Re: About the Kikaku script for Mende, and an existing font for it

From: Philippe Verdy (
Date: Mon Mar 15 2004 - 19:31:08 EST

  • Next message: Mark E. Shoulson: "Re: in the NEW YORK TIMES today, report of a USA patent for a met hod to make the Arabic language easier to read/write/typeset"

    (Still not working, retrying to send this message with different options, there
    may be a problem in the Unicode mailing list server)

    For your information, here is a document better documenting the beautiful glyphs
    used in the "JG Mende" True Type font by Jason Glavy, and comments from the
    author of this font.

    The two first files below include the reference on the net to the location of
    the beautiful TrueType font "JGMende" on the web by its author, Jason Glavy:

    One of its question is directly referencing discussions for the standardization
    of the Ki-ka-ku syllabary within Unicode (see hist first message quoted below).

    Also the document was created with the JGMende font for which I created a code
    point mapping table as it currently uses PUA assignments: (Excel format), or (first worksheet) (second worksheet)
    The second worksheet groups character variants and order them according to Dr.
    Dalby list by their phonetic values.

    I have located the attachment related by Jason Glavy (who authorized me to quote
    him below in this message) at this URL:
    And another attachment which is a scan of the Dr. Dalby '67 list of Kikaku
    glyphs with their Latin transcription:
        (large: 1683 KB)

    May be there exists some orthographic rule that specifies which syllable form
    should be used, or if this choice is made because of esthaetic reasons, or
    practical reasons with hand-drawing... So most of the 195 syllables have a
    secondary variant, and sometimes a third one.
    Some distinct letters also have equivalent Latin transcription in Dr. Dalby
    list, but this may be superficial and a caveat of the Latin transcript (Jason is
    not an expert, and other people that have the whole book from Dr. Jason '67 may
    want to verify this).

    Jason sent me a PDF quoting a reference alphabet for the Latin script and a list
    of glyphs composed with his font with corresponding phonetic values. You may see
    it in:

    Comments welcome. Apparently Jason is quite busy and would like that others
    discuss these issues (Jason just made a font to help someone requesting it to
    him, but has apparently no time to discuss it with Unicode). If you want direct
    contacts, you may try contacting him to get other contacts he has to discuss
    this script. You'll see a part of his email I quoted here below.

    ----- Original Message -----
    From: "Jason Glavy" <jglavy (at) ac (dot) cyberhome (dot) ne (dot) jp>
    To: "Philippe Verdy" <verdy_p (at) wanadoo (dot) fr>
    Sent: Sunday, March 14, 2004 2:25 PM
    Subject: RE: Comment about your beautiful JG Mendé font

    > Hello Philippe,
    > Please see the attached PDF file. My primary information has come from
    > Dalby '67. In the attached PDF file, I have indicated the Modern
    > Orthography as well as the Kikaku script, along with IPA values. Whiile
    > there are only around 195 characters in the script, of those characters
    > there are generally one or more different variants for each glyph. I have
    > been in contact with Dr. Konrad Tutscherer who is apparently making a study
    > on the Mende and other West African scripts. Because he is so busy, he
    > hasn't been able to do an in depth analysis of my font.
    > Anyway here is the phonetic values of the Kikaku syllabary. If/when you
    > talk with the pandits of Unicode, kindly ask them how they will resolve
    > which glyph is "standard" and which glyph is "alternate", since I haven't
    > got a clue.
    > Best regards,
    > Jason Glavy,
    > Yokohama, Japan

    ----- Original Message -----
    From: "Jason Glavy" <>
    To: "Philippe Verdy" <>
    Sent: Monday, March 15, 2004 6:09 AM
    Subject: RE: Comment about the JG Mendé font (for the Ki-ka-ku syllabary script)

    > Hello Philippe,
    > I am attaching a PDF scan of Dr. Dalby's Mende list. It is constructed much
    > like your list, with phonetic representations of the character values. It
    > is numbered according to unique character, with variants in parentheses.
    > You will immediately see that some sounds (w,mb, etc.) have more than one
    > unique character and even variants. I cannot and do not attempt to explain
    > this. In personal communications with Dr. Tutscherer, he said that Dr.
    > Dalby's study was a good "start" but incomplete. I don't know how
    > incomplete, so if/when Dr. Tutscherer returns from West Africa, I hope he
    > will contact me and then I can pass that information on to you.
    >> Before I send the following email to the Unicode List, do you authorize me
    >> to quote you?
    > REPLY: Sure, just keep in mind that I am not an expert, and based my fonts
    > on the study conducted by Dr. Dalby, as faithfully as possible. If anyone
    > requests me to add/modify/delete characters, providing they send me the
    > proper data, I would be more than happy to modify my font. Above all, I
    > strive for accuracy and authenticity, so if my font is inaccurate, I hope
    > someone will tell me where so that I may fix it.
    >> Also, I have studied your interesting document, but after counting
    >> scrupulously the various variant syllables shown for the "same"
    >> phoneme, I found only 184 of them, and not 195.
    > REPLY: (see the attached scan)
    >> What is wrong in my worksheet? I see sometimes more that 2 variants for
    >> the same syllable. For example the 4 variants listed for "wa", "wè", "mbu",
    >> "mbè", "mboe", "la": are they variants or 2 characters with a secondary
    >> variant? (this case would mean we have then 190 Kikaku letters, but still
    >> not 195).
    > REPLY: (see the attached scan)
    >> May be we should add the long vowels (how are they represented in Kikaku
    >> outside of /le:/ shown by your font at E0098 and its variant at E0099? By
    >> an additional isolated vowel between E0127 and E0133 in your font?)
    > REPLY: Since I do not know enough about this script, it might be better to
    > just follow Dr. Dalby's grid....that is until Dr.Tutcherer publishes more
    > data
    >> Or should I add the 4 combining tone marks diacritics (accute for high tone
    >> 55, grave for low tone 11, hacek for raising tone 24, circumflex for falling
    >> tone 42, which are listed for the Latin script but apparently not listed in
    >> the Kikaku script) Unlike the Latin, there is no graphical tone
    >> representation in the Kikaku script available in your font (which shamely
    >> does not include also all the Latin characters used by Mende, but only
    >> basic ASCII).
    >> Also I note that the positions E0149 and E014F both show the same glyph
    >> (similar to a big ">") and the same interpretation as NING (i.e. /ni/). Is it
    >> a duplicate in your font?
    > REPLY: It is a error

    This archive was generated by hypermail 2.1.5 : Mon Mar 15 2004 - 20:10:32 EST