Today, scientific research projects often can be characterized as ‘interdisciplinary projects’, i.e. projects that integrate knowledge and know-how from different (sub-)disciplines. However, interdisciplinarity comes with all kinds of problems. This paper addresses frequently reported problems related to interdisciplinary communication. Consider, for example, the following observation by a cultural and a physical geographer:
“Given different epistemological approaches of different disciplines, finding a way to work together to produce a coherent output is exceedingly difficult, not least because disciplines have developed their own specific technical languages reflecting these differences.” (Jones & McDonald 2007, p. 491)
One cause of impeded interdisciplinary communication is terminological ambiguity, i.e. within the context of an interdisciplinary project, some terms are shared by multiple disciplinary jargons, yet it is unknown whether the concept behind each them is the same in all jargons. In other words, interdisciplinarians might be unknowingly using different ontologies, thereby inhibiting clear and efficient communication. As interdisciplinarians already experience heavy workloads and generally are not trained to deal with communication problems, they would clearly benefit from using a tool that enables to deal with terminological differences. What follows is the outline of such a tool.
The tool has two goals: (i) to identify terminological ambiguities, and (ii) to resolve these ambiguities. The first goal entails three subgoals. First, the shared terms are to be identified. Next, it should be checked whether the concepts underlying shared terms vary across the source disciplines. Finally, the differences between the underlying concepts need to be articulated in detail.
To reach the first goal, the tool makes use of natural language processing techniques. For the first subgoal, term extraction software is used to generate lists of key terms for each disciplinary jargon whereupon the set of shared terms is determined. To reach the second subgoal, the tool starts from Harris’ hypothesis, stating that words occurring in similar contexts have the same underlying concept, where ‘context’ is understood as a number of words before and after a term (1968). The different concepts underlying a shared term are determined by checking the concordance with other terms (with underlying concepts) by means of statistical filtering. The third subgoal requires representing all concepts underlying a shared term and comparing them to determine the differences. Based on the identified contexts of shared terms, representations are generated using a combination of (i) Thagard’s diagrams containing kind and part-whole relations between concepts (1992), (ii) Kuhn’s relations of similarity and dissimilarity (1977), and (iii) Chen and Barker’s work on attributes and values of concepts (2000).
The second goal comes down to developing a new ontology in which every shared term has only one underlying concept which integrates the former variety of underlying concepts in such a way that it both accommodates the needs of the project and remains acceptable for all of the involved researchers. To reach this goal, the generated representations will be used to set up a moderated group discussion.
Nanoscience is an inherently interdisciplinary field of study. Because it developed around a scale, rather than a set of laws or phenomena, it invites research programs from fields as diverse as materials science, biology, physics, chemistry, engineering, and design. For instance, gold nano-cubes are synthesized and characterized by chemists and physicists; modeled on computers by mechanical engineers; studied for their color-changing properties in stained glass by art historians, designers, and materials scientists; and manipulated for smarter drug delivery by chemists and biologists.
This scale-centric character of nanoscience means that knowledge in nanoscience is often grouped not along disciplinary lines, but rather around instrumentation techniques (as Mody (2011) has argued), around individual materials, as described above, or around particular applications. Consequently, the structure of knowledge in nanoscience is better understood as clusters of Galisonian “trading zones,” rather than a taxonomy of laws, theories, models, and heuristics. These trading zones permit contributions from diverse research perspectives—including those from history and philosophy of science.
I have spent over 3 years working with a nanoscience laboratory with the aim of understanding the structure of knowledge in nanoscience. Through this work I have become convinced that philosophers and historians of science can impact the development of new knowledge in nanoscience alongside practitioners in STEM fields. My talk shows how contributions from history and philosophy of science can provide new knowledge in nanoscience by describing how philosophical reflection on the concept “surface” led to reforms in experiment design in my lab.
As interdisciplinary fields such as climate science, genetic engineering, or behavioral economics gain prominence in contemporary scientific practice, an increasing number of investigators of scientific knowledge began to probe problems that arise in this context. One example of these problems is the question of disciplinary integration, both at the epistemic and social levels. However, most of these studies deal with the problem from a theoretical or formal perspective. In my paper, I offer an approach to the integration problem via the “new experimentalist” thesis that “aspects of experiment might offer an important … resource for addressing key problems in philosophy of science.” (Mayo, D. “The New Experimentalism, Topical Hypotheses, and Learning from Error,” in PSA: Proceedings of the Biennial Meeting of the Philosophy of Science Association, Vol. 1994, Volume One: Contributed Papers (1994), pp. 270-279, p. 270.)
The experimental episode that I focus on is the discovery of the acceleration of the universe, which was made by two independent research collaborations in 1999 (to be awarded a Nobel Prize in 2011). The aim of the paper is to examine how one of these collaborations, known as the Supernova Cosmology Project (SCP), had to deal with several significant obstacles stemming from inter-disciplinary integration, in order to reach their final knowledge claim. The SCP team is centrally located at the Lawrence Berkeley National Laboratory (LBNL) at the University of California, Berkeley and all the founding members of the team came from a particle physics background, which governs the evidential culture of LBNL. This particle physics culture proved to be a key epistemic factor for the SCP research. Using the concept of inter-experimentality, understood as the study of the interaction of differing practices of empirical confirmation in inter-disciplinary contexts, I analyze both the formation history of the team and their attempts to be recognized as competent epistemic actors within the astronomy community, which regarded them as particle physicists invading astronomers’ turf.
There were two forms of integration problems that the SCP had to deal with. Internally, the particle physicists and the astronomers within the SCP had to work together throughout the data collection and analysis procedures. Externally, the group had to legitimize itself within the astronomical community, who were the main judges of their work. In both these cases, there arose problems. For example, it became clear very early on to the astronomers in SCP that the particle physicists of the group did not know enough astronomy and made elementary mistakes during the data collection processes at the telescopes. Moreover, in several oral history interviews I conducted, astronomers of the SCP complained about the “over-confidence” of the particle physicists that the statistical analysis would take care of the problems that arise in data collection. The evidential cultures of particle physics and astronomy clashed within the team.
Externally, the SCP had to overcome a cultural barrier to launch a successful research program. For several years, the group had enormous difficulty in obtaining observation time at the big telescopes. As they were perceived as “outsiders,” the astronomy community was reluctant to grant access to the SCP researchers for “they did not know what they were doing” (This remark was made by a senior astronomer who is an expert in supernova research. He later joined the rival group, known as the High-z, which was composed only by astronomers). In order to legitimize their methods, the SCP had to publish a premature paper explaining their research endeavor to the astronomical community, using questionable data. Even though the results of the paper later turned out to be wrong, particle physics trained members of the group still think that it was a necessary move, for it gave them both communitarian and institutional validation for further research (SCP also had issues of losing support from their home institution, LBNL, due to not having produced results for a number of years).
In the body of my paper, I draw several lessons from this episode, following a historical presentation that documents the above claims. I argue that we need to be attentive to inter-experimentality in order to understand empirical confirmation in inter-disciplinary research. More specifically, I claim that perceived disciplinary hierarchies and different cultures of evidence play key roles in the confirmation practices of inter-disciplinary research and in certain contexts, this may impede progress. Both in the micro-level contexts of data collection, statistical analysis and announcement of the results, and the macro-level settings of institutional support and access to research facilities, the sociology of inter-experimentality is seen to have direct epistemic consequences, as I aim to demonstrate.