If academics will just insist on not assigning copyright, they may get further; publications need authors, and to demand copyright assignment will become a competitive disadvantage.
"The ACM copyright form now explicitly says that authors can post a copy on their personal websites" --scott_s
Actually, ACM can be pushed further than that. When ACM solicited an article from me for their Communications magazine, I (nicely) refused to execute their copyright assignment, even with the personal website permission. So ACM proposed instead a simple licence to publish (royalty-free, worldwide, non-expiring, any medium, non-exclusive) which I was happy to sign while preserving my copyright ownership, and which gave them everything they need. But recently IEEE has refused a similar licence arrangement, insisting on a copyright assignment or no publication (which was what I chose). I expect ACM to prosper over IEEE.
I really don't think the journal model works well anymore.
It used to be that distributing papers was burdensome, so they compiled them into journals and professionals paid for a subscription. Built into the cost was peer review -- they couldn't just publish ALL the articles, because there was limited space, so they had to have professional reviewers filter out all but the "best."
So now that distribution has been solved, the only value-add that a journal can claim is that filtering process, which ostensibly is costly.
It all hinges on the question of whether peer review can happen without compensation. I think it's pretty well-established now that peer review without compensation happens constantly. Welcome to the internet, right?
So the correct new model is something like a digg/HN system that aggregates articles from various fields, and allows peers to review, respond to, and upvote important papers. I think it'd need to be more complex because of voter credibility concerns, but it seems workable. I can't say the same for the current journal model.
I think it's pretty well-established now that peer review without compensation happens constantly. Welcome to the internet, right?
As far as I know, even before the internet era, you never got paid for referee reports. I've certainly never been paid for one, though I've only been writing them since 2006.
edit: actually, this sounds awesome. Is it out there? You'd have to change the rules somewhat for more persistence and maybe moderate the submissions (for general quality). Grouping them would also be good.
I'm referring to something more like hacker news in spirit (good papers get upvoted) but much more constrained. Give voting to qualified academics (maybe fractional votes to grad students as well?) with more persistence and multiple categories/tags.
Creative commons and the arxiv give you a very simple solution.
1. Upload to the arxiv under a creative commons license.
2. Submit, and post to your webpage.
3. If the journal ever demands you take the paper down, do so. Then download the creative commons version from the arxiv and post that (or just link to it).
Many journals will not consider an article if it has been previously published.
Publishing to your own web page may not count but it will surely not make the journal happy, especially if you hope to get published in the same journal again.
There is a small but growing number of open access journals that do not require the author(s) to assign copyright. The PLoS family and the Frontiers In family both come immediately to mind, and I am sure there are others.
Simple solution: a requirement of receiving public funds should be that authors are required to make their work available for free online (possibly via a community license).
This way the journals don't get a choice as to what copyright scheme they'd like to use.
"The ACM copyright form now explicitly says that authors can post a copy on their personal websites" --scott_s
Actually, ACM can be pushed further than that. When ACM solicited an article from me for their Communications magazine, I (nicely) refused to execute their copyright assignment, even with the personal website permission. So ACM proposed instead a simple licence to publish (royalty-free, worldwide, non-expiring, any medium, non-exclusive) which I was happy to sign while preserving my copyright ownership, and which gave them everything they need. But recently IEEE has refused a similar licence arrangement, insisting on a copyright assignment or no publication (which was what I chose). I expect ACM to prosper over IEEE.