It would be useful to include in the EpiDoc Guidelines some example of structured ("genetic") bibliography in the div[bibliography] section. (As discussed on Markup in November 2014.)
I went through the markup list archive of Novembre 2014 when that bibliography topic was discussed.
It seems that "genetic" relations between bibliographic references can be expressed in 2 ways:
1. by classifying the references in categories ('primary', 'secondary', 'derived', principalEdition', 'translations', etc.)
2. by indicating, at the bibliographic item level (<bibl>), the source it is based on, wether an autopsy of the text bearing object or another edition.
In case 1. the genetic relation are implicitely derived from the semantics of the categories and does not try to establish precise relation between the editions, but it is quicker to encode and seems quite a pertinent level of simplicity to cope with that kind of information.
With case 2. it is possible to be very precise and get precisely interlinked bibliographic references, but needs identifiers and pointers, and assume a very well informed editor.
Some examples attribute values for case 1.
Could be a div[@type='bibliography'][@subtype] or a div[@type='bibliography']/listBibl[@type]:
secondary, edition, citations, corrections, illustrations, otherPublications, principalEdition, translations, ancientEdition, modernEdition (values collected by Tom Elliott)
Some examples for case 2
These examples are based on the editorial convention where editions based on an autopsy of the object are presented first. If an edition was based on another edition, it is then presented inside parenthesis, just after the reference it is derived from.
For example, this bibliographic statement (provided by Eleonora Santin): Masson 1979, p. 59, nº 1 and p. 60 fig. 1 (Bile 1988, nº 57, p. 56)
Could be expressed with mixed content:
<p><bibl><ptrtarget="#Masson1979"/><citedRange>p. 59, nº 1 and p. 60, fig. 1</citedRange></bibl>
(<biblsource="#Masson-1979"><ptrtarget="#Bile1988"/><citedRange>nº 57, p. 56</citedRange></bibl>).
</p>
or (with explicit mention of the autoptic nature of the edition):
<p><bibltype="autoptic"><ptrtarget="#Masson1979"/><citedRange>p. 59, nº 1 and p. 60, fig. 1</citedRange></bibl>
(<bibltype="derived"source="#Masson1979"><ptrtarget="#Bile1988"/><citedRange>nº 57, p. 56</citedRange></bibl>).
</p>
Remark:
The examples here suppose a master bibliography where the entries identified by 'Masson1979' and 'Bile1988'. The genetic relations could be expressed in an even more precise way:
<p><biblxml:id='bibl01'type="autoptic"><ptrtarget="#Masson1979"/><citedRange>p. 59, nº 1 and p. 60, fig. 1</citedRange></bibl>
(<biblxml:id='bibl02'type="derived"source="#bibl01"><ptrtarget="#Bile1988"/><citedRange>nº 57, p. 56</citedRange></bibl>).
</p>
The type values proposed in the discussion were @type "FS" (from Stone) and @type "FOS" (from other source) with @prev to indicate the genetic relation.
I would suggest using @source instead of theses two @tye values, because it can be used for every kind of external sources (printed editions but also squeezes, photographs, etc. as long as the have an identifier one can point to).
The absence of a @source attribute would mean just "not derived".
If there is a necessity of being more precise, other @type category could be used like 'autoptic', or 'derived' or anything else for the editor.
The same references could be encoded in a structured way I guess, letting the XSLT doing the parenthesis inclosure. My bet though is that editors would prefer the mixed content approach because it gives them more flexibility to add commentaries on the editions within the parenthesis which would be difficult to handle with a structured bibliography, especially if there are several reference inside the parenthesis, and the scope of the commentary is more than one ref.
The question still remains wether it would be useful to recommend values for categorisation of types of references. As I think this is closely linked to the editorial specifications of the collection the inscription is published in, I would try to harmonize too much things there, but I'd like more comments from others on that point of view.
Last edit: Emmanuelle Morlock 2017-12-03
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
@ellimylonas and @sdigiulio discussed this and although we think that it is unlikely many editors will want to characterize the authoritativeness and source of the editions in the bib., if they do @gethia has proposed using@source on<bibl> with a recommended set of values as follows: please suggest better terms.
1. "photograph"
2. "autopsy" (edition prepared from viewing original object)
3. "squeeze" (all forms of physical reproduction, including estampage and rubbings)
4. "facsimile" (microfilm, engraving, drawing)
5. 3D scan (this could be a "squeeze")
This can be laid out in a section on the bibliography page, and would belong in a master bibliography section.
However, we think it's difficult to classify a reference as a primary source, secondary source, generic information, in a master bibliography because the same publication could be a primary source for one object and a secondary reference for another that is just mentioned or cited. So this classification, which be handled with a @typeattribute seems to belong in the bibliography for an individual object. Not a master bibliography. Or can be handled, as in the example already guidelines, by having separate <bibliography> elements.
We recommend @type on the bibliographic references in the inscription with the following suggested values (from above): secondary (secondary source) edition (any edition other than princeps) ancientEdition (papyri?) modernEdition (papyri?) citation (text or object is cited) corrections illustrations otherPublications (is this different from secondary? @paregorios) principalEdition or editioPrinceps translation
If these values are on the <div type="bibliography"> then they may be plural.
If this makes senes, we can add it as a section in the bibliography Epidoc page.
Last edit: Elli Mylonas 2020-08-12
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I admit I did not read through all the past conversations, but I do not really understand the list of values in @source . I think I liked the examples with the use of @type (values= derived, autoptic), @source (pointer to another <bibl> or another pointable thing), eventually @prev, all of which into <bibl> inside the inscription, not in the master bibliography, as in the comment from 2017-03-12. I think this would make a nice addition to the recommendations in the GL. The point made than about the ability to use @source to point to whatever is "pointable" is, IMHO decisive and does solve the issue discussed, because one can have one to many pointers to things which are in or out of the project.
I would prefer the list of values for @typeto be suggested for <listBibl> only, because I would not like to change the way @type charachterizes <div> and ... because we do so in BM, where we have "manuscripts", but these include scrolls, inscriptions, etc. and we use edition and secondary as values for @type in listBibl, which strikes me as an interesting point of interchangeability.
The use of @type (values= derived, autoptic) could be limited to listBibl[@type='edition']
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
ok
I went through the markup list archive of Novembre 2014 when that bibliography topic was discussed.
It seems that "genetic" relations between bibliographic references can be expressed in 2 ways:
1. by classifying the references in categories ('primary', 'secondary', 'derived', principalEdition', 'translations', etc.)
2. by indicating, at the bibliographic item level (<bibl>), the source it is based on, wether an autopsy of the text bearing object or another edition.
In case 1. the genetic relation are implicitely derived from the semantics of the categories and does not try to establish precise relation between the editions, but it is quicker to encode and seems quite a pertinent level of simplicity to cope with that kind of information.
With case 2. it is possible to be very precise and get precisely interlinked bibliographic references, but needs identifiers and pointers, and assume a very well informed editor.
Some examples attribute values for case 1.
Could be a div[@type='bibliography'][@subtype] or a div[@type='bibliography']/listBibl[@type]:
secondary, edition, citations, corrections, illustrations, otherPublications, principalEdition, translations, ancientEdition, modernEdition (values collected by Tom Elliott)
Some examples for case 2
These examples are based on the editorial convention where editions based on an autopsy of the object are presented first. If an edition was based on another edition, it is then presented inside parenthesis, just after the reference it is derived from.
For example, this bibliographic statement (provided by Eleonora Santin):
Masson 1979, p. 59, nº 1 and p. 60 fig. 1 (Bile 1988, nº 57, p. 56)
Could be expressed with mixed content:
or (with explicit mention of the autoptic nature of the edition):
Remark:
The examples here suppose a master bibliography where the entries identified by 'Masson1979' and 'Bile1988'. The genetic relations could be expressed in an even more precise way:
The type values proposed in the discussion were @type "FS" (from Stone) and @type "FOS" (from other source) with @prev to indicate the genetic relation.
I would suggest using @source instead of theses two @tye values, because it can be used for every kind of external sources (printed editions but also squeezes, photographs, etc. as long as the have an identifier one can point to).
The absence of a @source attribute would mean just "not derived".
If there is a necessity of being more precise, other @type category could be used like 'autoptic', or 'derived' or anything else for the editor.
The same references could be encoded in a structured way I guess, letting the XSLT doing the parenthesis inclosure. My bet though is that editors would prefer the mixed content approach because it gives them more flexibility to add commentaries on the editions within the parenthesis which would be difficult to handle with a structured bibliography, especially if there are several reference inside the parenthesis, and the scope of the commentary is more than one ref.
The question still remains wether it would be useful to recommend values for categorisation of types of references. As I think this is closely linked to the editorial specifications of the collection the inscription is published in, I would try to harmonize too much things there, but I'd like more comments from others on that point of view.
Last edit: Emmanuelle Morlock 2017-12-03
Bumped -> Future.
@ellimylonas and @sdigiulio will discuss both of these questions (with reference to USEP) and report back with a proposal.
EM and SDG will report back.
@ellimylonas and @sdigiulio discussed this and although we think that it is unlikely many editors will want to characterize the authoritativeness and source of the editions in the bib., if they do @gethia has proposed using
@sourceon<bibl>with a recommended set of values as follows: please suggest better terms.1. "photograph"
2. "autopsy" (edition prepared from viewing original object)
3. "squeeze" (all forms of physical reproduction, including estampage and rubbings)
4. "facsimile" (microfilm, engraving, drawing)
5. 3D scan (this could be a "squeeze")
This can be laid out in a section on the bibliography page, and would belong in a master bibliography section.
However, we think it's difficult to classify a reference as a primary source, secondary source, generic information, in a master bibliography because the same publication could be a primary source for one object and a secondary reference for another that is just mentioned or cited. So this classification, which be handled with a
@typeattribute seems to belong in the bibliography for an individual object. Not a master bibliography. Or can be handled, as in the example already guidelines, by having separate<bibliography>elements.We recommend @type on the bibliographic references in the inscription with the following suggested values (from above):
secondary (secondary source)
edition (any edition other than princeps)
ancientEdition (papyri?)
modernEdition (papyri?)
citation (text or object is cited)
corrections
illustrations
otherPublications (is this different from secondary? @paregorios)
principalEdition or editioPrinceps
translation
If these values are on the
<div type="bibliography">then they may be plural.If this makes senes, we can add it as a section in the bibliography Epidoc page.
Last edit: Elli Mylonas 2020-08-12
I admit I did not read through all the past conversations, but I do not really understand the list of values in
@source. I think I liked the examples with the use of@type(values= derived, autoptic),@source(pointer to another<bibl>or another pointable thing), eventually@prev, all of which into<bibl>inside the inscription, not in the master bibliography, as in the comment from 2017-03-12. I think this would make a nice addition to the recommendations in the GL. The point made than about the ability to use@sourceto point to whatever is "pointable" is, IMHO decisive and does solve the issue discussed, because one can have one to many pointers to things which are in or out of the project.I would prefer the list of values for
@typeto be suggested for<listBibl>only, because I would not like to change the way@typecharachterizes<div>and ... because we do so in BM, where we have "manuscripts", but these include scrolls, inscriptions, etc. and we use edition and secondary as values for@typeinlistBibl, which strikes me as an interesting point of interchangeability.The use of
@type(values= derived, autoptic) could be limited tolistBibl[@type='edition']Made simple recommendations for some
<div type="bibliogaphy" subtype="xxx">values for the@subtypeattribute , and for<bibl type="xxx">close with commit 2822