Wikipedia is free. Why can’t scientific journals find a similar model that makes them free not just to the readers but also to the authors?
Let’s consider the differences:
_wikipedia content is provided for free by contributors, this is true for scientific papers as well,
_wikipedia content is not moderated, but moderation/reviewing of science is provided free by academics,
_wikipedia content is formatted by the authors, this did not use to be the case for science but it is certainly possible now,
_wikipedia is a charity and gets donations, this did not use to be the case for science but it should be possible, journals can be charities or run by charities,
_wikipedia only employs a handful of people, this is true of modern electronic journals as well,
_wikipedia is happy to have its size grow exponentially, this is probably not true for journals.
So the only problem I can see would be that a free journal would become victim of its own success and receive too many submissions. Moreover the fees charged to the authors can work as a kind of quality filter, really clever science comes (generally) from well funded institutions. But if this is the real reason that open access journals have fees there should be more openness about it. As a filter it is very coarse and other methods to limit the number of submissions that are sent could be devised. For instance authors may be required to submit the raw data at the same time as the manuscript. This would have the added advantage that a keen reviewer may take a look at the data and discover something different, maybe even more interesting.
So, unless someone tells me which factors I have forgotten, it seems to me that there are no real good reasons why open access journals should charge the authors fees for publishing their work.