Reinventing science? From open source to open science

Community member post by Alexey Voinov

alexey voinov
Alexey Voinov (biography)

Science is getting increasingly bureaucratized, more and more driven by metrics and indices, which have very little to do with the actual scientific content and recognition among peers. This is actively supported by the still dominant for-profit publication mechanism, which harvests products of scientific research for free, processes, reviews and edits them using voluntary work of scientists themselves and then sells the resulting papers back to the scientific community at obscene costs. The original ideals of scientific pursuit of truth for the sake of the betterment of humanity are diluted and forfeited in the exhausting race for grants, tenure, patents, citations and nominations. Something has to change, especially in the era of post-normal science when so much is at stake, and so little is actually done to address the mounting problems of the environment and society.

In computer programming open source emerged in the 1980s largely in opposition to attempts at licensing code and the growing dominance of Windows with the annoyingly secretive policies of Microsoft. It came as the idealistic philosophy of software development that stems from the so-called “gift culture” and “gift economy” based on this culture. Under gift culture you gain status and reputation, not by possessing things or ideas other people want, not by occupying a dominant or special position, but rather by of giving your time, your creativity, and the results of your skill. Some of the best known off-springs of the open source paradigm are the first web browsers that helped the Internet to happen (Mosaic, Apache, Mozilla, etc.) and the Linux operating system.

Gift culture, and the dedicated protection of a free intellectual commons, is a powerful incentive that seems to be most appropriate for the scientific pursuit in general, not just computer engineering and programing. While some federal and international agencies already require open source delivery of code and data as a funding prerequisite, there is also a growing trend towards ‘open access’ for the scientific publications that result from federally funded research.

We need to promote a shift in the traditional reward system for scientific research. While individual efforts and talent drive science, we have reached a point where team efforts are required for most breakthrough achievements. While gift culture is inherent to the scientific community, it has been significantly compromised by privatization of science and severe competition for diminishing resources. It is paramount to reaffirm our commitment to the fundamental precepts of science and encourage the rebuilding of a culture of collaboration and information sharing.

However the shift to ‘open science’ that would be based on the same paradigms as ‘open source’, and would support transparency and inclusiveness, not just for the results of research but for the research process itself, is yet to be achieved. It will be hard to adopt this approach due to competing pressures or simply excuses that results are intellectual property, the student has not graduated yet, the paper is not yet accepted, code is clumsy, documentation is not complete, or that a few more tests or fixes need to be made…. Academic institutions are also to blame, as they look to the commercialization and patenting of faculty-conducted research as source of income to enhance University revenues.

It would also be helpful if we could shift some emphasis away from peer-reviewed publications toward less traditional products like well-documented models, data and code. More appreciation should be given to the reviewers themselves by acknowledging their efforts and giving them credit. Journals should encourage open, collaborative reviewing rather than blind reviews that breed competition and suppression of new ideas. There is an obvious need for new award and credit systems that would stimulate sharing and teamwork rather than direct personal gain. Again there is much to learn from the experience in licensing that is available from the open source community.

The development of appropriate supporting infrastructure, licenses and standards can certainly affect the collaborative behavior of participants. There is evidence for this in how infrastructure for user-generated content has revolutionized development and access to information through social networks, in efforts like Wikipedia and Myspace. Here again there is much to learn from the open source community. At the same time there is clear evidence that the reward system for research is already changing. The Nobel Prize awarded to the Intergovernmental Panel on Climate Change (IPCC) team is probably the first time such a large group was acknowledged, and is a tangible indicator that some challenging science questions are now inaccessible without huge teams that work in collaboration.

Funding agencies could certainly help by further supporting and encouraging community research, but this may take a long time. The model adopted by SESYNC is unique and holds much promise. It already largely depends on voluntary contributions of scientists, who do all the work on their own, at their own cost, with support provided only for communication – remotely through web services, and in-person during a small number of meetings with travel paid for.

What other initiatives are out there that could be built on?

The post is based on some ideas that previously appeared in:
Voinov, A.A., et al., (2010). A Community Approach to Earth Systems Modeling. EOS, Transactions. American Geophysical Union, 91, 13: 117–124.

Voinov, Alexey. (2008). Systems Science and Modeling for Ecological Economics. Academic Press (Chapter 9).

Biography: Alexey Voinov PhD is Professor of Spatio-Temporal Systems Modeling for Sustainability Science at the University of Twente Faculty for Geo-information Science and Earth Observation in the Netherlands. His academic and teaching interests evolve around spatial dynamic modeling of ecosystems and sustainability science in application to decision support and policy making. In particular he is interested in integrated modeling and participatory modeling, integrated assessment, systems analysis in ecology and economics, energy and natural resources, with applications in watershed management, agroecology, energy policy. He is a keen advocate of stakeholder involvement in modeling and decision making. He is a Principal Investigator of the Participatory Modeling Pursuit funded by the National Socio-Environmental Synthesis Center (SESYNC).

5 thoughts on “Reinventing science? From open source to open science

  1. I appreciate the perspective, and wholeheartedly agree with trying to maintain a gift culture in science, promoting a freely accessible information commons. It is definitely useful to be mindful of how the things we do fit with the culture that we want to encourage.

    However, I do have a few comments/questions:

    – How would you deal with tension between cultural and economic incentives? It makes sense to do things that promote the culture you want to live in, but more basic levels of Maslow’s hierarchy of needs must still be addressed. Does maintaining a gift culture require a gift economy (and therefore cultural change), or can the incentives of a different economy be aligned? Commercial work based on open source seems to indicate that commons can actually provide a competitive edge e.g. with scale and network effects. But would that still be compatible with your vision?

    – Rather than condemning bureaucracy and metrics, I would suggest that their potential is simply not yet being realised. For example, citations do, in principle, capture something meaningful in being a visible expression of support. From that point of view, the metrics just need some work…?

    – I think anonymity still has an important role to play in peer review. It encourages comments to be taken at face value rather than being linked to the reputation of the reviewer. It encourages (sometimes brutal) honesty by reducing the risk of social embarrassment. Clearly, though, we do have problems with reviewer accountability, associated with unconstructive and impolite reviews. As I also mentioned on Gabriele’s post
    ( https://i2insights.org/2016/10/11/methods-section-purpose/ ), I think the way forward is for more dialogue about what is expected of reviewers – confidential feedback from editors, anonymous feedback from authors, and visible debate within a disciplinary community about what is good and bad reviewer (and author) behaviour. But the current context does seem to be that just because reviewers give their time freely, they are the ones in a position of power… What would a gift culture perspective say about this issue?

    • – How would you deal with tension between cultural and economic incentives?
      * I do not see this as an ‘either-or’ relationship. Like you point out, commercial work embraces and benefits from open source software. Similarly you can also assume commercial applications based on scientific research. Do you do your science to be rich? Probably not. I bet most of good scientists would be much better off financially working for business.

      – For example, citations do, in principle, capture something meaningful in being a visible expression of support.
      * Support? Not necessary. You also have to cite papers when you totally disagree with them.

      – I think anonymity still has an important role to play in peer review.
      * I’m not saying that anonymity should be totally excluded. I’m suggesting that the default option should be an authored review. Of course there are situations when you may be putting your career at risk by writing a negative review. But in most cases there is really nothing to be afraid of. On the contrary, it will only make reviewers more responsible for what they are saying, Totally agree with the idea of more feedback to reviewers. Name disclosure could also help, allowing authors to directly communicate with reviewers to clarify certain aspects and perhaps even ask for help improving the papers. Of course with due acknowledgement of reviewers input. Up to the point of co-authorship in some cases, as we do in EMS.

      • Thanks for the reply. I would like to clarify a few points:
        * I’m not talking about scientists becoming rich, but rather maintaining livelihoods. The competition for resources is such that academia does not always provide a reliable livelihood. While I do agree that open and commercial science can be made to be compatible, I don’t think it necessarily comes naturally.
        * I mean “support” in a social rather than agreement sense – citing a paper you disagree with effectively legitimises their involvement in a debate. This also illustrates that distinction of types of citation is important, hence why efforts to identify “influential citations” like https://www.semanticscholar.org are interesting.
        * Good to hear your optimistic yet pragmatic view on review. Let’s see what we can do!

  2. There is also a great TED interview with Linus Torvalds about the importance of open source. It’s called Linus Torvalds: the man behind Linux and was posted in April 2016.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s