Re: Uppercase variant of U+00DF LATIN SMALL LETTER SHARP S ("German sharp s", "ß")

From: Otto Stolz (
Date: Thu Feb 17 2005 - 08:08:51 CST

  • Next message: Otto Stolz: "Re: 03F3 j Greek Letter yot"


    let me remind everybody participating in this thread that “ß” cannot
    appear at the start of any German word; hence there is no need for
    an uppercase variant in mixed-case spelling. The whole problem is
    a typografical one, confined to typesetting in all capitals (German:
    “Versalien”) and small caps (“Kapitälchen”).

    Am Mittwoch, 16. Februar 2005 um 15:10 schrieb Werner LEMBERG:
    > Using ‘ß’ within uppercase words is plain wrong — the ‘Duden’
    > explicitly says (even for the new German orthography) that it has to
    > be replaced with ‘SS’. In seldom cases (‘Masse’ vs. ‘Maße’) it should
    > be replaced with ‘SZ’ to avoid ambiguity.

    The current official rule: <>.

    Note that “ss” may be used where “ß”is not available, and generally in Swiss
    spelling (cf. Explanation E2). Note also that in all caps (or small caps)“ß”
    is replaced with “SS”, and never “SZ” (E3).

    Karl Pentzlin schrieb:
    > This is a pure normative point of view.

    This is not only the normative, but also the historical view.
    “ß” has developped from a ſ-ʒ ligature (U+017f + U+0292; the latter was
    later used to denote the [ts] sound which is now written as “z”, thence
    the German name “Eszet” for the “ß”).

    In all-uppercase, or small caps, writing, there was simply no need to
    ligate “S” and “Ʒ” (or, later, “S” and “S”), so “SƷ” (or “S”, respectively)
    was the normal spelling.

    Until 1941, the “ʒ” glyph was used in Fraktur to denote the “z”
    character (the Antiqua type had the “z” glyph for that character);
    so everybody was used to equate Fraktur “ʒ” and Antiqua “z”; hence,
    “SZ” (in Antiqua) was often used as the upper-case equivalent for
    lower-case “ß” (there is no such thing as all-uppercase Fraktur,
    exept in “der HErr”, or occasionally “der HERR”, meaning “our Lord”).

    After WW II, “SS” prevailed as the upper-case equivalent for lower-
    case “ß”, and is now the only official one (cf. supra).

    Karl Pentzlin:
    > The living language has developed another way and proves the capital
    > forms “SS” and “SZ” as plain wrong.
    > It is very hard to find “ß” capitalized as “SS”, except in schoolbooks.

    This is a blatant distortion of the facts.

    Actually, most people use “SS” when uppercasing “ß”; many,
    including most Swiss, even use “ss” in place of official “ß”.

    Karl Pentzlin:
    > The “ß” in capitalized German text is fact.

    Fact indeed, but not the mainstream — at least in the area where
    I am living.

    > I never have seen the “SZ” variant except in the Duden itself or on
    > very old technical drawings.

    So did I (mainly in old maps). But “SS” is common, these days.

    > E.g. in my car papers from 2004, issued by a communal authority,
    > my street is printed “KANZLEISTRAßE”.

    This sort of spelling can indeed be found in car papers, idendity cards,
    and passports, and it looks indeed ugly. (I reckon, our authorities
    lack good taste in typography, as also manifested in the glyphs chosen
    for the newest series of car plates; luckily, the old design is still

    This is the reason why I would not object to having a variant selector
    for an “ß” glyph to fit into an all upper-case, or small caps, word.
    But that usage does not warrant the creation of a “CAPITAL SHARP-S”,
    invalidating/obsoleting the main-stream spelling.

    Sill, I confess: a cold shudder runs down my back, every time I see
    an “ß” in all-uppercase context. A wider “ß” glyph will not change
    this, as I learned when viewing Andreas Stötzner’s evidence.

    > The Deutsche Post (German Post) requires forms (for new address
    > after moving) to be filled in by all capital letters, but explicitly
    > to write “ß” not “SS”.

    This is simply a technical measure to facilitate OCR; in the address,
    or phone, books the address will be written in mixed-case, as usual.

    > If a “ß” occurs as such in capitalized text, this can be interpreted
    > in two ways:
    > a.) “ß” is a case-invariant letter, like Cyrillic U+04C0.
    > b.) There is a capital letter “ß”, for which the glyph of
    > the lowercase “ß” may be used.

    More possible interpretations:

       c.) There are technical limitations forcing to use all-capitals,
           where mixed-case was really intended, and there is simply no
           capital “ß”, so the only (viz. lower-case) “ß” was used.

       d.) The writer is ignorant, or does not care about orthography,
           at large.

    In my opinion, c.) applies in the overwhelming majority of cases,
    with a considerable fraction to be attributed to d.)

    > Andreas Stötzner's proposal at
    > shows evidence that a capital sharp s is in fact used in some “good
    > typography”, but that in “every day use” the (hitherto) lower case
    > letter is used as upper case also.

    Let me remind everybody that Andreas Stötzner has collected evidence
    only for his case, viz. “ß” in all-capitals context,but not for the
    “SS” spelling. Yet, the latter exists, and — according to my perception —

    > To handle “ß” in capitalized text correctly, theremay be another way
    > than encoding a new character “LATIN CAPITAL LETTER SHARP S”:
    > 1.) Change the properties of U+00DF “ß” to “caseless” like U+04C0.

    This property should not be changed, as “ß” is definitely lower-case
    (cf. discussion above), and is subject to an established, official
    uppercasing rule. Alos, that rule must not be changed in the Unicode
    data files.

    > 2.) Add a note to the description of U+00FD in the standard that the
    > “SMALL” in the name is historical.

    Not yet, I reckon ;-)

    By all means, do not delete the current remark that the uppercase is “SS”.

    And, by the way, correct the historical mark about its origin, cf.
    <>. (The ſ-s ligature (U+017f
    + U+0073) was originally *not* used for German, so I did not mention
    it, above. At home, I have a facsimile of a baroque print with a bi-lingual
    explanation: the french part has ſs, the German part however ſʒ.)

    > 3.) Define a standardized variant sequence U+00DF U+FE00 “LATINSMALL
    > LETTER SHARP S capital form”.

    If it is felt that a particular encoding for “ß” in all-capital context
    is needed, at all, then I deem this the way to go. However, I am still
    not convinced that a particular encoding is needed for those cases.

    Best wishes,
       Otto Stolz

    This archive was generated by hypermail 2.1.5 : Thu Feb 17 2005 - 08:10:07 CST