5

Convergence as Practiced

The panel discussion summarized in this section brought together experts who have studied and overseen collaborations in a variety of settings and can provide their perspectives on what the National Center for Science and Engineering Statistics (NCSES) may usefully measure. Lora Weiss (Pennsylvania State University), the first panelist, discussed some implications of convergence. To recognize and remind others about the importance of continuing work on convergence technologies, she began by quoting the National Strategy for Critical and Emerging Technologies (C&ET) October 2020 report, which stated that “[i]n fact, many technology breakthroughs occur at the intersection of two or more disparate technologies,”¹ The National Strategy for C&ET emphasized interdisciplinary work and the need for academic, public, and private partnerships to address societal challenges.

The Journal of Higher Education recently published an article² questioning whether hiring initiatives at U.S. research institutions that focus on interdisciplinary clusters represent an ideal rather than delivering actual collaborative relationships. Commenting on the authors’ highlighting of Penn State as an exemplar institution, Weiss explained that the interdisciplinary research institutes at Penn State were established based on existing and genuine partnerships between interdisciplinary institute leaders and deans

___________________

¹ The White House. (2020). National Strategy for Critical and Emerging Technologies. Available: https://www.whitehouse.gov/wp-content/uploads/2020/10/National-Strategy-for-CET.pdf.

² Bloom, Q., M. Curran, and S. Brint. (2020). Interdisciplinary cluster hiring initiative in U.S. research universities: More straw than bricks? The Journal of Higher Education 91(5):755–780. doi:10.1080/00221546.2019.1688615.

of academic colleges, all of whom shared expectations and a vision for the center. This shared outlook, Weiss noted, is crucial: If researchers remain within a single discipline, interdisciplinary work cannot function.

The institute directors are furthermore top scholars in their area of expertise, which helps in the recruitment of junior hires. Incoming researchers are more likely to be interested in interdisciplinary work when they know that the director has experienced the tenure process and understands the experiences of junior faculty and that such work will not harm their promotion paths.

Each of Penn State’s core research facilities is staffed by research professionals, which constitutes an important component of the institution’s success. These dedicated staff allow researchers to focus on their research rather than become absorbed by activities such as maintaining and operating equipment. The final element of Penn State’s success is its seed grants. The university invests in high-risk, high-payoff collaborations that lead research in new directions and help to foster new partnerships. Roughly half of the projects receiving Penn State seed grants receive external funding. Weiss cautioned that if this ratio were to approach 100 percent, it would signal that the supported projects are not taking enough risks.

To demonstrate the range of interdisciplinary collaboration at Penn State, Weiss presented a diagram that shows all internal Penn State collaborations (spanning 24 campuses and other extensions) that have generated at least 15 shared publications. The university’s external partnerships further extend its collaborations and allow diverse perspectives to flourish through interactions and relationships that Weiss and colleagues constantly fine-tune.

In his talk, Peter Schiffer (Yale University) spoke about administrative structures and incentives that encourage convergence. He noted that measures of convergence outcomes are difficult because they try to measure ideas and impacts that are brand new, requiring researchers to work to measure things almost before they exist. On the other hand, he observed that convergence can be fostered when an institution’s culture allows it, which depends on administrative structures, policies, and incentives. Therefore, structures and incentives within an institution may be surveyed to provide indirect insight into whether an institution is creating an environment to promote and support convergence.

Schiffer described several specific administrative structures that create such an environment. First, support for diversity, equity, and inclusion in research is critical because diverse teams produce better outcomes, particularly in areas that do not represent a single, well-defined discipline. Large institutions and centers that reside outside of disciplinary centers and allow people to collaborate across disciplines are also important. Collaborations can be promoted even further through campuswide events,

talks, or research sessions that informally bring together researchers from different fields and facilitate social and professional interaction. Schiffer reiterated Weiss’s observation that professionally staffed, centrally managed core facilities make it easier for people to meet and collaborate using the same instruments to develop new ideas.

Funding mechanisms are also important. A critical mechanism is an accounting system that effectively manages funds across multiple units. If a single person in one department controls all funds, interdisciplinary teams may find it difficult to work collaboratively. Universities can similarly help build a fundraising arm for transdisciplinary work that can help start or seed new projects. Traditional university fundraising is ill-equipped for this work because it is organized around alumni. Transdisciplinary convergent efforts typically do not have alumni because they do not grant degrees and are relatively new. Movement toward convergence is facilitated by offering interdisciplinary degrees to students. Degree programs can bring students together across disciplines beyond the obvious connections between biology and chemistry or between physics and math. Students can also create bridges that join faculty across disciplines.

Beyond structure, Schiffer outlined policies that support convergent research. Tenure and promotion policies are a particularly difficult case because most universities depend on external reviewers selected for their disciplinary expertise. Convergent work can be recognized by explicit policies in the tenure and promotion process. Graduate students should be able to work with different or multiple advisors, regardless of their admission department. Postdoctoral students and research scientists should likewise be able to hold appointments that cross units. Internal funding policies can also support convergent research by ensuring that individuals are not forced to compete with each other for things like research incentive funds associated with indirect cost reimbursements, but instead are rewarded when convergent activities take place. Finally, Schiffer addressed incentives, particularly for faculty members. If internal, generous, and stable funds support convergent research, then faculty will pursue convergent research opportunities. Offering faculty discretionary funds to investigate their newest ideas will provide creative researchers with the opportunity to try new things in a manner not directly dictated by a dean or department chair. Also important is dedicated staff to assist in proposal writing and grant management. Public relations activities that celebrate the productive crossing of disciplinary boundaries can also incentivize researchers, particularly when participating researchers are recognized instead of just specific departments.

Schiffer ended his talk by adding some caveats to his suggestions. First, research on the ideas presented continues to evolve and should be studied over time. Even successful ideas need a supportive institutional culture to

flourish. If the leadership culture does not support convergent research, it will not happen regardless of incentives, structures, and policies. He acknowledged that many of his suggestions are aimed at larger institutions and single institutions. Tailored approaches may be required for smaller schools and an interinstitutional convergence measure may be needed to capture activities occurring across institutions. Finally, Schiffer cautioned that not all research should be convergent. Convergence depends on disciplinary advances, and incentives and policies that promote convergence should not disadvantage researchers who work best within their specific discipline and whose work may ultimately lead to a disciplinary breakthrough or a convergent effort.

Dan Gallahan (National Cancer Institute) opened by stating this workshop has underscored the importance of understanding how we do science. As a funder of science in the Division of Cancer Biology at the National Cancer Institute (NCI), he has seen billions of dollars toward basic cancer research move through his office. At various points in time, he has used all the different terms spoken during the workshop: team science, integrated science, interdisciplinary, transdisciplinary, and others. Unfortunately, many of the research efforts launched before these terms were fully defined, and though they are related, they do differ in meaning.

Based on his background in cancer research, he noted that several aspects of convergence in science relate to cancer and our scientific understanding of it. Gallahan said that cancer incidence and mortality rates have remained steady. About 20 years ago, NCI realized many of its efforts did not hold or address the same view of cancer as other research communities. New approaches to cancer research were needed, which incentivized an incorporation of the ideas of convergence, team sciences, and integrative biology into current research. These approaches increase the variety of perspectives and abilities available for use in investigating imaging and the different scales involved in cancer research.

Gallahan stated that NCI has established two programs to broaden perspectives on cancer: the Physical Science Oncology Network and the Cancer Systems Biology Consortium. Gallahan explained that multiple viewpoints can come together through a combination of enhanced insight into a topic and efficient problem-solving skills. To begin this work, NCI convened meetings of investigators, physicists, engineers, and mathematicians to share their perspectives about the complex problems that exist in cancer research and ways which these problems can be addressed. The aim of the Physical Science Oncology Network is to identify new ways to view these problems rather than to solve these problems outright. The Cancer Systems Biology Consortium brings together a different set of communities that focus on computational approaches to enhance understanding of cancer’s development and to model the body’s immune response to cancerous cells.

This goal requires input and expertise in modeling and computational biology as well as in technology.

Both programs were very successful beyond the standard bibliometric measures of output. Each project involved scientists from multiple disciplines and required (based on proposal requirements) an educational and outreach component. Investigators were retrained to ensure that they possessed the vocabulary necessary to communicate across disciplines. Each program also ensured that a new generation of researchers entered the field.

Gallahan emphasized a need for critical evaluation of the programs to determine their success and future directions through engagement of an external expert panel and extensive quantitative and qualitative data collection. Evaluation should consider the diversity of publications, the extent to which the principal investigators represent multiple disciplines, the use of new terminology, and novel grant applications (incorporating new investigators with applications in new areas). Another measure of success could be the development of new programs such as the spread of computational modeling at the National Institutes of Health (NIH), which was almost nonexistent at NIH only 10 years ago. Gallahan is encouraged by the training that has been developed for young investigators and the high quality of cross-disciplinary collaboration by those new investigators.

Maryann Feldman (University of North Carolina at Chapel Hill) was the final panelist of the session and used her time to represent the research done by a group of scholars who were formerly known as the Science of Science and Innovation Policy (SciSIP) Program, and now reorganized as the Science of Science Discovery, Communication, and Impact (SoS:DCI). The group seeks to understand how knowledge is created and transformed into socially and economically useful products. Feldman herself is most interested in this work’s spatial dimension, an interest that springs from her background in economic geography. Economic geography itself is a convergent topic that investigates the spatial organization of economic activity as well as complex relationships and emerging activity patterns within a place.

A great deal of work is being done on the emergence of new academic disciplines, varieties in organizations, and the creation of entirely new industries. The idea of an ecosystem is key to this understanding, and it involves the set of institutions and constituent factors that enables creativity and legitimacy. The question is the emergence of something distinct and new, even as it blends other existing types of expertise.

Feldman stated that the emergence of new academic disciplines and departments is a clear signal of legitimacy. An academic discipline requires accreditation, which is a formalized professional process. Feldman’s group investigated the emergence of biomedical engineering as a distinct discipline,

and preliminary findings suggest that a university requires a strong medical school as well as an engineering school, combining these strengths with the ability to bring in large research grants. Unfortunately, if the university is a Historically Black College or University (HBCU), then the probability of a biomedical engineering department’s emergence decreases. The university’s geographic proximity to firms that use biomedical knowledge increases the likelihood of department creation. Specifically, the density of research firms within a 60-mile radius is directly related to the probability of a university adding a new biomedical department.

Emerging industries can also be tracked using the technology categorization scheme used by the U.S. Patent and Trademark Office (USPTO). The USPTO categorizes patents and plans to establish a new patent class for emerging fields. The categorization allows researchers to see what firms and universities are developing inventions in a particular area and location.

She closed her comments by expressing the state of convergence and the opportunities that it presents. In the COVID-19 era, a time characterized by a broader recognition of societal problems, business schools are pondering the timing to redirect strategies away from shareholder gains toward stakeholder and user needs. Feldman highlighted the importance of underlying values, including sustainability and social justice.

DISCUSSION

Klein asked whether each presenter could connect their presentations and findings to the previous presentations, particularly those of Dan Stokols (University of California, Irvine) and Gary Anderson (NCSES). Holbrook added a request that the panelists address how their own experience practicing convergence might inform a definition that could be used by NCSES.

Weiss said that it would be ideal to have consistent language for distinguishing between the terms “interdisciplinary” and “convergence.” Consistency across universities will be particularly critical as the tenure process considers convergence. Penn State uses the term “interdisciplinary” throughout the institution. All interdisciplinary institutes comprise faculty from various departments, and hiring is done collaboratively between the institutes and university deans. The practice of cohiring allows departments to gain faculty members for disciplinary advances, while new hires allow the interdisciplinary institute to broaden its base of expertise.

Gallahan agreed that a clear definition is useful when evaluating and understanding metrics, but he does not feel hindered by the lack of a clear definition at present. As faculty are encouraged to engage in team science and interdisciplinary work, Gallahan said challenges might arise as junior scientists and postdoctoral students advance through their careers. These

scientists may be part of different programs but may feel compelled to follow a more classic trajectory of publishing first-author works in a narrow way to line themselves up for promotion. Gallahan said that funders need to recognize that they are part of the problem. A researcher may put forth impactful interdisciplinary work that nonetheless may be difficult to justify to a traditional tenure promotion committee.

Schiffer added that the definition depends on the goal. Why does NCSES want a definition of convergence? Will the measure be used to determine whether convergence is happening or to measure the impact of convergence research? Schiffer stated that slight variations of the definition may be needed, from which different measurements would follow.

Feldman noted that NCSES provides data measuring university research and development (R&D) organized by traditional discipline, but it does not yet consider research center activities and the contributions made within universities.

Holbrook referred to earlier presentations that noted a desire to measure convergence because it may have a greater societal impact, while disciplinary research may have a more significant scholarly impact. The different practices and outcomes of convergence may bring about solutions to society’s most pressing problems, which constitutes a motivation for measuring convergence. The flourishing of disciplinary scholarship should continue, but convergence must be the focus because it promises greater societal impact, which will influence its definition, Holbrook said.

Gallahan asked Holbrook to clarify whether he was referring to measuring how well convergence research solved societal problems in the past, or how much convergence research is currently occurring within an institution. Holbrook stated he did not know the answer to this exact question, but he believes that the first priority of NCSES should be the need to understand whether convergence can be measured, and if so, how to best carry out measurement.

Weiss reiterated Schiffer’s previous point that the traditional, single-focus disciplines cannot and should not be forgotten or discounted, as interdisciplinary research depends on advances within these disciplines. She also maintained that the best of disciplinary fields come together, and that each investigator’s contribution must be recognized. Gallahan agreed that modes of recognition could change, but he questioned who in academia is paying attention to these changes. Gallahan added that such changes may be suitable for more senior investigators who have already established their credentials, but unconventional modes of recognition might be detrimental to junior and untenured investigators. The responsibility to acknowledge interdisciplinary contributions falls to both academic institutions and funding agencies. Gallahan addressed a balance between focused research and more interdisciplinary programs. He views NCI as heavily weighting Research

Projects Grants (R01), so providing metrics for interdisciplinary programs could assist in quantifying the success of future interdisciplinary work.

Feldman spoke about typologies, referencing Pasteur’s quadrant.³ This classification distinguishes between basic and applied science, with the most fruitful topics of investigation having an applied orientation that adds new knowledge. Using academic entrepreneurs as an example, she argued that the most productive academic scientists are those engaged with commercial activity that translates their research into practice. Academia brings in creative graduate students but then trains them within a disciplinary box. Consequently, this practice could cause a potential decrease in creative productivity of recent doctoral graduates.

Gallahan concurred with Feldman and noted the increasing tension between supply and demand. At NCI, application rates have nearly doubled in the last 5 years. If the budget does not match that rate, then award rates will flatten. Additional considerations need to be made for collaborative research and bibliometric acknowledgment. Researchers are less likely to have first-author publications as their collaborative work increases, which may prevent them from advancing within academia.

Nora Cate Schaeffer (University of Wisconsin–Madison) asked about what happens as convergent teams and their membership shift over time. Do the groups tend to stay together and investigate different problems, or do individuals leave a group to join a different group based around a different problem? Weiss replied that research funded with partnership-forming seed grants continue to work together. At Penn State, the relationship is seeded along with the research. Anecdotally she finds that once an investigator has collaborated in interdisciplinary research, they become more open-minded and willing to collaborate with other researchers in the future for different projects. Schiffer and Gallahan agreed with Weiss’s statement, and Schiffer added that institutional support for working together in teams is critical. Strong support for working together in interdisciplinary teams makes future research feasible.

Andrew Zukerberg (National Center for Education Statistics) asked whether areas of convergence can turn themselves into new disciplines. Feldman responded that this is possible, pointing to the example of neuroscience. Initial work began with the limitations of neurology and searches for new tools to investigate problems. She said that more work is needed in this area to allow new disciplines to blossom. Ideas have previously been described as “sleeping beauty science,” that is, fundamental ideas lying dormant until an ecosystem concept is applied that allows a quorum of understanding to emerge. Business demands or funding calls are often the

___________________

³ Stokes, D.E. (1997). Pasteur´s Quadrant: Basic Science and Technological Innovation. Washington, DC: The Brookings Institution.

driving force to introduce a new discipline, but this method often does not consider larger social implications.

Stokols responded to Feldman’s comments to consider Pasteur’s quadrant and the value added when research outcomes have near-term applicability. A longer latency period between the initial concept and the buildup of empirical evidence can be translated and applied in other areas over time. Stokols suggested that investment in convergence research that investigates basic work could have translational applicability in the future. He suggested distinguishing between different types of convergence when developing a taxonomy. Gallahan noted that Pasteur’s quadrant is a constant challenge across all disciplines and said that he often finds himself defending basic research.

Josh Schnell (Clarivate) asked whether a national measure’s utility depends on how a research portfolio balances, for example, basic, applied, and translational research. Schiffer pushed back on the concept of a single national measure of convergence, stating that a portfolio of metrics is more likely to be of value. Feldman agreed with this sentiment.

Schaeffer asked whether convergence could include translational research. She also asked whether social sciences and the humanities are considered niche areas dependent on a specific problem. Gallahan answered that translational research could be combined with other types of research. Schiffer agreed with Gallahan regarding translational research and added that social sciences and the humanities are not niche fields. Historically, the humanities have not been thoroughly integrated with the STEM fields at universities, but humanities scholars can play necessary and important roles that enhance the value of STEM research and vice versa.

Weiss expressed skepticism of the idea of a single definition and metric for convergence. Holbrook questioned whether Weiss would be more willing to accept a single definition of convergence if it were measured in multiple ways. Weiss acknowledged that she could accept a single definition with multiple measures, but only if the definition captured nuance, which could be very difficult. Holbrook commented on the work that Anderson presented, in which web survey participants are shown a single definition but can use a hyperlink that provides more detail and multiple examples. He asked whether providing additional definitional information in the hyperlink could still be considered providing a single definition, which Weiss found acceptable.

Schiffer added his description of convergence, which provides a framework that recognizes the importance of addressing several facets of research, particularly team dynamics and institutional incentives for facilitating collaboration, in order to accomplish an overarching aspiration rather than a monolithic goal. Gallahan agreed with Schiffer’s description and endorsed that definition of convergence.

This page intentionally left blank.