Tidbit from today's meeting with a senior editor of the Lancet:

The major source of revenue of big journals is reprints, not subscriptions nor ads. Publish a clinical trial and drug companies will pay ~$2m for 100,000 reprints to distribute to doctors. Aside from the issue of bias this raises, which I'm reasonably confident they handle properly, is the copyright issue. This is probably good news for open access. The core insight which led to the GPL-- that copyright can be used to liberate works as well as constrain them--here could be used to allow open access to the public but restrict redistribution rights so that they still have most of their revenue stream.

Mentioned PLoS has good lay summaries (apparently there's resistance there because BMJ does them), which she thought was great. Didn't get a very satisfying answer on the open access question, however. Carving out open access exemptions for specific fields is not enough.

Also had a good talk about why science writing is so bad. It was most gratifying to learn that they edit heavily and detest the passive voice. I'll have to go check out some articles later to see whether that comes out in the final product.

--Ari