Posted by Jukka K. Korpela on 05/09/07 19:35

Scripsit Oliver Wong:

> In an informal setting (e.g. a blog), I don't mind
> small changes like case or spelling corrections to go unbracketed,
> but in a formal setting (e.g. an online academic paper), all changes
> should be denoted.

That's a matter of convention, and indicating changes of letter case belongs
to a particular formal or "legalese" quotation style in English. It's
sometimes seen in other languages as well, probably due to influence of the
English practice, but it looks rather ridiculous and odd outside its
established scope of use. The Chicago Manual of Style presents case changes
(for contextual reasons) as permissible (see clause 11.8) but mentions
another, more rigorous usage "appropriate to legal writing and textual
commentary".

> I never ran into the wordwrapping problem
> mentioned above, because when I changed the case, I just bracked the
> whole word, e.g. "[This]"

That's possible but slightly misleading, since it suggests that the entire
word has been changed or added.

Regarding the word wrapping problem, it exists in some browsers, including
IE, and the practical cures are the CSS property shite-space that has been
mentioned and the nonstandard but widely supported <nobr> markup. Using
Unicode control characters would be the theoretically correct way but does
not work well.

Unicode line breaking rules really permit a break between "]" and a letter,
even if no space intervenes. But what matters is really how browsers behave.
Their behavior has just tangential connection with Unicode rules. Note that
HTML specifications do _not_ require Unicode conformance.

--
Jukka K. Korpela ("Yucca")
http://www.cs.tut.fi/~jkorpela/

[Back to original message]