The Open Scholarship Initiative

Goals, Approach & Methods


Exactly what is OSI advocating?  It’s open—not particular open agendas or specific solutions (yet), but a robust, realistic, sustainable framework for moving open forward, which means being able to discuss issues and options as a community, collaborate on efforts, and adapt solutions and approaches the community can get behind and help grow. The OSI2016 report wanders into the weeds a bit more about what OSI is trying to accomplish and why, and also describes what OSI2016 was able to accomplish. Specifically, the goal of OSI as described in this report, is “to build a sustainable, robust framework for direct communication and cooperation among nations, universities, researchers, publishers, funding organizations, scholarly societies, libraries, policy makers, and other scholarly publishing stakeholders, in order to shape the future of scholarly communication, beginning with scholarly publishing and the issues that surround it, to support a climate for finding common understanding and workable solutions and to help this stakeholder community move toward these solutions together.“ The eventual outcomes of this effort will include:

  • Achieving open goals faster and on a more predictable trajectory by bringing all stakeholders to the same side of the table to push together toward their common goals (while continuing to work out their differences on tangential issues),
  • Creating multiple platforms for working on scholarly communication improvements together as a broad stakeholder community (these platforms will expand as OSI’s ability to collaborate and communicate increases),
  • Increasing the efficiency and effectiveness of stakeholders by facilitating the development of a common roadmap of goals, policies, and standards in scholarly communication, and
  • In the end, increasing the amount of science information available to the world and the number of people who can access this information.


Apart from annual meetings and reports, how exactly does this group plan to achieve these lofty outcomes? The short answer is not in one step. Because the OSI community is coming at this issue from so many different directions, the best strategy for encouraging more open is to acknowledge and value where each stakeholder group is in the process and then figure out what course adjustments can be made to the system to continue to expand open and what assistance this community can offer—new incentives, coordinated policies, collaboration efforts, formal partnerships, new studies, pilot products, expanded perceptions, and so on—to help flip to a more open mindset going forward and thereby accelerate the growth rate of open publishing and also increase the volume of scholarship available in open format.

As the OSI2016 report notes, “No single actor in a multi-stakeholder system like this can enact system-wide change unilaterally; a mechanism for collaborative action needs to exist but it doesn’t currently exist in scholarly communication on a broad scale.” Therefore, OSI has been designed to work on this change collaboratively and deliberatively, in a way that involves input from all stakeholders in the scholarly communications community, and always with an acute awareness that the new world of scholarly communications being designed needs to be accepted by the research community and be of benefit to this community, needs to work in every country, institution and field of study, and needs to be reliable and effective over the long term.

It is at these intersections of idealism and reality, of open knowledge and intellectual property, and of politics and policy that OSI’s most important work will be done—-determining the best balance point between embargoes and immediate release, designing data repositories that scientists actually want and will use, curbing the unintended consequences of publish or perish without dismissing the importance of publishing in academia, improving access to scholarship for underserved regions of the world without unintentionally making the access problem worse, and more.

The OSI report notes that scholarly communication is changing and that this change presents opportunities and challenges. Some of the change that is happening involves shaking up the current system to utilize publishing tools and approaches that may be better suited to an Internet-based information world. But not all current and needed changes fall into this category. Indeed, some of the most needed changes do not. The general guidelines for action as defined by the OSI2016 group are therefore as follows (with the specific recommendations contained in workgroup reports):

  • We don’t have a clear, coordinated action plan for improving open. What needs to happen today, tomorrow and the day after? Who are the actors, what are the mileposts, what are the likely impacts, and how do we measure success? (Note that these concerns don’t necessary suggest that OSI itself should create and evaluate specific programs of work. Rather, this is a commentary on the need for OSI to identify what it can do and how it will operate, and then farther down the road, what kinds of synergies OSI can encourage.)
  • Some change will need to involve reforming the communications culture inside academia, where old publishing methods, measures and perceptions can drive author choices and be used as proxies for merit when evaluating grant awards and tenure decisions. And some will need to involve examining our own biases that publishing is a binary proposition involving either open or closed, subscription or APC-based, right or wrong. Open, impact, author choices, peer review and other key concepts all exhibit a range of values. Identifying non-binary measures for some of these values (as proposed by several workgroups) may be helpful insofar as allowing stakeholders to focus on improving areas most in need of change and comparing progress and best practices across disciplines, institutions, publishing approaches, funders and so on.
  • Any widespread change is going to require a widespread effort. There are simply too many stakeholders with different interests and perspectives who influence different decision points. No single stakeholder or group will be able to affect this kind of change unilaterally.
  • How do we make these reforms in response to the needs and concerns of authors rather than in spite of authors (authors are not a homogenous group with common interests or opinions, of course, but there was some sense among delegates that reform efforts could be better attuned to what authors needed)?
  • How do we make changes across disciplines (which have different needs) and that also effectively build on the efforts of the many stakeholders in this space?
  • How do we reform the system without losing its benefits?
  • How do we move from simply repairing dysfunction to creating a more ideal publishing world and reaping the benefits that such a world could provide in terms of participation, efficacy, efficiency, and discovery?
  • Developing standards and norms would be helpful as we move forward, as well as answers to a number of key questions.


The two most important communication instruments OSI has been using in this process so far are:

  1. Listserv conversations. The OSI listserv was a robust tool in 2016 for tossing around ideas and will continue to play an important role in 2017. When engaging in this forum, delegates try to abide by a code of conduct that respects that everyone in OSI is committed to working on scholarly communication issues together. As members of this select group, the role of delegates is to contribute facts, share perspectives and experiences, and to the extent possible, help bridge the gap with fellow OSI colleagues by trying to see the big picture, find common ground, propose new solutions, and debate ideas. “What if” and “why not” discussions will be increasingly important in 2017.
  2. Workgroups. OSI2017 may end up with 20 different workgroups focusing on a wide variety of issues such as studies, peer review, impact factors, and information overload (see Workgroup members are the sherpas who guide OSI  through particular issues and challenges and recommend solutions for consideration. Last year, OSI2016 delegates worked exclusively with one workgroup team to come up with the papers published on the OSI website at These papers were an opening volley of cross-stakeholder group collaboration and hit on familiar points in some sections, and broke new ground in others. The next phase of workgroup recommendations will begin to combine analysis with consensus and action plans for full-group consideration. Participating in workgroups is an opportunity open to all OSI members and not just meeting delegates.

Other communication instruments may evolve over the coming year (the OSI2016 group talked about forming tiger teams, for instance—maybe geographically close groups of delegates who could talk about OSI at meetings and institutions), but for now, the listserv and workgroups are OSI’s two main drivers for action.

Going forward, OSI planning groups have put forward a draft governance plan which will be discussed at the upcoming meeting. Also to be discussed will be the strategic recommendations from OSI2016 delegates to create a united front for OSI by taking actions such as:

  1. Sign a statement of purpose affirming that all OSI delegates are working toward a common goal of making research information more accessible. A draft of this statement was prepared last year but the general sentiment was that most people weren’t quite ready to take this step.
  2. Endorse a pluralistic approach to these challenges. OSI doesn’t necessarily need to endorse a specific solution, for instance. Rather, it would help as a first step to simply acknowledge that many solutions exist on the same open spectrum and that we’re all working toward the same end goal of more openness. By embracing all of these efforts, taking credit for them as a community, and then continuing to onboard new ideas and efforts, we can energize this space and make it a safer place for others to join. There are innovation advantages to this approach as well— to having a pluralistic approach, that is. As many delegates have noted, we’re not entirely positive which way is up right now and having a broad approach will allow the market to find the best solutions—not all gold, or all green, but somewhere in between. Second, few universities have wanted to be the first to pioneer new forms of scholarly communication—kudos to Harvard and others for taking the risks here. But over time (as Susan Fitzpatrick has noted), few universities will want to be the last to adopt a new all-encompassing ethos of open we’re promoting, with an array of stable and well-populated pathways that all tilt toward open, and that over time can continue to tilt more in this direction as policies and practices evolve. If we can push this approach to a critical mass of acceptance, it can reach a tipping point (as Ivy Anderson has mentioned) where it clearly makes sense to participate—where the benefits of participation clearly outweigh the risks.
  3. Endorse a collaborative approach. Change the tenor of this effort to one where we’re all working together instead of at cross-purposes.
  4. Keep building the case for open (it’s not as strong as it needs to be to convince skeptical policy makers) by supporting and/or conducting studies, and otherwise trying to improve the evidentiary base.
  5. Lay the foundation for a common vocabulary. For instance, help ensure that funding rules don’t discriminate against any of a rainbow of open options and are consistently applied internationally, and make sure we’re all talking about the same thing (OA still means different things to different people, even when a definition of open is included in a glossary).
  6. Find inroads to reforming systemic issues, from the commoditization of PhDs to publishing fraud to publish or perish hysteria and a hundred other issues in between.
  7. Formally get behind appropriate efforts and enhance working relationships with other groups in this space so we can collaborate on efforts and strengthen our approach and outcomes.

OSI2016 Summary Report


Why is collaboration needed? What proof is there that collaboration will succeed? On the one hand, it’s clear to many people who have followed the changes happening is scholarly publishing over the years that much tension and uncertainty currently exists. To this, OSI delegate Rick Anderson noted in a recent OSI listserv conversation that “All of us have an imperfect understanding of ‘the bigger picture,’ and we should…try hard both to listen to and to learn from the perspectives of those who spend most of their time working in a different part of the system than we do.”

Having a forum where issues can be discussed that reach across stakeholder groups is critical, as it is with many other societal concerns. Imagine OSI’s approach to improving scholarly publishing as being akin to auto manufacturers needing to establish common standards, or environmental regulators working toward common goals with a wide variety of stakeholders in the private sector, state and local governments, and federal and international governments.

In scholarly publishing, a variety of independent stakeholders are independently working to create a similar class of products that should ideally be interoperable and that have significance to society—the production of knowledge of consequence to medical research, industry, environmental protection, and so on, using public money in most cases. This information isn’t entertainment, nor is the type of information we’re likely to easily find in newspapers or online (without access privileges), but research that we’ve invested in, that we monitor, and from which we increasingly expect to receive a return on investment. And in the production of this good, we have no universal guidelines—no coordinating body that says how it will be done, where it will be stored and preserved, how it can (or can’t) be used, and so on.

Ensuring that this process has reasonable guidelines that protect the benefits owed to society is the best way to protect the outputs from this system.

So to create these guidelines—or at least to begin having this conversation—we need to create some kind of working group, some kind of representative body or forum that can work toward developing a system of joint responsibility for its proper care and development.

Please see the OSI2016 report (linked above) for more FAQs.