Re: Case-folding dotted i

Note that the "soft" dot over lowercase i and j is not mandatory even
if it is usually written now. Initially it did not exist at all, but
it appeared in the Middle Age to disambiguate the cursive form of the
script (between series of i and u). The same could have happened to
disambiguate series of m and n, but it was less ambiguous).

It appeared above j by analogy with the letter i (from which the
letter j was itself derived) and was kept as it could potentially
disambiguate some cursive form of j with p (but the solution was, in
cursive scripts to add a loop to its leg to join in back to the base
line, so the soft dot above j is less justified).

If you look at some mediaval manuscripts,you'll have difficulties to
read words like "immunity" without the dots ! The letters i, u, m, n,
all look very similar without the dot. If this text is in German, the
dots may also add their own confusion with the diearesis (is it "ii"
or "ü" ?)

In non-cursive forms, the normal "i" and "j" may have their dots
suppressed, EXCEPT if the text is written in a Trukic language,
because their dots are "hard", not "soft". For this reason Turkic
texts *should* encode the hard-dotted lower case i as i+dot above, and
not just as i alone. But when the language used in the text is clear,
the extra encoding of the explicit "hard" dot above is almost always
forgotten and for legacy reasons, most Turkic texts do not use this
extra dot above, but it does not mean that its presence is incorrect
(it will be needed in multilingual documents, or when using some
Medieval-style fonts that do NOT display any dot above U+0069 and
U+006A and that require the explicit U+0307 to render the hard dot
needed for Turkish).

2013/1/23 Philippe Verdy <verdy_p_at_wanadoo.fr>:
> 2013/1/23 fantasai <fantasai.lists_at_inkedblade.net>:
>> 0130; F; 0069 0307; # LATIN CAPITAL LETTER I WITH DOT ABOVE
>>
>> This says that capital dotted i case-folds to latin lowercase i (with a dot)
>> plus a combining dot. Why? Does this not imply two dots??
>
> No, because the lowercase i U+0069 (also the lowercase j U+006A) just
> has a "soft" dot which disappears if there is any extra diacritic
> above the letter, including if this is an explicit dot above.
Received on Wed Jan 23 2013 - 16:48:25 CST