On Wed, 31 Aug 2011, Martin Holmes wrote:

> We've had some discussion recently about the use of <label> for purposes not 
> associated with lists, a usage which seems to be hinted at by some parts of 
> the Guidelines, and by its general availability in a variety of locations, 
> but which is not explicitly endorsed. Examples already in the Guidelines show 
> it used in <application>, <etym>, and <postscript>, and I have another 
> situation which I think is suitable for the use of <label>.
> In many of our 17th- and 18th-century French texts, topic headings are often 
> shown in the margin of the text, as if they were marginal annotations. These 
> appear occasionally where a new paragraph brings a change of topic, but most 
> often right in the middle of paragraphs, to signal the mention of something 
> significant. Up to now, I've been using <argument> for these little headings, 
> but that's rather unsatisfactory because its semantics aren't quite right, it 
> can't appear inside paragraphs or other content elements without modifying 
> the schema, and it requires an additional <p> or <ab> inside it before you 
> can add text.
> <label> seems to me to be exactly what these little marginal items are. 
> They're not <note>s -- marginal notes also exist, and they're different. So 
> I'm proposing to use <label> for them instead of <argument>.
> Does anyone see any objection to this, or have any other suggestions?\

Hi Martin,

   Such marginalia, serving to 'track' the text and provide a running,
visually arresting guide to its progress, are almost the rule rather
than the exception in the books of the 16th and 17th centuries, at least
judging from the ones that we meet.  Some are a little more note-like, 
some more heading-like, and some more milestone-like; some, when gathered 
together, form a continuous argument, most do not. I've never been sure 
how best to treat them, or whether a single method suits them all
equally. We certainly have no qualms tagging them as <label>s.

-- Most of them, I'm afraid, end up as <note>s, mostly because it
    has proven very difficult to supply any reliable guidance to our
    keyers on how to distinguish them from notes, and it is too
    time-consuming for our editors to pick out the more label-like
    notes, re-tag them, and likely also move them to a more appropriate
    spot in the text. <note> with a distinctive @type might be an
    acceptable answer.

-- a few tend to form a continuous argument, are meant to be
    read sequentially down the page, and turn into nonsense
    if left dismembered in separate note tags :

        <note>The author here speaks about</note>
        * * *
        <note>the dreams he has had</note>
        * * *
        <note>lately, and dismisses</note>
        * * *
        <note>them as the product of bad</note>
        * * *

      : these we gather up as convenient into a single <argument>
      tag at the head of the appropriate <div>.

--  if they can be interpreted as marking definable <div>s, we
     move them into <head> tags and place <div> boundaries at those

--  if they are repetitive and represent some kind of sequence, we
     turn them into <milestone>s. (We take a broad view of milestones),

     <p><milestone unit="reason" n="1"> asldkjfla;k</p>
     <p><milestone unit="reason" n="2"> asdlfkajsd;l
     <milestone unit="reason" n="3"> as;dlfkjasd;l</p>

--  most of the remainder are tagged as <label>s, as you suggest.

--  I suppose <seg> is an option.

Paul Schaffner | [log in to unmask] |
316-C Hatcher Library N, Univ. of Michigan, Ann Arbor MI 48109-1190