I wrote about this recently [1], and here's my best shot:
Contrary to what some readers have said in e-mails to me, or inferred from what I’ve written, I’m actually not at all opposed to peer review or peer-reviewed publications. But the important thing these days isn’t a medium for publishing—pretty much anyone with an Internet connection can get that for free—but the imprimatur of peer-review, which says, “This guy [or gal] knows what he’s talking about.” A more intellectually honest way to go about peer-review would be to have every academic have a blog / website. When he or she has an article ready to go, he should post it, send a link to an editor, and ask the editor to kick it out to a peer-reviewer. Their comments, whether anonymous or not, should be appended to the article. If it’s accepted, it gets a link and perhaps the full-text copied and put in the “journal’s” main page. If it doesn’t, readers can judge its merits or lack thereof for themselves.
The sciences arguably already have this, because important papers appear on arXiv.org before they’re officially “published.” But papers in the sciences appear to be less status-based and more content-based than papers in the humanities.
I think this change will happen in the humanities, very slowly, over time; it won’t be fast because there’s no reason for it to be fast, and the profession’s gatekeepers are entrenched and have zero incentive to change. If anything, they have a strong incentive to maintain the system, because doing that raises their own status and increases their own power within the profession.
Given how cheaply one can find or buy a website / blog these days, I'm not sure where all this money is going.
Researchers and scientists could have asymmetric keys and peer review in that fashion. The authors could append cryptographic signatures to the article. Then we don't need centralized journals, just a list of academics whose scientific rigor is trusted.
I have often thought about doing something similar to this. Basically, you publish a signed version of your paper on your own website, and then other scientists can sign it with notes. It would be kind of like a "Facebook like" system. (By "sign", I mean a cryptographic signature.)
"82 PhD scientists have reviewed this paper, including 3 people in your close network and 2 more people from your extended network." ...or something like that.
You could do a network-graph search. Start with some professors that you trust, and some keywords that you're interested in and start walking the graph from there.
I don't know what the profit motive would be for someone to build this though. Nor the motive for individual reviewers to carefully review someone's paper and provide detailed useful feedback...
