Print

Print


Hallo Oliver,
I would also tend to favor a systematic and deterministic encoding  
using <biblStruct>. <biblFull> was indeed intended to have in one  
element a sort of copy of the header, but <biblStruct> better matches  
the kind of bibliographic representation we need for project like  
yours or similar digital library corpora. As a matter of fact we  
started to compile a page recording encoding models (in German) using  
biblStruct:
cf. http://colab.mpdl.mpg.de/mediawiki/TEI_Bibliographic_Information
which in turns points to a "special issues" page that you can  
contribute to:
http://colab.mpdl.mpg.de/mediawiki/TEI_Bibliographic_Information/ 
special_issues
Schönes Wochenende,
Laurent


Le 17 mai 08 ā 16:40, Oliver Duntze a écrit :


> Dear TEIers
>
> I'm trying to encode some rather complex bibliographical data in the
> <sourceDesc>-Element of the TEI-Header, and I'm wondering if there is
> any best practice for encoding the bibliographical record of printed
> sources. Especially, I'm not quite sure about the decision between  
> <biblFull> or <biblStruct>.
>
> For instance the guidelines at diglib.org are recommending  
> <biblFull>. Unforunately,the examples given there are quite simple.  
> It seems to me, that <biblFull> is lacking some important elements,  
> especially for encoding the scope of journal-articles or parts of  
> multivolume works.
>
> An example: Novalis' essay "Die Christenheit oder Europa" was edited
> posthumously by L. Tieck and F. Schlegel in the first part of the 4th
> edition of Novalis' "Schriften" (collected works). In an unstructured
> format that would be:
>
> <bibl>Novalis: Die Christenheit oder Europa. Ein Fragment.  
> (Geschrieben
> im Jahre 1799). In: Novalis: Schriften. Vierte, vermehrte Auflage.  
> Hrsg.
> v. Ludwig Tieck u. Friedrich Schlegel. Bd. 1. Berlin: Reimer, 1826, S.
> 187-208. (Copy used: mylibrary, callumber xyz)</bibl>
>
> It's no problem encoding this in <biblStruct>. But when I try to
> encode it in the <biblFull>-Format, it looks like this:
>
> <biblFull>
> <titleStmt>
>   <title level="a" type="main">Die Christenheit oder Europa</title>
>   <title level="a" type="subordinate">Ein Fragment.(Geschrieben im
> 				Jahre 1799)</title>
>   <title level="m" type="main">Schriften</title>
>   <title level="m" type="subordinate">Bd. 1</title>
>   <author>Novalis</author>
> </titleStmt>
> <editionStmt>
>   <edition>Vierte, vermehrte Auflage</edition>
>   <respStmt>
>     <resp>Hrsg. v. </resp>
>     <name>Tieck, Ludwig ; Schlegel, Friedrich</name>
>   </respStmt>
> </editionStmt>
> <extent>S. 187-208</extent>
> <publicationStmt>
>   <publisher>Reimer</publisher>
>   <pubPlace>Berlin</pubPlace>
>   <date>1826</date>
> </publicationStmt>
> <notesStmt>
>    <note type="location">
>      <name type="library">mylibrary</name>
>      <ident type="callNo">xyz</ident>
>    </note>
> </notesStmt>
> </biblFull>
>
> As far as I can see, there are at least three problems:
> - The scope of the essay (p. 187-208) may be encoded in the
> <extent>-element, but I would prefer that for information like "20  
> pp."
> or "10MB". Using <biblStruct>, it could be encoded in the
> <biblScope>-element.
>
> - It is difficult to encode that the essay was printed in the first
> volume of the "Schriften", because the volume-number is not exactly  
> a subtitle (in <biblStruct> this would be encoded in an <biblScope>- 
> element too).
> It may be possible to put this information in the <seriesStmt>. But a)
> the "Schriften" are not a series, but *one* work printed in several
> volumes. And what if b) the there's also a 'real' series title:  
> E.g. if
> Novalis' Schriften were printed as part 3-6 of a series "German  
> writers
> of the 19th century"?
>
> - Finally, it is not clear if the <author>-element is referring to the
> author of the essay or to the author of the monography. In this case
> both are identical, but it could be two persons. E.g. if you like to
> encode the bibliographical record of Tiecks introduction to Novalis'
> work, which was printed as a foreword to the fist part of the  
> "Schriften".
>
> I would be thankful for any information if these problems also  
> occur in
> other digitization projects, and how they are dealing with
> bibliographical metadata.
>
> Thanks,
> Oliver
>
> -- 
> ***************************************************
>
> Dr. Oliver Duntze
> Deutsches Textarchiv
> Berlin-Brandenburgische Akademie der Wissenschaften
>
> Jägerstr. 22/23
> 10117 Berlin
>
> Tel.:     +49 (0)30 20370 523
> E-Mail:   [log in to unmask]
>
> ***************************************************
>