7.1.1Interdisciplinarity – Perspectives and Approaches

This step concerns conceptualisation. In the first place, the possible perspectives and approaches with respect to interdisciplinarity as a quantifiable phenomenon need to be determined. This comprises, among others, the delineation from other, related concepts like multi- and transdisciplinarity, and the framework within which interdisciplinarity should be conceived and interpreted. As early as in 2016, in the context of data integration from the perspective of funding and performing organisations, the issue of the different perspectives of (subject) assignment, their ambiguity and their effect on data processing and evaluation was raised in a systematically manner. In particular, the activities of each researcher, each application, project and research output can be viewed from different perspectives, such as the professional background of the researcher, the researcher’s organisational affiliation, the administration and management of projects and the cognitive assignment of research outputs. The assignment of a scientist’s research activity to a discipline may thus differ for each of the perspectives adopted, where different assignments may actually point to research in a potentially interdisciplinary environment.
From these perspectives of knowledge integration from different disciplines, bibliometricians usually choose two main approaches, where

  1. the cognitive approach is based on information flows, i.e., on knowledge integration in the framework of knowledge diffusion (e.g., Porter et al., 2006; Zhang et al., 2016), and
  2. the organisational approach is mainly based on the researchers’ activity assignment and scientists’ affiliation in the framework of research collaboration and co-authorship (e.g., Abramo et al., 2012).

The cognitive approach is usually related to the analysis of the cited references in a scientific publication. The idea behind this approach is that interdisciplinary research is reflected by the use of information from different and possibly even unrelated topics in a new context. The advantage of this approach is that any international multidisciplinary citation database can be used with the supplementary option of benchmarking the cognitive aspects of interdisciplinary work. A second, however less common, cognitive approach uses textual (dis-)similarity of documents. Text analysis allows the measurement of the cognitive similarity or distance of documents, or parts of documents such as title, abstracts, keywords (Rousseau et al, 2017).

Both approaches have drawbacks. Reference analysis is always confronted with the different communication and citation cultures in the disciplines, which is to the detriment of the accuracy in some disciplines of the social sciences but most notably in the humanities. The lexical approach may yield incommensurable results when based on full text analysis. One need only think of disciplines in which a highly formalised language is the standard vs. those which use natural language. Possibly, a hybrid solution, i.e., the combination of the two models, has the potential to overcome these limitations (cf. Glänzel & Thijs, 2017).

By contrast to the previous approach that is independent of research collaboration or team composition, the organisational approach requires the manifestation of interdisciplinary research through the collaboration of scientists with different professional backgrounds and/or organisational affiliations belonging to different research fields. In this approach, researchers are, for instance, assigned to (unique) disciplines (Abramo et al., 2012), or assigned on the basis of author affiliation. The main disadvantage is apparent: the assignment criteria and process restrict this approach to a local level without the opportunity of a global benchmarking – at least for the time being. Bibliometricians are, however, studying a combination of the two approaches (Zhang et al., 2018) to overcome the limitations of the underlying methods.