bibl as loose model is not loose enough

[GVogeler]

The specification of bibl in the current specification is considered an element for any loosely-structured bibliographic description. For this reason it takes several msDesc and biblFull subelements as possible content. It excludes elements like incipit or decoNote. As incipits can replace titles and the physical features of a bibliographic entity are more than just dimensions I would suggest to extend the biblPart model by these (and maybe further) elements.

[ebeshero]

@GVogeler Are you thinking that we should extend
http://www.tei-c.org/release/doc/tei-p5-doc/en/html/ref-model.biblPart.html to include
http://www.tei-c.org/release/doc/tei-p5-doc/en/html/ref-model.msItemPart.html ? Or does this go too far?

[ebeshero]

@scstanley7 You might have some thoughts on this one for classification purposes...

[GVogeler]

@ebeshero yes, this seems sensible to me.

[jamescummings]

I'm not adverse to that. A more flexible <bibl> element is always desirable for me.

Of course it isn't possible just to have model.biblPart include model.msItemPart since they both have some of the same elements we'd end up with ambiguous content models, I think? We might have to refactor things into intermediate classes. It would seem a bit weird to have both msIdentifier and msItem together there to me, but I'm not that worried.

[lb42]

I otoh am decidely averse to the idea of recklessly adding everything in msItemPart into msBiblPart. . For example, model.msItemPart includes msItem, so msItems can self-nest. Which bibls most definitely cannot. (Cue long boring argument about biblipraphic standards) By all means make a principled selection of members of the former which could also be in the latter, though

[scstanley7]

Besides <incipit> and <decoNote>, do we have a sense of what other elements would need to be? model.respLike is already a member of model.biblPart and it sounds like we want to leave out model.biblLike to avoid nested <bibl>s.

Is there any reason we shouldn't do model.msQuoteLike and a freestanding <decoNote> or does that leave out something crucial to resolving the issue raised in the ticket?

[GVogeler]

I can imagine quoteLike (consider a text on a biblical quotation) and filiation (consider a text identified as the copy/translation of another text) as well.

[GVogeler]

@scstanley7 @lb42 considering references of review articles which use the bibliography of the reviewed item as a core identification element I don't see why the bibl/bibl-nesting allowed already by model.biblPart should create a semantical special risk.

[GVogeler]

I don't want to put brakes into the implementation of already any small extension to bibl.Part but I would at least like to point out that any msDesc could be considered a strict version of a loosely-structured bibliographic description as long as you accept that a frbr:item can be described by bibl. This would lead for instance to include model.phyDescPart into model.biblPart. Just for future considerations on bibl ...

[lb42]

@GVogeler I stand corrected: bibl is a member of biblPart and can thus self-nest. BUT I THINK THAT'S CRAZY! bibliographic entries are supposed to be flat. That's why we have relatedItem (which surely is a better way of handling filiation).

[lb42]

@GVogeler where does it say that a bibl corresponds with a frbr:item? I am pretty sure that it was not intended to originally: that was in fact one of the major reasons for inventing msDesc.

[martindholmes]

It's not crazy to nest bibls. A series contains monograph-level items, and monograph-level items contain articles or poems (analytic level). Real bibliographical objects nest, and <bibl> correctly reflects this, IMHO.

[GVogeler]

@GVogeler where does it say that a bibl corresponds with a frbr:item? I am pretty sure that it was not intended to originally: that was in fact one of the major reasons for inventing msDesc.

@lb42 I compared frbr:item and tei:bibl in a conditional phrase and did not state anything about the verbal form of the guidelines.

I see the conceptualization of bibl is in controverse, so I stick to my initial proposal to add incipit and decoNote to be nested in bibl as they are similar descriptive information as title and extent:
incipit identifies as bibliographic item which does not carry a title, decoNote adds essential physical information like illustrations similar to the use of extent.

[holfordm]

In medieval/manuscript studies there are a very large number of existing bibliographical tools that contain incipits (often many) and often explicits, as a way of exactly identifying the text cited (which may or may not have a title or titles). Stegmueller's Repertories of Biblical texts and Sentence commentaries, the Bibliotheca Hagiographica Latina, etc, are just a few examples. I think incipit and explicit would both be useful additions for anyone who wanted to convert these or similar resources to TEI or create similar resources in TEI.

[scstanley7]

Council agrees to (directly) add <incipit>, <decoNote> and <explicit> to the content model of <bibl> (via model.biblPart).

[jamescummings]

Fixed by 65d514e and tested.