not chose to care about this aspect, even though they definitely had to create a new (physical) cut for that character, they chose not to go for a Fraktur-based solution to reprezent zh (it would have been none the more complicated), but created a foreign looking character: it would never show up in book text, after all, it's just a phonetic character, actually, more of a symbol, than a letter!
    Actually, there's probably more to it. If you read works set in Fraktur,
    you will find that anything 'foreign' is always set in antiqua, the
    exception being well-known foreign names (e.g. Newton). Therefore, I
    don't find it surprizing at all that novel shapes are based on styles
    other than Fraktur.
    > Also, it would be helpful if you identified the dictionary/edition more exactly, citing exact title, publisher, year, place of publishing as it is common (you gave the publisher and the year, but not the exact title, just it being a German-French dictionary, nor the place of publishing).
    More to the point, the excerpt should be increased to include the full
    notational system plus a full page of showing it in running text.
    > As a closing note,
    > the phonetic notation in that dictionary differenciates between antiqua and fraktur a, ä, o, ö and uses them contrastively. Would this mean (considering antiqua to be standard style today) to encode phonetic symbol PHONETIC FRAKTUR A, etc., or would it be correct to use the MATHEMATICAL FRAKTUR range? Doesn't the mathematical range have different character properties?)
    This is an excellent question. Normally, there's nothing that prevents
    the use of the mathematical alphabets for other notational purposes, in
    particular, I would not expect Unicode to encode additional "script", or
    "sans-serif" characters for use in other notations, and by and large I
    see that the phonetic extensions clone Latin letters only for raised,
    rotated, lowered, etc. variations.

    However, there are some problems in this particular case.

    Whenever the entire *text* is set in Fraktur, the use of the
    mathematical Fraktur range is strongly recommended against. So you would
    never use it for the German translations of the French keywords.

    If you wanted to cite the notation in the context of a text set in
    Antiqua, then I would see no principled reason not to use the
    mathematical alphabets - however, they would not be expected to ligate!
    Therefore 'sch' as in the name of the character 'H' would not have the
    mandatory 'ch' ligature.

    Because of that, you would be better off if you could make a global font
    shift to Fraktur and then use special characters for the non-Fraktur
    cases. However, there you run into the problem that there is no
    mathematical alphabet for regular antiqua (roman) style that you could use.

    You could ignore the ligature issue, but that would neither be faithful
    to the original, nor would it have pleased readers trained in Fraktur
    (but there are few enough of those nowadays ;-)). Or you could create a
    special font that supports a ZWJ to enable an optional ligature, where
    needed. At that point, you might as well use private use code points
    because little enough could be expected to be interchangeable...


    This archive was generated by hypermail 2.1.5 : Fri Jan 04 2008 - 13:46:31 CST