W3C home > Mailing lists > Public > www-html@w3.org > August 2006

Re: sub/sup (was Re: samp, kbd, var)

From: Patrick H. Lauke <redux@splintered.co.uk>
Date: Tue, 22 Aug 2006 19:02:27 +0100
Message-ID: <44EB46B3.6060708@splintered.co.uk>
To: XHTML-Liste <www-html@w3.org>

Jukka K. Korpela wrote:

> The nature of subscripting and superscripting varies from purely 
> presentational to strongly structural.

I would take a slightly different slant and say that the visual 
convention of subscripting and superscripting has traditionally been 
used for different purposes, ranging from the purely presentational to 
the strongly structural.

> It's a mistake to lump all kinds 
> of usage under a simple model of just sub and sup.

Which is why I feel it doesn't make sense to effectively lump these 
usages (or rather, the reasons for the usage of the visual convention of 
sub/sup) into two such inappropriate elements.

> Those are traditional examples for HTML, but this does not make them any 
> better. The former is "semi-structural". The meaning normally remains 
> the same if <sub>2</sub> is replaced by mere 2, but subscripting is 
> still part of the notation system and not just esthetic styling, and in 
> some contexts it makes a semantic difference.

It makes semantic difference, so why use an element to merely replicate 
the visual convention, rather than having an element that conveys the 
semantic quality?

> The latter uses strongly structural markup. To make the point even 
> clearer, compare 2<sup>2</sup> by 22.

Yes, of course they're different, but that doesn't make the use of sup 
any more suitable. What should really be present here is an element that 
denotes "to the power of, in a mathematical vocabulary". As it stands, 
it only indicates that "there's a number 2, and it should visually be 
presented as superscript".

>> Surely this should be marked up more rightly with something like MathML?
> MathML is hopelessly complicated and thoroughly confuses structure and 
> semantics with rendering.

Yes, completely agree there. But again: I wouldn't criticise complexity 
(unless it's unnecessary complexity), as it's needed to mark up complex 
real-world content.

>> <span xml:lang="fr">M<sup>lle</sup> Dupont</span>
> That's of course purely presentational (which doesn't mean irrelevant),

I'd say it *is* irrelevant from a meaning point of view. Again, 
otherwise it should be handled by an element that conveys this 
distinction in meaning, not just "lle needs to be displayed visually as 
superscript" (which, for instance, doesn't translate to non-visual 
renderings such as audio output from a screen reader, for instance).

> just as 1<sup>st</sup> vs. 1st is for English. So should it be 
> eliminated?

If it's purely visual, it should be eliminated. If it's semantic, it 
should be replaced with an element that carries semantic meaning. I 
don't see any difference here from the "b versus strong and i versus em" 
type arguments.

> Do we really want to force people into writing foolish 
> "structural" markup like M<span class="sup">lle</span>?

If people want to have it purely for a visual styling, then yes, they 
should be forced into using that kind of stuff. If they need to use it 
for semantic reasons, there should be an element that conveys the 
semantics, and not the visual convention used to render a variety of 

Patrick H. Lauke
re·dux (adj.): brought back; returned. used postpositively
[latin : re-, re- + dux, leader; see duke.]
www.splintered.co.uk | www.photographia.co.uk
Web Standards Project (WaSP) Accessibility Task Force
Received on Tuesday, 22 August 2006 18:02:45 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:06:14 UTC