[tei-council] Fwd: TEI TITE question

[Sebastian Rahtz]

On 6 May 2012, at 13:24, Lou Burnard wrote:

> rather than the core though.  But why do we choose these rather minimal 
> indications of rendering? Shouldn't there be some tags for other kinds 
> of visual salience, notably for size of letters, letter spacing etc. ? 

some answers to this come from analysis of large amounts
of texts, surely. ie if there are 999 examples of people doing
<hi rend="bold"> and only 1 of <hi rend="doubleunderline">,
then it suggest we need <b> but not <dub>. 

letterspacing and size are tricky, as they more commonly involve
measurement of the amount of the effect (ok, so bold has the same
problem, but much less so).  But one could certainly _consider_
<larger>, <smaller> and <letterspaced> I suppose. and
<underline>.

> 
> I also worry about the way that visual salience is not a simple matter. 
> If you mark up font change alone you will often completely miss the 
> point -- for example, the word/s in Roman within a paragraph in italic 
> are the ones that should be tagged, not the words around them which 
> happen to be in italic.

Sure. But many people will surely want to use the TEI crudely for recording
font changes, and then release that text for someone else to do
semantic markup? I assume this is where the digilib levels of encoding
come in.

Sebastian