The Radical Collaboration of RDA and What It Means for Developing Institutional Data Management Services

The Research Data Alliance (RDA) is an organization dedicated to reducing barriers to data sharing and exchange. 1 While there are many technical barriers that must still be surmounted, it is a core principle of RDA that technical impediments are not the only ones. Often the more challenging barriers are the less visible social roadblocks and those blockades constructed at the intersections of the technical and the social. In my experience in developing and working in institutional data management services

In my personal experience as a member of RDA and co-chair of multiple subgroups, the primary, yet less-noted, obstacles that RDA removes are those between the diverse people who work with or care about research data in some fashion. Whether you are an information technologist, archivist, researcher, scientist, librarian, professor, program or project manager, chief corporate officer, managing director, policy developer or manager, funder, other data professional, otherwise impacted by data, or some magical unicorn-like combination thereof, you will find like minds at RDA in RDA brings people together across experiences, roles, and disciplines in the common cause of sharing research data in a responsible manner that supports the scholarly record of research.
conversation and keeping company with one another. You will find a similar roster of characters as stakeholders in developing institutional data management services.
RDA brings people together across experiences, roles, and disciplines in the common cause of sharing research data in a responsible manner that supports the scholarly record of research. One of RDA's main contributions is the establishment, support, and maintenance of a space where people collaborate to address some of society's grand challenges. As of July 2018, RDA has 93 groups, 2 which have collectively produced over 24 recommendations and outputs, which are "the technical and social infrastructure solutions enabling data sharing, exchange, and interoperability." 3 These products range from Machine Actionable Policy Templates, 4 an information and communications technology (ICT) technical specification, to Repository Audit and Certification Catalogues, 5 a harmonized procedure for certifying repositories, to Wheat Data Interoperability Guidelines, Ontologies and User Cases, 6 an aid to researchers in organizing and communicating their data, to 23 Things: Libraries for Research Data, 7 an overview of research data management resources and tools for librarians. The recommendations and outputs have been adopted or implemented by over 60 organizations, including universities, research centers, repositories, international research efforts, and more. The broad range of the challenges that RDA members address, the solutions they create, and the organizations that implement them speaks to the underlying success of RDA, creating the environs where radical collaboration can occur to address those challenges of research data.
Of course, not all of the current RDA groups, or the groups that have come and gone in the past, have been equally successful in developing solutions that reflect the full strength and participation of group members. Similarly, our institutions are often tasked with addressing society's grand challenges, and we see a comparable variability of success when it comes to developing local solutions for removing barriers to data sharing, in the form of institutional data management services. The intersections between RDA group formation, service development, and the concepts of radical collaboration offer interesting ground for exploring commonalities, differences, and possible roads to success that radical collaboration offers.

The Evolution of Collaboration
An important aspect of collaboration is its evolution. This is true of both RDA groups and institutionally based service development efforts, the differences being the available starting points and the supporting structures provided for the paths forward. Within RDA there is a process for forming groups that often starts with an informal gathering termed a "Birds of a Feather" (BoF), where people with a shared need or idea meet to talk through whether there seems to be sufficient expressed community interest to start coalescing around a shared end goal. The interactions in these BoF meetings vary and account for most of the potential starting points for collaborations. The exception, which I have not encountered in RDA, is the situation that frequently shows up in institutional settings where groups are notified of charges to go forth and develop services or otherwise "do." McGovern identifies interaction types in "Radical Collaboration and Research Data Management: An Introduction," 8 as the quadrants of a matrix, and names them as Radical collaboration, Exclusive interaction, Dominant coordination, and Passive sharing. I have used these quadrants as a foundation, expanding particularly on the nuances of Passive sharing to more accurately represent the variety of starting points that then evolve into collaboration modes: • Inadequate intersections is an interaction signified by lack of interaction; everyone is passive and there is little common ground or shared understanding.
• Notification presumes commonality of purpose but overlooks the necessity of building community through interaction, ultimately shortchanging the strengths and contributions of the potential community.
• Little listening occurs when people have committed to their idea, but not to the community, and have spoken but have not engaged interactively.
These six starting points provide the initial states for the evolution of potential collaboration.

Encouraging Radical Collaboration
Collaboration modes evolve from their starting points. In RDA, this evolution is helped toward the state of radical collaboration by the group proposal process, through which the Technical Advisory Board (TAB) reviews group membership and leadership, and recommends others who should be invited to sit at the table. This process supports balance, one of the core values of RDA, 9 which is equivalent to the radical collaboration concept of having representative places set around the table. Radical collaboration is further encouraged in RDA through the adoption and implementation of other of its core values: openness, consensus, and harmonization. In addition, the communitydriven nature of RDA leaves little room for groups that are run in a style of Dominant coordination. 10 Within an institutional setting, the starting points and motivations for collaboration are slightly different. In particular, groups formed in a Notification style or lacking community will to exist are still expected to perform. This creates a greater burden on both participants and leaders. A role equivalent to what TAB fulfills in the group proposal and refinement process of RDA is often difficult to coordinate at an institution, which may represent a more complex community, and where there is no single group similar to TAB. Lacking the strategic oversight of TAB or a similar group that is focused on inclusion, balance, and processes enabling representation, participants in an institutional setting must be self-conscious in considering with whom they are sharing a collaboration, who else should be involved, and how they will establish an environment that normalizes the behavioral expectations required for radical collaboration.
Institutional leadership also has responsibility in setting the stage for radical collaboration. A parallel to this is laid out in Manges et al.'s extension 11 of Tuckman's group development model, 12 where in the forming stage, leaders may engage in coordinating behaviors that encourage group success such as purposeful team selection, and facilitating team-identified goals and a team-developed shared mental model. Leaders may further promote radical collaboration by employing the principle of the roundtable and considering broadly who should be invited to it. In their facilitating practice, leaders may place special emphasis on the values of inclusivity and inquiry, which are part of the radical collaboration framework put forward by McGovern. 13

Figure 3: Evolution of RDA/community-driven collaboration modes
In both the RDA and institutional settings, groups that work well are ones where it evolves that experience is respected, differences are appreciated, and working together is considered an opportunity to learn, not to proselytize. Groups that don't achieve their potential are ones in which some subset of the participants arrives with a solution in hand, and they are unwilling to consider alternative solutions while other participants display an unwillingness to consider the offered solution in the problem space. This is potentially a case of Exclusive interactions, as elaborated by McGovern. 14 Exclusive interactions can also be detected in how language is used: is it multifaceted, do people explain what a term means to them, is there an effort made to construct cross-understandings? Or is there that built understanding developed out of co-created working definitions, which McGovern 15 points out as an effective tool for radical collaboration?
As a leader or participant in a group in any setting, it is necessary to be sensitive to the interaction types that are occurring and how collaboration modes are evolving. Language use both defines and betrays us. Listening to how a group speaks can reveal how they have defined themselves with the limits they have committed to work within, whether narrow or expansive. Equally, language affects how individuals are perceived in collaboration settings, and may betray our best intentions when it is assumed to indicate our implicit understandings and perspectives. We each have a responsibility within a group to listen with the assumption of good intentions and to speak in ways that can be heard. Coming to a shared use of language is a difficult and important part of forming shared goals, a shared mental model, and a shared commitment to engage in radical collaboration.
In building institutional data management services, language also has an important signaling effect. Given the wide variety of institutional groups that must work together to form a robust service, 16 it is rare that we will ever all mean the same thing when we use words like "repository" or "workflow" or even "data." Taking the time to work through assumptions and having a willingness to adapt understandings and language usage is imperative. Another way to term this is to deploy a caring curiosity: when you care enough to really figure out what it is an individual means and how they understand the challenge. It takes a level of confidence to accept spaces of ambiguity in this radical collaboration process and to work through them in a manner that is respectful and demonstrates hospitality to others' ideas.
Bringing Everyone to the Table   The metaphor of the table, particularly the roundtable, is useful here. If you are taking on the role of establishing a collaboration space, this is equivalent to hosting. A host is responsible for composing a guest list that is varied and inclusive. As the host of a roundtable, you don't do arrange the space so that guests feel welcome at it. As a guest at a collaboration or a dinner party, you participate, but you don't dominate. Hospitality is demonstrated by the host and engaged with by the guests. Just as with dinner parties, so too with radical collaboration, the concept of hospitality is paramount.
It is in this generosity of spirit that radical collaboration is found, and that efforts in forming institutional services can learn from organizations like RDA that promote conscientious efforts of inclusion, balance, and openness. Despite differences in how efforts are initiated, there is a common motivation to solve problems, find solutions, and address grand challenges. Coming together around these motivations with a sense of generosity and hospitality, which is exemplified by openness, consensus, harmonization, and balance, creates the space for radical collaboration and may provide the setting for some truly extraordinary meals. 3. "Adoption Stories," Research Data Alliance, accessed December 4, 2018, https://www.rd-alliance.org/recommendations-outputs/ adoption-stories.

Coming together around [shared]
motivations with a sense of generosity and hospitality, which is exemplified by openness, consensus, harmonization, and balance, creates the space for radical collaboration….
sit at the head of it, but you