On Aug 15, 7:54, Marc Wilhelm Kuester wrote: > (sharp s, U+00DF) was originally a ligature Yes. > between the long s (U+017F) and the normal s. Probably not, though Jan Tschichold indeed says so; I think I saw this in his "Meisterbuch der Schrift", ISBN 3-473-61100-X. In her book "Gotische Schriften in deutscher Sprache", ISBN 3-88226-281-8, Karin Schneider says (p. 189) that in the 2nd half of the 14th century in south-western Germany "s" (U+0073") and "z" (U+007A) began to be used interchangeably (my comment: in contrast to current use, "z" was then used for a particular "s"-like sound; current dialects of this area have only one s-like sound, viz. the voiceless "S", in contrast to the two different s-like sounds used in other German dialects, including standard German). In the sequel, K. S. says that the "ß" ligature evolved from "sz" (U+017F U007A) spelling in the beginning of the 14th century in that very area. In , an FAQ is monthly posted that contains a quote from Jacob Grimm explaining the spelling of the various S-like sounds, and the emergence of the "ß" ligature in even more detail (I haven't kept a copy, so you will have to wait until the FAQ will be posted again). Albert Kapr's wonderful book "Fraktur" (ISBN 3-87439-260-0) has pictures of several texts, and specimens of many gothic fonts. Most of them show rather clearly that "ß" is formed as a long-s+z ligature, as can be seen by comparison with the respective t-z ligatures. My own copy of a "Theuerdank" facsimile (by Maximilian I, printed at Augsburg, in 1519) shows s-z spellings (e. g. "ausz") where the ascender of the long "s" extends to the right over the "z". It has also "ss" spellings (e. g. "grosse") where both "s" are long, but I could not find any long-s + round-s spelling. This is in contrast to English and French tradition, as has been reported in this forum. At home, I have a bilingual facsimile (I think from the 17th century, but do not nail me down on this date) which has long-s + round-s spellings (latin running hand), in its French part, but clearly s-z ligatures in fraktur (gothic print), in its German part. Many contemporary roman fonts have an "ß" glyph that can be seen as a long-s+ round-s ligature, and most (but indeed not all) fraktur (gothic print) fonts have a clear long-s+z ligature. > In gothic script (= Fraktur) the 'round' s (U+0073) was used at the end of > syllables, long s within. Right; but I do not know how old this rule is. > Double s are frequent in German, and the collision between U+017F and > U+0073 resulted in the ligature "sharp s". This is not probable, as double-s is usually written as twice U+017F, and was originally written as twice U+007A, according to Karin Schneider, and later (13th century) sometimes as either U+017F U+007A, or U+007A U+017F. > This holds true only for lowercase letters as there has never been an > uppercase form of the long s. I agree. > The Unicode-Standard is quite correct in pointing out that the *only* > correct uppercase form of sharp s is SS. Almost so. As I have pointed out, recently, the old (still valid) spelling rules allow "SZ" instead, in very limited cases. > With the de facto disappearance of the gothic script from day to day use > in the thirties, In the thirties, gothic script was the prevalent script, in Germany. It has been abandonded by a decree, only in 1941. (You can read the original text, in Kapr's book cited above.) > the long s and with it the perception of sharp s as a ligature has all but > vanished. I guess, Marc Wilhelm Kuester wanted to say "has almost vanished". > While it is essential for the reproduction of texts which were originally > printed in gothic script Not only for gothic script. I have a facsimile of the 1926 (or was it 1916 - oh, I am not good at memorizing numbers) edition of a renowned German songbook, "Der Zupfgeigenhansel", that uses long-s (U+017F) in a roman font ("Antiqua", in German), complying with the normal German spelling rules (which later were used only for Fraktur). > and for scholarly editions of e.g. medieval works, > the long s is no longer used and would, if employed today for modern texts, > be considered a spelling error. Currently, there is a revival -- though a rather limited one -- of the Fraktur (gothic), for decorative purposes. Our banknotes (bills, for Americans) have the word "Banknote" printed in gothic on them (until we will be blessed with Euro banknotes :-). My favourite beer has the word "Hefeweizen" in gothic on its label. Other examples are inn-signs, and advertisments. A couple of books have their book, or section, titles in gothic. In these examples, round-s (U+0073) in places of long-s (U+017F) is considered as a spelling error by the knowledgable, though many contemporary printers, or sign-painters, do not know, or do not care of, the spelling rules for Fraktur, any more, the most often seen example being "Gasthaus" with twice U+0073 rather than the correct "U+017F" in the middle (and U+0073 only at the end). > Therefore I would opt against a > compatibility decompostion of U+00DF in U+017F and U+0073. It would > contradict current practice. I agree. However, current practice dictates a decomposition into double-s (twice U+0073) for comparison, and sorting, purposes. Swiss spelling uses double-s almost invariably for German sharp-s; the new spelling rules, as negotiated between German, Austrian, Swiss, and some other interested authorities explicitely state, in § 25 E2: : If the letter "ß" is not available, you write "ss". : In Switzerland, you may always write "ss". : Example: Straße - Strasse Cf. for the original text. > As an aside: For sorting sharp s is considered to be equivalent to "ss". Yes, but only in level 1, as has already been discussed in this forum. > For the sorting of the umlauts a,o,u we have two contradiction norms. For > phone books and the like they are treated as a,o,u+diaeresis, in libraries > and as a rule in scholarly contexts as "ae", "oe", "ue" Actually, it is the other way! (according to DIN 5007) Best wishes, Otto Stolz