Multidisciplinary Epistemic Groups — preliminary factors
I will name Multidisciplinary Epistemic Groups (MEGs) those collaborative research groups formed for the purpose of seeking a response to complex scientific questions by employing different research methods and disciplines. I will refrain from discussing the differences between multidisciplinarity, interdisciplinarity, transdisciplinarity and so on (for this, see, inter alia, Blackwell, 1955; Rosenfeld 1992, and Choi & Pak, 2006, in the context of health research; Stember, 2005, in the context of the social sciences; Wilshire, 1990, and Nicolescu, 2005, in the context of education; Isemonger, 2018, on their conceptual distinctions within the digital humanities). My only comment is that I take multidisciplinarity to be the least demanding form of collaborations with more than two members.
The requirement for diverse methods to be applied to one research problem can be motivated in different ways. For example, it can be brought about by increasingly complex scientific questions, or a growing interest in — and availability of — machine learning (ML) and artificial intelligence (AI) as approaches to these questions, or simply the belief that more collaboration is better. In either case, what result are groups of individuals from different scientific fields, and with different epistemic backgrounds.
This epistemic diversity leads to an important barrier to success within MEGs for several reasons. Firstly, there will be a conceptual division: each discipline within the group will work with a set of concepts and terms that are specific to their fields; e.g.: the philosopher of the social sciences might not agree with the economist on what is meant by reflexivity. Secondly, differences will be present in what regards methodologies employed by each discipline; e.g.: the ML expert’s approach to statistical models has little to do with the historian’s interpretation of Medieval texts. Thirdly, each discipline will hold different standards in what regards success in science; e.g.: are we looking for 5-sigma, a p-value below .05 or simply some result that fits with another project’s finding? This list of barriers is by no means exhaustive, and it relates solely with the epistemic diversity described. I therefore call these epistemic barriers, which can be accentuated by the number of such fields present in a MEG, and by the “gap” between any two fields; e.g.: consider how the economist might find it easier to understand a sociologist than, say, a microbiologist.
A further problem relates with how each field operates. I don’t think it is too controversial to assume that each research field is a world of its own, but I wish only to outline one common source of further complexity when it comes to studying MEGs: institutional norms. Very roughly, universities and research institutions at large fit within the larger sector of academia. Academia, in turn, is devised in such a way that it encourages the publication of scientific work through an incentives system that does not recognise the pursuit of truth in the way we might expect or might want. As a parenthesis, I do not take truth to need defining in this post as anything more than “one relevant goal of science”. Returning to the topic at hand, the incentives structure sustain what is commonly referred to as the “publish or perish” culture. In short, this culture means an individual scientist is incentivised to publish their work in academic journals because this will allow them to secure better jobs and have access to more interesting opportunities in the future. The pursuit of truth may be deemed secondary. Within this system, academic journals will be of greater use to the researcher depending on their perceived prestige, which will differ from field to field. The amalgamation of these forms of prestige and opportunities is commonly called credit. We can thus take this publish-or-perish culture to mean that researchers constantly weigh two options throughout their careers: between the pursuit of truth and the pursuit of credit. Ultimately, this will result in some form of compromise — unless one is willing to throw truth out the window and strive for some type of scientific fraud, or one decides to ignore context, basic scientific standards and all philosophical debates, and claim to have uncovered Truth-with-a-capital-T.
Now consider that each scientist is in this position, even when working within MEGs. Each scientist is — in part — in it for themselves. They know the journals where their publications (and the MEGs’) will be the most profitable (for themselves), and they know the collaborators within the MEG who they can produce the most interesting work with in the future. Individuals’ goals can therefore be in conflict with some, in line with others’, and dependent on a third group’s. I call each of these relations occurring within MEGs divergent, convergent and dependent. Along with epistemic barriers and institutional norms, we thus have a third dimension for the study of MEGs: individuals’ interrelations, which can become more complex as MEGs have more and more members. **But these interrelations are as sporadic as the MEG itself; they are all time-bound. So, on top of being subject to all these pressures to impress the right collaborators, obtain the highest amount of credit given the circumstances, and potentially figure out how to work across different fields they are not experts in, these pressures are increased by their time-bound nature. And where does this time-pressure come from? From a fifth dimension of complexity for MEGs (and all research programmes): funding.
Funding for universities and institutes could be said to come through two major streams: government-sponsored research-funding bodies (which I will naively call “government”), and businesses and charities (which I will naively refer to as “industry”). But under what premises is such funding granted? Here I wish to mention only one such premise: that the research meet the specific interests of the funding body. For example, government might be interested in the development of ML technologies that support their military programme; a financial institution might fund research into default-prediction models; a charity might focus on their own charitable objects; and so on. In this sense, MEGs are subject to the ethics of their funders. We can take this to be quite the constraint: “academic freedom” (whatever that means) is subject to the ethics of others. Researchers, in a sense, have their autonomy limited by the source of funding. The importance of credit is heightened here, given the previous depiction: increased credit means increased opportunities to work in areas researchers are more interested in. In this way, the focus on seeking credit is arguably greater in the case of early-career researchers. Returning to the credit-truth compromise, it seems that the balance tips towards the more noble pursuit of truth as one advances in their career.
I have briefly outlined some rather individualistic notions related to MEGs, but I think they are relevant factors regardless of the theory of such collaborations we wish to present. Whether focused on the psyche of the scientist, or the collectives formed, an analysis of MEGs will require that of epistemic diversity and the resulting barriers, institutional norms and the resulting interrelations, time pressures, and the ethics of funding bodies.
