Re: U+0345 COMBINING GREEK YPOGEGRAMMENI not usable in other scripts as "hook below"?

From: John Hudson (
Date: Sat Jun 17 2006 - 14:28:11 CDT

  • Next message: John Hudson: "Re: Glyphs for German quotation marks"

    Alexej Kryukov wrote:

    > On Saturday 17 June 2006 07:01, you wrote:

    >>Well yes and no. Pretty much any Greek typography expert will
    >>probably tell you that the adscript is correct following the relevant
    >>uppercase Greek vowels in the polytonic system. One does sometimes
    >>see subscript iotas in this context, and they are understood by
    >>readers, but they are not correct according to the canons of quality
    >>Greek publishing.

    > This is not exactly true. The fact is that iota subscript below
    > capital vowels was rather uncommon for European typography (outside
    > Greece) of 19th and 20th centuries. As a result, one can see that in
    > 99% of editions of classical (ancient) authors combinations with iota
    > subscript are always capitalized to an uppercase vowel + regular
    > lowercase iota. This practice is described in any manual of Ancient
    > Greek language, and so most Western classicists consider any other
    > practice unusual or even illegal. However, this rule was never
    > considered mandatory in Greece itself: in fact most Greek editions
    > of the same time have subscript iotas even below capitals.
    > For this reason I probably would never use iota subscript
    > with capitals when publishing monuments of Ancient Greek literature
    > (since the standard orthography here is formed by the Western practice)
    > but would consider these forms even preferrable for any other types of
    > texts (i. e. Byzantine texts, liturgical books of Greek Orthodox Church,
    > etc.), since in this case the subscript glyph will better correspond to
    > the canons of quality Greek publishing.

    I am cc'ing this message to my colleague Gerry Leonidas, who is an expert in the field of
    Greek typography and type design. Perhaps he can comment on what you have written, since
    it does not conform to what I understand.

    The use of adscript iota with uppercase letters certainly precedes the 19th and 20th
    century. I have seen examples from the 16th century. Yes, these were in editions published
    outside of Greece, but as Konstantine Staikos has documented, emigre Greeks were heavily
    involved in that publishing.

    > Note that in most Unicode fonts prepared by Greek designers all
    > combinations with subscript/adscript are implemented by this way (see
    > for example Particularly I prefer
    > such fonts (and follow this practice in my own fonts) for the following
    > reasons:
    > -- if one absolutely dislikes iota subscript below capitals, (s)he
    > always can type a capital letter and small iota separately, while the
    > opposite is hardly possible;

    If users do have a preference, it is also possible to handle this entirely at the glyph
    display level. There's no reason why a font couldn't support both usages, and there are
    several mechanisms by which the user could select which he prefers. This has the advantage
    of not requiring changes to the text encoding that, of course, affect the ability to
    search for and compare strings.

    > -- generally speaking, all digraphs encoded as a single character
    > are very inconvenient for typesetting, because they are not spaced out
    > when we change letterspacing for the piece of text which contains them.
    > That's why IMHO the ideal solution for this situation might look as
    > follows: the font itself contains only combinations with iota subscript
    > encoded to the appropriate slots, which, however, are replaced with
    > capital vowels followed by lowercase iota by applying an OpenType tag
    > (which may be enabled by default). Unfortunately, I don't know such
    > an OpenType tag which would allow replacing a single glyph with multiple
    > glyphs in Wester scripts...

    The <ccmp> feature can be used to decompose individual glyphs into multiple glyphs.
    Decomposing digraphs is one of the uses for it. However, note that not all digraphs should
    necessarily be spaced out when overall spacing is increased: the Dutch IJ digraph is often
    seen spaced like e.g. R IJ N -- even in inscriptions carved in stone, where font or
    application limitations certainly were not to blame.

    John Hudson

    Tiro Typeworks
    Vancouver, BC
    I am not yet so lost in lexicography, as to forget
    that words are the daughters of earth, and that things
    are the sons of heaven.  - Samuel Johnson

    This archive was generated by hypermail 2.1.5 : Sat Jun 17 2006 - 14:31:55 CDT