MODS: The Future of XML Bibliographic Metadata?
I’ve come across this question in two different forms today. Raymond Yee asks:
Question: As someone interested in marking up bibliographic metadata in XML (or with the RdfSpec), what specification/standard should I use? I’ve not found good answers yet to that questions.
Similarly, someone asked on the RefDB list:
- Is there really no established standard for representing references in XML? By TEI, for example, or something linked to the Dublin Core…
I confess to not really understanding RDF, nor its applicability to bibliographic metadata, despite the best efforts of Steve Cayzer to explain it to me. This doesn”t mean I’m an RDF sceptic; it just means I don’t understand it.
As for an adequate XML bibliographic metadata standard:
IMHO MODS offers structurally-sound comprehensiveness, without being too unwieldy. I see nothing else out there that is suitable for the range of documents I store and cite. Coupled with the fact that it is backed up by the Library of Congress, and part of its comprehensive XML strategy, makes MODS a more-or-less no-brainer.
Is it perfect? No, of course not. It was designed for the library community, and there are things that are missing as a result. There is no obvious way to code publication status, for example. There is also no facility to code abbreviated names.
Still, rather minor critiques, and MODS has an extension element that allows adding markup from other namespaces. And the forthcoming (any day now) v3 of the schema adds a crucial new element for bibliographic markup that I have seen nowhere aside from the seemingly defunct BibX: “part.” It is used to describe details associated with so-called analyticals: book chapters, journal articles, web pages, etc. If nothing else, this structure from BibX will live on in MODS.
Ultimately, though, technology is social as much as anything, and the success of any standard depends on projects implementing it to solve real problems.
By “abbreviated names” do you mean personal, corporate and geographic names? (I’m assuming you don’t mean titles.) I guess this would be another area where MODS reveals its library origins, since in a library system, any alternate names used should be coded in the 4xx or 5xx fields (”see” or “see also”) in the appropriate authority record. Burden on the authority file rather than the individual record…
I just happened upon your site this morning and I definitely agree that MODS is intriguing for coding bibliographic information. It has its limitations though, as you note. I’m not as familiar with MODS as I am in MARC, but MARC doesn’t really do that great a job of denoting certain types of relationships anyway, despite the plethora of 7xx, 8xx, 4xx and some 5xx fields. And I see that in mapping MARC to MODS, all those fields are kind of scrunched down into the “relatedItem” field.
I wonder how this scrunching down affects the lack of the crucial element of “part” that you mention. With analytics that are series, 4xx and 8xx fields are often used to denote that relationship. Not sure about parts that are constituent, since I haven’t seen many examples. Sometimes they’re implied with 700/710 fields (additional entries that have some form of responsibility for the work.) Actually, now that I look at the format (http://www.oclc.org/bibformats/en/7xx/), it appears that 773/774 could be used to denote the constituent/host relationship, but I’m not sure how widely this is done in actual practice. Or, of course, if it wouldn’t just get crazy unwieldy for your purposes to use relatedItem (with type=”host” or type=”constituent”) to indicate those article-to-journal or chapter-to-book relationships.
Probably easier just to wait for the new schema.
On names, yes, I’m looking for something equivalent to the abbreviated titles already there. If I need to handle a name like the FBI or an elaborate court name, it would be nice to get it in the record correctly.
With respect to authority lists, I think the next step for MODS is to incorporate them into the validation process. It would be nice to have my own XML-based authority list that I could just reference from withiin MODS. For me, of course, there would be nothing particularly fancy about it. It’d just be a list of names, with additional info not carried in MODS.
Any suggestions on what this ought to look like? I’ve never seen an authority list.
As for the analytics, in MODS one indeed would use the relatedItem “host” and within that the new “part” element to cover volume, issue, page numbers, etc.
Another limitation I was reminded of the other day is that there are no attribute on the “place” element to distinguish, for example, an event place (say a conference location) from a publication place (say a proceedings).
Still, MODS is much better than what those of us in the personal bibliographic management field are used to!
To see the authority record for FBI, go to http://authorities.loc.gov and look up “FBI” as a name authority heading. Click on through to the authority record and toggle between the MARC and labelled displays if you like. You’ll see that “FBI” is listed as one of the alternate titles, but the “official” heading to use is the “United States. Federal Bureau of Investigation” long form. In a way, this type of authority file doesn’t distinguish abbreviated names from other forms of alternate names, though it does distinguish inverted names, jurisdiction names and names in direct order via the MARC indicators. I’m not sure if there’s an equivalent in MODS…?
The trick to effectively using the authority file is a sense of seamlessness, I think. In an integrated library system, you could type in “FBI” and be able to find everything listed under “United States. Federal Bureau of Investigation”. Now that I think about it, I’m not sure you’d want to adopt this sort of methodology directly in terms of actual record creation. I mean, such a scenario would maybe find you validating a record, and the system suggesting the official form and substituting it in your record. But you would want the abbreviated form to remain in the record itself? The “referencing” aspect of the authority file is definitely key, and I’m sure there’s some way to do this referencing without actual substitution.
Not sure now that using an authority file would necessarily address your issue now, actually… Think too much in the library box sometimes so this is a nice way to stretch.
Common use of MARC relies a lot on external library practices (what are the things that are typically cataloged, rules of input, etc.) which speaks to MARC being a more, um, old-school form of markup I think, not to mention a rather entrenched one. It’d be nice to be able to specifically articulate some of those practices in the markup: tagging event and publication places for conferences, for example, instead of just knowing that event place names aren’t typically specifically tagged as such and often (but not always) just appear as part of a title field.. or perhaps as a geographic subject heading if it’s relevant that way.
Enjoying your site…