PhDs take years, are an "original contribution to knowledge", and are generally paid for by the taxpayer. They are then generally read by a handful of people -- generally a supervisor or two, and your examiners -- and then left to rot as a PDF on a university document repository of some kind.
I don't think this is good enough.
"Open Data" has increasingly come to mean large, public data sets accessible by APIs. I argue that in addition to this, we should be thinking about publishing our research in the most open and beautiful ways possible. Publishing as, for example, HTML5 in addition to PDF gives several benefits.
- Research can then be crawled by a search engine. PhDs are so specific, finding them directly is usually after searching for an author. A web-friendly form can skip this, allowing people to find unusual combinations of words that the author may not anticipate.
- It allows access to other collaborative tools. For example, a thesis could have a per-paragraph or per-section commenting system, allowing people to respond to a specific aspect.
- A proper digital pipeline reduces work. As O'Reilly (amongst others) are currently showing, HTML5 is a solid publishing format that exports well to PDF, ePub, iOS and other modern formats. Instead of creating several different versions of the same document, a single source can generate multiple outputs.
- Modern versionning tools can be used. This one requires greater digital literacy, but potentially github and local webservers can be used for edits and changes, and forks and edits done by other people. "With many eyeballs, all bugs are shallow" could one day apply to research as well.
- HTML is a free, open source format. Academics should be trying to abandon proprietary software full stop, in my view. HTML is a good alternative to current, terrible tools for thesis writing such as Microsoft Word, and is much less intimidating than LaTeX, with a variety of good editors.
Therefore I will be attempting to convert my PhD thesis to HTML5 over the next few weeks, keeping these things in mind. The output from my current editor, LyX, is not nearly as clean as I would like, so this may take some time: hopefully with a vision of what a PhD can look like online though, I can create a positive blueprint for future tools.