[pallandt]

If I remember correctly Elsevier's rules, in case the article is from a subscription-based journal (non-open access) they retain copyright to the final published peer-reviewed & formatted article. The author is allowed to share their final published paper only in a 'scholarly' context, which seems to mean anything but mass dissemination; you're allowed to share it with your peers privately, but not post it on your site for example. I've just re-checked and it seems this rule is still enforced: http://www.elsevier.com/journal-authors/author-rights-and-re...

[legutierr]

OK, but to be fair, this seems to be slimy behavior on the part of Elsevier, not Academia.edu. Why should academics give up self-publishing rights just to go through peer review and get published?

[maccard]

Because, before the internet was widespread and used for distribution, the only way to get your research out was to be published in journals. Part of this deal involves handing over the rights to your paper to the publishers. This is pretty standard procedure: JK Rowling isn't going to be going giving rights to penguin books for the Harry potter books while she has a contract whereby Bloomberg publish the books currently, it's a similar idea. Additionaly, the measure of an academics success is based on his (or her) citations, and if your citations are from "website of some guy", nobody is going to take you seriously. It's a vicious circle, and the whole system needs reforming, however the thing is, conferences such as SIGGRAPH, FOCS, INFOCOM etc etc all play an important role, that they offer peer reviewing, and you can guarantee quality (if something is published in SIGGRAPH, then you can normally rely on it). Until there are viable alternatives, which are accepted and can ensure that the quality of the articles appearing is of a high enough quality, this isn't going to change.