Re: Rare extinct latin letters

From: Philippe Verdy (verdy_p@wanadoo.fr)
Date: Fri May 30 2003 - 10:42:53 EDT

  • Next message: Edward C. D. Hopkins: "Re: Aramaic Roadmap (was: Persian or Farsi?)"

    From: <Peter_Constable@sil.org>
    > Patrick Andries on 05/29/2003 06:15:10 PM:
    >
    > > Could letters like l molle
    > (http://pages.infinit.net/hapax/abcmeigret.jpg
    > > ) or long-tailed A (between O and P in Baf's alphabet http://pages.
    > > infinit.net/hapax/abcbaif.jpg), letters which I believe cannot be
    > > composed from other existing Unicode characters, be considered for
    > > Unicode encoding
    >
    > If there is a user need, then probably yes.

    I'd like to see some real historic publication that is not an attempt to reform the French orthograph, using an invented "new" alphabet only used by 1 author.

    Such text is probably interesting to study as it gives hints on how French was *spoken* when it was written (i.e. interesting for phonetic studies), but I doubt it has a real language value until there is some real usage of these modified alphabets created only as a proposal for a future reform of the orthograph that was never applied.

    Both samples demonstrate that the usage of diacritics does not match the strict categories applied in Unicode (for example the Unicode distinction between a cedilla and an ogonek is not clear in those texts, where the difference really seems to be considered as a form variant for the same diacritic).

    Also the letter forms are not quite clear, because the metal fonts use characters that were apparently manually manufactured individually. So the glyphs are near but not enough distinctable from these small scanned images. This would require a more complete analysis of the text, if such text exists.

    For now, these appear to be simple and normal form variants from other characters already encoded in Unicode, provided that a convention is applied to choose which Unicode diacritic best fits the displayed characters.

    For example, my reading of the A with long trail, or G with a hook, or E with long trail makes me think that these are ligatures created from a base letter and a following U. The already encoded "attached hook above" diacritic could represent correctly these ligatures, considered in those texts as individual letters in the tentative "reformed" ortograph, based on phonetic rules rather than strict historical radicals.

    The form of the L moll or N moll letters look like if it was a tilde diacritic above N OR ENG (and the same could be used above L, even if it does not strictly look like the same composed glyph, as a historic form variant of the LATIN LETTER L WITH TILDE, currently encoded in Unicode as a pair of abstract characters).

    If this is not enough, may be we could create only a new diacritic for the long leg attached on right, or the moll sign (the rotated and mirrored J sign, which may also be a sort of curl) normally detached above the base small letter and ligated on the right for the L letter or capital letters.

    It's hard to decide from these two images.



    This archive was generated by hypermail 2.1.5 : Fri May 30 2003 - 11:14:15 EDT