FROM NILS HANWAHR IN THE SCIENCE POLICY CENTRE
We kicked off our Science as a public enterprise study with a public event at the Southbank Centre last month. The full recording of the first debate is now available.
(Apologies for the mixed quality – the audience mikes were not picking up well that day.)
We have also distilled some of our highlights:
Why should Science be open? – introduced by Paul Nurse (President of the Royal Society), with a speech from Mark Walport (Director of the Wellcome Trust), and a panel discussion chaired by Geoffrey Boulton FRS (Chair of Science as a Public Enterprise working group).
03:00 min. Paul Nurse introduces the event and the project:
- A fundamental characteristic of a free and enlightened society is that knowledge should be open.
- Science needs to check if its institutions are as effective and as open as they should be, concerning both the sharing of information between scientists as well as how the public engages with scientific information.
- However, do we want complete and open access to all stages of the scientific process; and what about concerns about consent, privacy and the control of how information is being used?
08:45 min. Mark Walport, a member of the Science as a public enterprise working group, lays out the steps towards more open practices, as well as some of the reasons to be cautious.
- Open access publishing provides free content to all users online, including supplementary material and some data that relates a specific research paper.
- The next step is to share observations and scientific raw data at a large scale. This would allow other scientists or the public to use existing datasets to reproduce, check and understand the conclusions of a research paper.
- Ultimately, one might aim to open up the entire process of experimentation to voluntary collaborators. In this case, it would not only be raw data that is shared, but all aspects of the scientific process, such as emails and draft papers, as well as models, methods and codes.
- The original argument for open access publication is that openness for reuse and replication and testing lie at the heart of the scientific method.
- However, openness does not automatically entail equality. Walport reminds us that on a global level, free trade does not equal fair trade. In open science, research from developed nations might feed on data from developing states that do not have the necessary computing power.
- Furthermore, citizens increasingly demand transparent evidence underlying complex policy decisions. Science should be accountable to the public as its source of funding.
20:10 min. Geoffrey Boulton introduces the panel discussion on why should science be open.
- Science is about understanding the cosmos around us and our position in it and thus science is a public enterprise and not just an instrument of economic growth.
24:45 min. David Dobbs, freelance science writer, praises the rise of open access publishing in the last years.
- The scientific paper has become a ‘bottleneck’ of big data rather than a simple means of communication between researchers.
27:00 min. William Dutton, Oxford Internet Institute, voices concern that the focus on the publishing of data is too narrow. Democratising science is not just about publishing.
- Open science is essential for academic freedom, but there need to be incentives, when it makes sense for the quality of the research.
30:20 min. Stephen Emmott, Microsoft Research, points out that there is a difference between data and the models employed to analyse it.
- Science is fundamentally about reproducibility and falsifiability. However, model-based research is difficult to repeat without knowing the code and model assumptions. Open science needs to be more than open data.
34:00 min. Discussion with the audience. There were a number of challenging questions and interesting comments from the audience. Here are some of the main points:
- Should someone regulate interpretation of open data which has been misrepresented by others?
- What about data curation and appropriate metadata? Who will spend the time and money to provide the extra services open data may require?
- International collaboration and data sharing across borders can be crucial to address international problems such as water security or environmental issues.
- Public trust depends on confidence in the process of open science, which will be based on transparency.
Twitter discussion during the event was also quite active – here is the TwapperKeeper archive of #openscience for the day of and day after the meeting: http://twapperkeeper.com/hashtag/openscience?sm=6&sd=7&sy=2011&shh=00&smm=00&em=6&ed=8&ey=2011&ehh=23&emm=00&o=&l=25000&from_user=&text=&lang
The call for evidence for the Science as a public enterprise study is still open. We welcome contributions until 5 August. For details and an online questionnaire go to: royalsociety.org/policy/sape/call-for-evidence/