Why do academics tend to come up with overly complicated things?
While visiting conferences in Digital Humanities, and now lately the LiquidPub workshop, I have seen many great and well-thought out ideas, especially considering the complex problems many of the projects are trying to solve (such as encoding complicated documents).
However, I have also seen quite a few proposals and/or projects as well, which seem to try things that are quite far-fetched, if not outlandish. What most surprises me is that these concern things which are clearly complicated and not well-understood, and yet are treated in a highly technical and theoretical manner (as opposed to experimentally).
An example of this is the idea of micro-attribution in papers. The basic idea is simple: Individual parts of papers should have their own author(s). And that credit for each part should be attributed to it's author only (or mainly).
While this seems to be a simple, and sensible idea, it opens a psychological/sociological can of worms. Namely: What does working together mean in academia? How much credit would each team-member deserve? How does writing relate to research? And is it really bad (both in a moral or practical sense) when a mentoring professor gets some, or most of the credit (I don't think it necessarily is) ?
To legitimately propose or tackle such a thing, on should at least be, or involve a sociologist of sience. And on a more practical plane, the incentives for micro-attribution are wrong, as it asks lead-researchers to give up credit.
But the real problem here is that among academics complexity is considered good. This could be because simple solutions may seem too simple to be intelligent, or because simple solutions have already been proposed by others. Similarly to the situation in modern art, only the novel, the new - no matter how impractical or bad - is credited, as opposed to finding out what things work (experimentation). While often, if not almost always, some things have not been proposed or done before, because they simply ...were a bad idea (so much for toilets outside restrooms in museums).
Making the web work for academics will require a lot of experimentation and tweaking in the details of simple systems, ideally based on real user-feedback. It is an empirical matter that will not be helped that much by ever more complicated proposals that are implemented in a half-baked way (if at all) and then described in detail in publications. Occams Razor should be the guiding principle, at least until a system actually takes off, and gets a significant user-base.