If this needed a distinct encoding, then I'd be inclined to encode the letter-diacritic combinations as atomic characters because they form a single outline, and the interaction of the diacritic with the letter depends on the particular letter. This is similar to encoding of letters with retroflex-hook tails as atomic characters.
But since you've mentioned a historical connection to ogonek should these just be considered glyph variants of ogonek letter variants such as 0105 "ą" and 0119 "ę"? Peter -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Janusz S. Bien Sent: Tuesday, August 31, 2010 7:57 AM To: [email protected] Subject: ,,semi-virgula'' I apologize for possible duplication. I've send this message already yesterday, but it hasn't appeared on the list... Enclosed please find a quotation from Die Altpolnischen Orthographien des 16. Jahrhundrerts Eingeleitet und herausgegeben von Stanisław Urbańczyk unter Mitwirkung von Reinhold Olesch Böhlau Verlag 1983 which mentions a diactritic called there "semi-virgula" and illustrates its use with the letter e. The diacritic has been definitely used also with the letter a, although sometimes it had a slightly different shape as illustrated in the next attachment. It was used also with a letter which I interpret as the latin alpha. Unfortunately I don't have a good scan of the original usage, so I demonstrate it in the last attachment by a contemporary rendering in the dictionary of 16th century Polish. I have two questions. First, is "semivirgula" a good name? Google shows that it often refers to semicolon. "Caudatum" seems to me even more misleading because it should refer simply to ogonek (to the best of my knowledge, the distinction caudatum/ogonek is made only internally by the editors of the dictionary of 16th century Polish). The characters with "semi-virgula" were replaced later by ą and ę and usually are transcribed this way, but sometime the distinction should be preserved, as in the dictionary mentioned above. So the second question concerns of course encoding. Can the diacritic be interpreted as an already exisiting combining character? I didn't find a appropriate one. I will appreciate very much your opinion. Best regards Janusz
