[whatwg] the cite element

Andrew W. Hagen contact2009 at awhlink.com
Wed Jun 3 20:54:27 PDT 2009


The cite element should be slightly changed. Under this proposal, the 
cite element should be used only for titles of works, but may be used 
for other things that web authors may wish to cite. This conforms with 
how the cite element is used in practice.

In the current HTML 5 specification, the cite element can only represent 
a title of a work. This has several negative implications. First, it 
goes against what the word "cite" means. The common English usage of the 
word "cite" includes making reference to non-titular authorities. For 
example, a writer may cite Aristotle. See 
http://www.merriam-webster.com/dictionary/cite

Furthermore, the current restriction makes the cite element useless for 
works which do not have a title. See a list of such works at 
http://en.wikipedia.org/wiki/Untitled

Trying to enforce a "titles-only" rule for the cite element is 
impossible. The best that can be done is for small bands of advocates to 
ringingly criticize any web author who breaks the rule. That is herding 
cats. It is not as if browsers will refuse to render <cite 
style="font-style: normal">Lincoln</cite> or that validators can 
distinguish that from Gore Vidal's <cite>Lincoln</cite> (a historical 
novel). The restrictive rule cannot be enforced.

Finally, HTML 5 has a broad definition for some elements, such as kbd. 
The kbd element can represent any form of user input, even if it is not 
made with a keyboard. In current-work, one example is given of 
<kbd><kbd>Shift</kbd>+<kbd>F3</kbd></kbd> for Shift+F3, even though in 
that keyboard chord, the user would not actually input the "+" character 
on the keyboard. It is so broadly defined that 
<kbd>Shift</kbd>+<kbd>F3</kbd> would also be valid. Some elements, like 
kbd, are very broad.

Logical consistency cannot be perfectly maintained when specifying the 
next version of HTML, but it should be a goal, and we ought to regret a 
logical inconsistency between the cite element and elements like kbd. 
One is narrow. The other is broad. Broadening the definition of cite 
will not cause harm. It would only allow web authors to fully embrace 
the cite element.

This solution is workable. The cite element's default style is italics 
in display mode, and this proposal would not change that. If a web 
author writes: <cite>Aristotle</cite>, the web author can live with it 
or re-style the cite element as desired.

To conclude, slightly broadening the cite element would improve HTML.

Andrew Hagen
contact2009 at awhlink.com



More information about the whatwg mailing list