[tei-council] the pdf guidelines, a battle not a war

[James Cummings]

Lou's Laptop wrote:
> James Cummings wrote:
>>
>> I don't know if I'm just being overly pedantic (what, who, me?) but 
>> are we being presentationally dishonest in providing a full citation 
>> after an example?
>>
>>   
> I think, yes, you are being a bit pedantic, but it raises quite an 
> important and interesting question as to the status of examples in the 
> Guidelines and elsewhere.

I thought I might be, but instead of (just) mentioning it to Sebastian I thought 
I should post it because I feel that one of the things we should work on at some 
point is the whole framework of examples in the Guidelines and how they 
function.  (Especially with regards to our developing efforts in 
internationalization.)

> As you rightly suggest, the vast majority (possibly even all) of the 
> examples in the Guidelines are not quoted directly from previously 
> tagged works, but instead represent (sometimes several) ways in which 
> parts of some pre-existing work might or should be tagged. 

Yes, the only example I thought might be real is one of the wife of bath ones, 
or maybe one of the msDesc ones, but I had no easy way to know/check.

> It had never 
> occurred to me before now that this was not entirely obvious, but 
> clearly we need to state it, if only to avoid the misconception that you 
> draw attention to.

For some reason it has been entirely obvious to me in reading the web version, 
it is only when seeing the citation under the example that I suddenly felt 
uneasy about what we were claiming.

> For me the important point here is that the examples demonstrate how to 
> encode real texts rather than stuff made up for the purpose. This keeps 
> us honest, and also imposes the need to cite a source, so that the 
> skeptic may go and determine for themselves whether or not the proposed 
> tagging makes sense in context. And it shows that one USP of the 
> Guidelines is their ability to mark up (readings of) real, pre-existing 
> text, with all their complexity, rather than just provide markup for 
> materials born digital.

I would in no way disagree with this.  Yes, the examples should always be real 
texts, for the reasons you give (if not others as well).  In doing so, however, 
we shouldn't necessarily ignore existing TEI ones.  Although it would probably 
be better for an adjunct document/workshop, wiki, or something _like_ TEI by 
Example, I could see a pedagogic purpose in showing how an existing project has 
done something (and their reasoning), and then demonstrating alternative methods 
to accomplish the same thing.

> I don't honestly know whether the community of established TEI practice 
> is yet sufficiently mature or stable enough that it would be helpful to 
> include examples of how *in practice* people use TEI for a wide range of 
> materials. 

It comes down to the problem of people being unwilling to (easily) share their 
dirty linen.  I don't think we can really judge.  Certainly though, this is 
probably true of TEI P5 at the moment though.  Does this mean it would be a bad 
thing to incorporate examples of TEI encoding from external sources in the 
Guidelines?  If one day in the extremely far distant future *all* the examples 
(except where pedagogically useful to show parallel methods) were from real 
existing projects one could cite, would that be a bad thing? (Assuming also that 
it didn't compromise the breadth and scope of the examples, etc.)

> I don't there are any cases in the Guidelines where we cite an existing 
> TEI encoded text. It's kind of a boot strapping problem!

An intriguing one I feel.  I'm happy with Sebastian's solution which uses the 
numbers created in the works cited index.  For some reason that seems entirely 
different compared with having the text there under the example.  I'll be the 
first to admit that this is a completely illogical feeling though. ;-)

Best,
-James

-- 
Dr James Cummings, Oxford Text Archive, University of Oxford
James dot Cummings at oucs dot ox dot ac dot uk