I’d like to have a session about citation formats and bibliographic processing. Not sure if this would be a hackathon or a general discussion, probably a bit of both.

The thing is, citation formats evolved in the days of paper – they’re a form of text based hypertext. In the old days when you referenced something you had to put enough bibliographic detail in your text so that people could find it. We still have to format articles, theses, essays etc with redundant text-formatted references and bibliographies to submit them to publishers and markers, even though we’re using machines to manage all the references. And we’re still teaching students to do this, sometimes by hand.

In many disciplines we have online resources so in many cases a citation could be a URI referencing a good quality stable data source. But URIs are not always going to be the way to go, in which case the bibliographic data could be embedded in text in a way that makes re-processing easy.

This session could look at what can be done to rationalise citation practices so that an author can use existing bibliograpic databases (via stuff like the Open Bibliogrpahy project, Zotero, Mendeley, CrossRef et al) without having to maintain their own, unless they want to of course, and downstream consumers (publisers, readers, markers etc) can choose how they would like to view, reuse or otherwise process the references.

In the sciences there are many disciplines where citing by DOI would be sufficient to cover almost all use-cases, but this is certainly not the case in the humanities.

We could talk about:

• How to embed citation-by-reference and citation-with-bibliographic-data in HTML and how to choose which to do. (I have some ways of doing this using HTML5 Microdata I’d like feedback on)
• How to produce said HTML using tools such as Word, Wiki formats, Pandoc, LaTeX, WordPress etc. (I have made some progress on a tool using Zotero + MS Word producing HTML5 that can be re-formatted automatically to suit the reader, with bib-data embedded in the HTML for machine processing as well).
• What are the limits of this approach? There will be lots of areas of the humanities where trying to construct bibliographic entries for the stuff you are referencing will be hard.