One of the exciting things about Opentech UK is meeting people who can contribute to your own work from the directions that you hadn’t thought of. So I met up with James Hetherington, from AMEE (http://wwmm.ch.cam.ac.uk/blogs/murrayrust/?p=2601 ) a group that provides an open source carbon calculator. We got talking and the subject came to bibliography and our ideas of helping the clarity of climate research through an Open Bibliography approach.
He then showed me an impressive report from the Netherlands
Where they had reviewed in detail the statements made in, or reported about, the IPCC report. This was a very careful analysis and categorised a large number of error types. I should emphasize that the conclusion was that the errors in the report or in the reporting of the report were not substantive. Nonetheless they showed that there were bibliographic errors in the report and categorised their types.
Two examples of the types of bibliographic error were:
- Bibliographic entry was incorrect (e.g. typographical errors)
- Referenced material did not exist (including disappearing web pages)
This is exactly the sort of thing that our Open Bibliography project #jiscopenbib has been set up to address. It encourages us to offer a bibliographic framework that might be used for validating the bibliography of climate added also got me thinking about how unit test methodology it might be used for bibliography.
Imagine that the world’s scholarship was referenced by an open-bibliography which was available pervasively. Every time somebody referenced an article or book or document or report that had an entry in this bibliography the system could immediately check whether the reference was correct. It could also follow the link to an online resource (and we expect that an increasing number of such links will be online) and check whether it existed. Any failings could be reported in the same way as software unit tests. The check could be run as frequently as desired, and certainly daily. For example a student writing a thesis could check their bibliography at frequent intervals.
There is no technical reason why this could not be put together very quickly, at least for electronic journals which are online. A simple collection of all the table of contents of all journals constitutes such a bibliography. This is straightforward to do from the online journals themselves and as far as we know breaks no intellectual property rights or contracts. Several publishers have already expressed support for the idea, which is to their own benefit since it advertises their material.
So part of what we will be building are the tools which allow us to do this.
If you’d like to help – whether you are an author, reader, publisher, funder or librarian – just let us know