If the first part of inventing is to try to think without constraints, the second part is to accept reality.
The reality of citations today is that they are messy, and that they have always been messy. At no point has there been a mechanism for following the citations in a document in a direct way to access the source, with the exception of in Doug Engelbart’s Augment for a brief, beautiful spell, starting in the 60s and fading away during the 80s.
The history of paper citations is long and interesting but the short history of digital citations is what’s pertinent to my future work with interactive text. This is a history replete with dreams of digital dimensions cut short by commercial concerns. At the JATS conference this last week I had the rude re-awakening that author’s work interactively with their sources, using third party citation management software or whatever else suits them, and then they strip all the semantics out of their work to present a PDF with the visual appearance of a traditional (and this is not a very long tradition but it appears to be very sticky) of a reference section at the end of a document with no active links to sources, only paper-link ‘references’.
I have moaned in previous posts about the loss of connective knowledge of having paper-simulacra be the goal for presentation and communication of digital documents http://wordpress.liquid.info/jats/ and how the importance of the cosmetics of display style is holding connection possibilities mute.
There is also the evergreen problem of users not having access to databases of academic documents with their meta-data, or only the meta-data, to help them get overviews and get to grips with the large volume of (potentially) relevant source materials for their work. But this is how it is. Creating a powerful visual interaction of a potential ‘citation space’ is about as real-world useful as creating a large, outdoor unicorn sanctuary.
The job then, of software systems aiming to augment a student’s thesis writing will be to ease the process of entering a citation once found and to export it in as useful a way as possible for the next reader, which is why we will have to mark up PDF properly and include all the meta for future citations. That was the in and out.
What needs to be designed is the best way for the author to deal with citations once in their system, to analyse and cite them.
The analysis of citations are both at the individual and collective level. At the individual level it’s about reading the documents in different ways and at the collective level it’s about seeing connections and patterns to generate useful understanding and insight into the problem space.
“I have read this document and I have the body text, my annotations and I know the basic identifying information to cite it. Now what can I do with it?” Is a prime question. Considering that the user will have read the document, the act becomes one of finding the right document when making a certain point and about expressing something from that document in a way which the reader can easily check.
Note. This post was typed to one glass of Madeira. I’m not good with drinking but I think I managed to get my thoughts down on this.