In her study of the ideal structure and function of scientific communities, Helen Longino argued in favor of applying the Habermasian ideal democratic speech conditions to scientific communities. Thus, well-structured scientific communities would (should) be structured so that members:
a. are willing to genuinely engage with each other;
b. treat each other as moral, political, social, and intellectual equals; and
c. are willing to modify their beliefs in response to criticism and evidence.
Item c is actually not so hard. Peer review of published work and peer pressure in general exert some force. But how in the world can item b be accomplished?
Though many philosophers of science have written on and built on Longino's work, not much has been done to identify how to nudge real scientific (or our own academic) communities in directions that would meet these ideal conditions.
Atul Gawande suggests that one thing that can improve the communicative equality of communities of professionals working together is to use checklists. Huh? So simple? The bridge between theory and practice is constructed of little practical steps.
From Michael Goldman's review of Gawande's The Checklist Manifesto:
The safe surgery checklist that Gawande helped develop instructs the entire [surgery] team to introduce themselves by name and role and to discuss the unusual aspects of the case and potential problems. The checklist distributes power, so that a nurse reading a checklist acquires the authority, as a member of of the team, to stop the surgeon from omitting a critical step or making a stupid mistake.
Scientific communities do not usually work under the constraints and risk of crisis that surgery teams do, but the larger point is that formalizing working methods and instituting time for face-to-face communication between team members can change the social and power dynamic of collaborative teams.