Darwin took decades to publish while Newton practically wrote his Principia so as not to have to bother answering questions from other physicists. Throughout science history the attitudes and methods of scientists have varied as dramatically as the personalities.

With the rise of government-controlled science over the last five decades, the need to publish in order to get grants has spiked dramatically - and that has meant a more closed-vest approach.  Official publication is where the money is at.

Yet this decade has seen a rise in technology designed to foster sharing.  Science 2.0 was created to allow for new methods of publication, collaboration, participation and communication and there are various tools like Mendeley and GalaxyZoo making sharing and participation work. One aspect of that still under construction is Open Science or, as Prof. Jean-Claude Bradley calls it, Open Notebook Science.  So far, it is embraced by only a few.  While academia is progressive politically, when it comes to new methodologies they are pretty conservative.  Given a choice between publishing in a subscription journal like Nature or a pay-to-publish journal like BMC, scientists will go for the one with the highest impact factor - quite reasonable, given that careers are at stake.

Yet open science has had a big win - the Human Genome Project has always had an open data policy. So why isn't it done more?  Michael Nielsen writing in the Wall Street Journal says the government has to mandate access - but more government mandates are not likely to help anything without a cultural change because scientists will simply circumvent it; the ClimateGate scandal did not happen because the data was fraudulent, it happened because the scientists were caught conspiring to block peer review for opposing research and avoiding requests to examine their data. As former members of the CRU at East Anglia said, it was more like a cult than a culture.

In the current research climate, sharing any result is a negative; if an experiment is fruitful, and a researcher spent two years on it, their competitor now does not have to spend two years to get there.  If the experiment is a null result, they wasted two years while a competitor does not have to do so - the pioneers are penalized.

Sharing is also not valued at review time, the same as writing on Science 2.0; outreach is considered to be outside the 'real' work of science.

There are only two solutions; one is to migrate back away from government-controlled science. In the golden age of Bell Labs and other basic research efforts, publishing was a source of pride and not a job requirement. The private sector has happily turned over the bulk of basic research to the government because it foists off the cost of failure on millions of taxpayers and they can still buy the winners.  Researchers in academia like the illusion that they are more independent if the government is controlling the money.  Biotech is the only area where that remains unworkable and the private sector does the bulk of the basic research itself to speed up the process.  The other solution is to mandate open sharing of data, just like the NIH did a few years ago.

Open science should have a much easier time than open publishing did; there are no high-powered Democrats getting paid by billion dollar publishing companies to block it. The community should embrace it because it would be egalitarian and thus no pioneers 'getting arrows in the back'.  But to find out, it has to be tried.