On Monday, I tuned in via live-stream to an interesting keynote by Geoff Bilder, Director of Strategic Initiatives at CrossRef, who talked about the importance of scholarly cyber-infrastructure (video archive here). Bilder was speaking at Shaking it up – a one day workshop in Boston to discuss “the changing state of the research ecosystem”.
Bilder started by lamenting the lack of importance placed on infrastructure, which is often seen as “unsexy” and goes unnoticed.
https://twitter.com/mrgunn/status/536883907111956481
It can also be seen as an afterthought in many projects, or regarded as outside of the project’s scope:
https://twitter.com/mrgunn/status/536886425439191040
Bilder then proceeded to consider the challenges of implementing centralised infrastructure, specifically mentioning issues of trust:
https://twitter.com/arrust/status/536887467228811266
https://twitter.com/scilib/status/536887462677999616
Wishlist of requirements for scholarly web infrastructure
So what would be the ideal founding principles of new infrastructure? Bilder outlined the following:
1) It should transcend disciplinary silos, geography, institutions
2) There should be non-discriminatory membership of the organisation.
3) The organisation should be non-lobbying – individual members can lobby, but don’t let organisation become involved with things that may become “mission creep”.
https://twitter.com/mrgunn/status/536889048061972480
4) The organisation should be financially sustainable – and preferably create a surplus.
https://twitter.com/mrgunn/status/536889603354296320
5) The organisation should make money from providing services not data.
https://twitter.com/mrgunn/status/536889898729754624
https://twitter.com/mrgunn/status/536890425924411392
6) And finally, as built-in insurance against the organisation “going evil”, the software used should be open source and information deposited as open data so that the entire system could be reproduced elsewhere, if necessary.
From scholarly comms to a science of scicomm
While this event was clearly focused on the scholarly web i.e. sharing academic research data, infrastructure is also something that’s been mentioned with respect to science communication more broadly. Specifically, Bilder’s talk brought to mind a post by Brooke Smith of COMPASS, who earlier this year argued that we need to construct “a metro for science communication“. More recently, Alice Bell has also underlined the need for science communication infrastructure.
So how relevant is Bilder’s list to the scicomm infrastructure discussions? In terms of transcending silos, there have been various conversations about sharing knowledge from different disciplines such as sociology, marketing and the digital humanities. This information exchange has been loosely termed “the science of science communication” and formed the theme for a couple of Sackler symposia held in Washington DC in 2012 and 2013. However, there’s no centralised infrastructure to facilitate that kind of information exchange on an ongoing basis nor, as far as I know, a firm set of new collaborations formed as a result of those meetings.
Have we even reached the point of widely acknowledging that such infrastructure is necessary? Or are we better off relying on smaller personal networks to seed the adoption of new norms via an ongoing process of “refining our craft” – as Dan Kahan outlined recently, using journalism as a specific example.
If creating infrastructure to support the science of science communication is a goal, perhaps it would be most realistic to start by creating a simple, (mostly online) meeting place where people from these different fields can exchange ideas and form new connections, gradually testing out the benefits of working together (rather as Liz Neeley outlines in the latter half of her post here). Maybe then we can scale to considering case studies of successful projects that might be replicated elsewhere in similar scenarios, and then move towards building a larger set of best practice resources. Something more akin to the “applied science of science communication” that John Timmer discussed last year, where we are eventually able to translate tested ideas into standard practice.
(Of course, this skips some of the other items on Bilder’s list – such as membership questions and funding models – things that others have already been pondering).
Social in silico 