Michael Nielsen, speaking at our Policylab earlier this month, described two revolutions in open science – two periods of significant change in the way scientists share discoveries.
The first revolution in the 17th century went hand in hand with the formalisation of the scientific method and the formation of the Royal Society. In just over a century, science went from secretive basement alchemy to peer-reviewed journals and public demonstrations of experiments.
Today, the proliferation of huge datasets, models and software in scientific research require scientists to rethink how they share their work. (See previous post from the launch of our Science as a public enterprise study.) For Nielsen, this is the beginning of a second open science revolution and an imperative for updating tools for sharing research practices and findings.
There are currently very specific incentives for scientists to share their work. Nielsen:
“Scientists are busy and they are very invested in sharing whatever knowledge they have through conventional channels – i.e. much of what we do is orientated toward journal publication. And while we have a reward system that is tied to publication in that way, the amount of experimentation that is going on with these sorts of ideas, like the Qwiki, is going to be extremely limited.”
[Download the full recording of the event. This statement comes 23:08 minute in.]
Qwiki is a quantum physics wiki set up to support sharing in that research community. It failed to catch on because, Nielsen argued, there is no conventional reward for taking part: “it’s much better to go to your tenure committee and say I wrote a single mediocre paper that no one is ever going to read.” Bankers are not interested in bonuses in monopoly money. Nor are scientists necessarily interested in something where the reward is not tied to their career.
A healthy transition to more open sharing practices can only happen when there are incentives for researchers to adopt this greater transparency. There needs to be rewards for participating in initiatives like Qwiki or providing full datasets with a research paper. In today’s data-heavy science, data are the bread-and–butter of many disciplines. Publishes articles are the artefacts pointing towards this underlying science. Research published earlier this month, showed that only 9% of submission to top journals provide online access to primary data sets. Sharing via the journal – the product of the first revolution in openness – is not, as it stands, enough. It does not reflect the full spectrum of scientific activity.
The Government announced last week that they are setting up an expert group to examine how access to research findings can be made easier and more transparent. But the Department for Business, Innovation and Skills drew a clear boundary round the area they are looking at:
“Focussing on academic publications, specifically journal articles, conference proceedings and monographs, the working group will take in to account parallel work relating to research data and other outputs being conducted by the Royal Society.”
There is still a lot to do to ensure access to journal articles produced from publicly-funded research. But the debate about open access has at least one foot in the past.
The bigger changes afoot are illustrated beautifully in Cameron Neylon’s ‘Story of the Future’ in an upcoming article . A young researcher works in an international, integrated education and research environment, where he will be rewarded for a career in data management as well as primary research, and he is already analysing data while at school.
The challenge for Science as a public enterprise is to set out the incentives for researchers that could make a future like Neylon imagines come true. This is no easy task. But one very obvious way to fail would be supporting only incentives for conventional – eg. journal – sharing mechanisms, with blinkers to the unconventional, like the Qwiki: especially when unconventional openness today may be a better convention for the future.