#347 clarify and rationalize encoding of pagination in bibliograp

RED
closed-rejected
Kevin Hawkins
5
2012-08-08
2012-03-05
Kevin Hawkins
No

Discussion

  • Lou Burnard
    Lou Burnard
    2012-03-15

    • milestone: --> 871209
     
  • Lou Burnard
    Lou Burnard
    2012-03-15

    This clarirfication seems useful. It does however involve quite a change in the meaning of bibl/biblScope which is likely to break many existing documents.

     
  • Kevin Hawkins
    Kevin Hawkins
    2012-03-19

    To clarify, you mean that under this proposal, use of <biblScope> within <bibl> would change significantly? Is that because till now people have been using <biblScope type="pp"> inside <bibl> for the page range relating to an entire bibliographic item, not just relating to the cited portion? Note that under paragraph (2) of the proposal, this practice would only be deprecated.

     
  • BODARD Gabriel
    BODARD Gabriel
    2012-03-20

    I'm not sure I understand--the suggestion is that <biblScope type="pp"> should only be used to refer to the cited subset of a work, whereas the full page-range of an article, say, should be encoded with <extent>? Would this not make <biblScope type="pp"> different from most other @types of biblScope in a bibliographic item, like "issue", "volume", etc., which are usually referring to the full range of an item not the cited range of a subset?

    Not sure I'm making sense. An example: a <biblScope type="issue"> in the journal part of a bibliographic reference to an article contains the issue number of the volume in which the article appears. It's the whole thing, in other words. But a <biblScope type="pp"> in the same bibliographic reference would, under this proposal, not refer to the page range of the whole article, but the page(s) of the subset of the article which the reference is citing. Am I understanding correctly?

    Is that not inconsistent?

    I should stress that I do like the idea of being able to distinguish between pp. 100-113 being the range of the whole article and pp. 100-101 being the passage I'm citing in that article--I just need convincing that this is the right way to do it. (This would be an expansion of the use of <extent> too, wouldn't it?)

     
  • Kevin Hawkins
    Kevin Hawkins
    2012-03-25

    Gabby,

    You understand correctly that the proposal is that <biblScope type="pp"> should only be used to refer to the cited subset of a work, whereas the full page-range of an article, say, should be encoded with <extent>.

    As for your example involving <biblScope type="issue">, I don't think of <biblScope type="issue">2</biblScope> as indicating "the whole thing". It's limiting bibliographic data to a particular issue of the journal.

     
  • BODARD Gabriel
    BODARD Gabriel
    2012-04-17

    An example that I think doesn't work in the proposed solution:

    In: "P. Romanelli, Leptis Magna (Africae Italiana: Monografie a cura del Ministero delle Colonie, 1), Roma, 1925" this volume is number 1 in a series called Africa Italiana. The citation is not to the whole series, but to this volume, so we use biblScope.

    In: "R. Bartoccini, 'Roman Roads' in Epigraphica X (1948) 150-7" this article is at pages 150-7 of the journal issue in question. The citation is not to the whole issue, but to this range of pages, so we use biblScope.

    The proposal implies that we will continue to use <biblScope> this way in my first example, but move to <extent> for the second example. I just don't see a logical distinction between the two that would justify this change of usage.

    (I would rather see a new element like <biblCitedRange> or something, for the usage that you are proposing we continue to use biblScope type="pp" for, i.e. "to refer to the cited subset of a work". Or are we just talking about a different attribute value: <biblScope type="pp"> for the whole article; <biblScope type="cited-pp"> for the cited subset?)

     
  • James Cummings
    James Cummings
    2012-06-29

    • assigned_to: nobody --> gabrielbodard
     
  • Kevin Hawkins
    Kevin Hawkins
    2012-06-29

    • assigned_to: gabrielbodard --> kshawkin
     
  • Kevin Hawkins
    Kevin Hawkins
    2012-06-29

    James assigned to Gabby, but per Ann Arbor minutes, I am supposed to read Gabby's comments closely and revise tcw23 to respond to these points. Still haven't gotten to it, but I will! So reassigning to me.

     
  • Kevin Hawkins
    Kevin Hawkins
    2012-08-05

    Let me first address Gabby's comment in parentheses. I believe he is saying that we can keep this all much simpler by changing nothing in the TEI except introducing a new element <biblCitedRange> for the cited subset of a work, which would occur anywhere that <biblScope> is currently allowed. We would then change many though not all examples in the Guidelines that currently use <biblScope> to use <biblCitedRange>. What would be left in <biblScope> would be those cases where you're giving the whole range of pages for an item -- basically, those that in the proposal would be moved to <extent>. At that point it would be even less clear to me now what the semantic difference is between <biblScope> and <extent>. So while it seems less disruptive, I don't know that we've solved much.

    In examining Gabby's comment outside the parentheses, I see that while this proposal brings a certain symmetry to the TEI, it also makes some choices between options without justification.

    Under the proposal, Gabby's first example would be:

    (1) <biblStruct>
    <monogr>
    <author>P. Romanelli</author>
    <title>Leptis Magna</title>
    <imprint>
    <pubPlace>Roma</pubPlace>
    <date>1925</date>
    </imprint>
    </monogr>
    <series>
    <title>Africae Italiana: Monografie a cura del Ministero delle Colonie</title>
    <biblScope type="vol">1</biblScope>
    </series>
    </biblStruct>

    I think he and I agree that we like this. But his second example would be:

    (2a) <biblStruct>
    <analytic>
    <author>R. Bartoccini</author>
    <title>Roman Roads</title>
    <extent type="pp" from="150" to="157">150-7</extent>
    </analytic>
    <monogr>
    <title>Epigraphica</title>
    <biblScope type="vol">X</biblScope>
    <imprint>
    <date>1948</date>
    </imprint>
    </monogr>
    </biblStruct>

    and yet, by the principle followed (though never explicitly explained) in the proposal of using <extent> for the whole and <biblScope> for the parts (the way to achieve goal 1 in the proposal), this second citation could just as easily be:

    (2b) <biblStruct>
    <analytic>
    <author>R. Bartoccini</author>
    <title>Roman Roads</title>
    </analytic>
    <monogr>
    <title>Epigraphica</title>
    <biblScope type="vol">X</biblScope>
    <biblScope type="pp" from="150" to="157">150-7</biblScope>
    <imprint>
    <date>1948</date>
    </imprint>
    </monogr>
    </biblStruct>

    That is, there's no justification for why the range of pages for an entire article in a journal or an entire chapter in a book should be in analytic/extent[@type='pp'] and not in monogr/biblScope[@type='pp']. Similarly, If you wanted to give the range of pages for an entire volume in a series that has continual numbering throughout, there's no justification for why you should use monogr/extent[@type='pp'] and not series/biblScope[@type='pp']. Putting it another way, we haven't achieved goal 2 of the proposal. And I'm not sure what to propose to fix this.

    Looking at (2a) and (2b), I realize that even these are misleading since *Epigraphic* was published for many years, not just in 1948. If we fixed this in (2b), it would in theory be:

    (3) <biblStruct>
    <analytic>
    <author>R. Bartoccini</author>
    <title>Roman Roads</title>
    </analytic>
    <monogr>
    <title>Epigraphica</title>
    <biblScope type="vol">X</biblScope>
    <biblScope type="pp" from="150" to="157">150-7</biblScope>
    <biblScope type="date">1948</biblScope>
    </monogr>
    </biblStruct>

    but then we would need to change the content model of <monogr> to no longer require an <imprint>.

    In fact, now that I think about it, I wonder whether the TEI should in fact use <series>, not <monogr>, for information about a journal, which would require changing the content model of <biblStruct> and the definition of <monogr> (instead of changing the content model of <monogr> to require an <imprint>). That, is, it would instead be encoded as:

    (4) <biblStruct>
    <analytic>
    <author>R. Bartoccini</author>
    <title>Roman Roads</title>
    </analytic>
    <series>
    <title>Epigraphica</title>
    <biblScope type="vol">X</biblScope>
    <biblScope type="pp" from="150" to="157">150-7</biblScope>
    <biblScope type="date">1948</biblScope>
    </series>
    </biblStruct>

    I'm not sure of the path forward. :(

     
  • Kevin Hawkins
    Kevin Hawkins
    2012-08-05

    • milestone: 871209 --> RED
     
  • BODARD Gabriel
    BODARD Gabriel
    2012-08-06

    I'm obviously a lot less specialist in bibliographic markup than Kevin, but my understanding of the different between <extent> and <biblScope> is that the latter allows you to express a range (of pages, volumes etc.) that defines a part (a scope) of a larger whole: pp. 12-20 of a book, issue 115 of a journal, etc., whereas <extent> is for a expression of total size: this book contains xiv+355 pages; this monograph is made up of 3 volumes; this film is 100 minutes long. (So I agree that biblScope as in (2b) is more appropriate than extent as in (2a)--although I'm still twitchy on the location of it.)

    The difficult distinction, therefore, is not between biblScope and extent, which as I see it are pretty unambiguous (extent is rarely if ever used in a footnote; biblScope perhaps not so often used in a library or bookseller catalogue), but between biblScope and biblCitedRange. This would involve explaining the difference between two kinds of page numbers (e.g.) which are currently, and fairly intuitively, tagged using biblScope.

    Two questions: #1 can we comfortably and unambiguously define the difference between a range of pages (or single page) that define the scope of a complete bibliographic item being listed, and a range of pages of single page that define the subset of the listed bibliographic item being cited? #2 do we feel it's necessary to distinguish between these two things, given that we've never done so before? I wonder if the distinction may not inherent in where in your bibliographic hierarchy the biblScope lives, but I'd have to sit down and look at some examples with the help of an expert like Kevin or Laurent to come up with a coherent proposal on this front.

    Incidentally, I'm not convinced that we should ever use <biblScope type="date">--that seems barbaric to me. The date in the imprint of the journal example refers to the publication date of Epigraphica 10, so it is not a further subset of that issue. If the journal Epigraphica had changed publishers over the years, presumably the publisher and pubPlace within that imprint would also refer to those in place when issue 10 was published, and not imply that it referred to the entire run of the journal. The presence of <biblscope type="pp"> in this same monogr is confusing, though: it's not the issue of the journal that is limited by these pages, but the article within it.

     
  • Martin Holmes
    Martin Holmes
    2012-08-06

    On Gabby's two questions:

    #1 can we comfortably and unambiguously define the
    difference between a range of pages (or single page) that define the scope
    of a complete bibliographic item being listed, and a range of pages of
    single page that define the subset of the listed bibliographic item being
    cited?

    I think we can, but doesn't the context usually do this? An item in a bibliography is surely going to contain the full page-range of an analytic item within its containing monogr, whereas a quotation in a text is going to refer to the specific page(s) of the quotation itself (and may link to a full bibliographic item). Perhaps there are cases where both would appear within the same biblStruct, but I haven't come across any.

    #2 do we feel it's necessary to distinguish between these two
    things, given that we've never done so before?

    I'm not convinced of the need to do this, but if it is required, then I'd be much happier with adding a new recommended value for @type (<biblScope type="citedRange">) than adding a new element.

     
  • Kevin Hawkins
    Kevin Hawkins
    2012-08-06

    On question #2, I think it's important to be able to make this distinction. If you have citations encoded in TEI that you want to be able to send as queries to CrossRef to retrieve DOIs, you want to send only page ranges for the scope of a complete bibliographic item, not ranges of pages or single pages that define the subset of the item.

    Gabby and I just skyped and came out with a proposal. I'm going to jot my notes and send to him to make sure I got it right; will share everyone soon.

     
  • Laurent Romary
    Laurent Romary
    2012-08-07

    Just read all the thread and for the record, I would favor a solution with @tape on biblScope.

     
  • BODARD Gabriel
    BODARD Gabriel
    2012-08-07

    Kevin and I have discussed this issue in depth off-line, and have come up with two new proposals as an alternative to the one suggested in this ticket. We have opened two new tickets:

    http://purl.org/TEI/FR/3555190 Improve guidance and restrict usage of biblScope
    and
    http://purl.org/TEI/FR/3555191 New element <citedRange> for bibliography

    I suggest we close this ticket and discuss those two separate proposals at the relevant tickets.

     
  • Laurent Romary
    Laurent Romary
    2012-08-08

    Seems to be a way to move forward. Let's try to sort out the two new tickets.

     
  • Martin Holmes
    Martin Holmes
    2012-08-08

    No objection from me.

     
  • Kevin Hawkins
    Kevin Hawkins
    2012-08-08

    Excellent, I'm now closing this ticket.

     
  • Kevin Hawkins
    Kevin Hawkins
    2012-08-08

    • status: open --> closed-rejected