On 4/24/05, John Blumel <johnblumel(a)earthlink.net> wrote:
On Apr 23, 2005, at 10:11pm, Dan Carlson wrote:
For example,
any of the various tags would generally be rendered in
italics: "cite", "var", "dfn" (for example). But if you
always used
the "i" tag, how would a context-less reader know whether the
italicized text was supposed to be a title, a variable, a definition,
or emphasized text? THAT is why the "em" tag is important.
I agree that cite, var, etc. do have semantics different from i/em.
It's only the i/em and b/strong dichotomies that I think are artificial
constructions without real semantic distinction. And it is the very
artificiality of these constructions that has resulted in em and strong
being ignored by most people.
You probably didn't realise it, but you are contradicting yourself
slightly - if you accept that a "cite" tag, or a hypothetical
"foreign" tag, has meaning (even if only a machine would care about
that meaning), then you understand that, at least sometimes, it is
useful to say *why* something is in italics. Well, if <i> just means
"it's in italics, for whatever reason" and <cite> means "it's
in
italics *because it's a citation*", how do you say "it's in italics
*because I want to emphasise it*"? The answer, in HTML, is <em>.
I think the key point is that making a word italic to emphasise an
attribute of it (e.g. its foreignness) is not the same as making it
italic to emphasise that word as against others in the text. I don't
think traditional typography uses italics for foreign phrases because,
say, "tete a tete" is always an important phrase, that you want to
draw attention to, whenever you use it; it's that it's "different",
and you want to *distinguish* it. The <em> tag is intended to
represent the kind of italics that mean "hey look at this, it's more
important than the rest".
The reason no-one uses <em> is, quite simply, because they don't use
the other semantic tags either - if, when you want to use italics, you
think *why*, then there will be times when <em> fits that reason
better than, say <cite>. <i> just allows you not to think about it,
because it covers both cases.
--
Rowan Collins BSc
[IMSoP]