Sabina Leonelli’s “Philosophy of Open Science”

Sabina Leonelli’s new book – “Philosophy of Open Science” – will be published later this year. However, there is an open access preprint available on the PhilSci Archive here.

In the book, Leonelli provides “a constructively critical reading” of the standard approach to open science which, she argues, is focused on sharing “objects” such as data and materials. In Leonelli’s view, this object-sharing approach has become an obstacle to promoting open science. Grounded in philosophies of science from Popper, Chang, Longino, and others, she puts forward an alternative, process-oriented view of open science that refers to efforts to establish “judicious connections among systems of practice.”

I propose a conception of openness as judicious connection, which is grounded in a process-oriented epistemology of science that recognises the situated, embodied and goal-directed nature of communication and collaboration among researchers.

What does “connection” have to do with openness?

The idea of openness is quintessentially linked to that of learning as going beyond one’s boundaries. This arguably applies to systems of research practice as much as to research groups and individual learners: establishing new connections often means expanding one’s learning, challenging existing assumptions around what is considered external or irrelevant to a given system, and considering whether new boundaries need to be established that incorporate the novel relations.

Importantly, Leonelli does not equate an indiscriminate increase in scientific connections with an increase in open science. Instead,

the building and maintenance of connections need to be judicious: they require skilled deliberation, whereby the new opportunities offered by the connection in question are evaluated within the contexts at hand. Indeed, openness can itself be understood as a dynamic and highly situated mode of valuing the research process and its outputs, which encompasses economic as well as scientific, cultural, political, ethical and social considerations (Levin and Leonelli 2017).

The differences between the object-sharing and “judicious connection” approaches to open science also have implications for inclusion. The standard object-sharing approach assumes that greater transparency leads to a more inclusive science. In contrast, Leonelli’s “judicious connection” approach flips the causal arrow and assumes that more inclusion leads to greater transparency.

The implementation of OS [open science] needs to start from consideration of what it may take to make research more inclusive, diverse and just – rather than expecting such an outcome to naturally follow from the ‘right’ choice of software, infrastructures, standards, publishing platforms, or whichever other technological or institutional fix is being devised to facilitate access to resources.

Finally, Leonelli’s distinction between the two approaches to open science is important because she believes that the standard, object-sharing approach:

threatens to blindly privilege specific ways of knowing, thus potentially disrupting sophisticated methodologies, inadvertently dismissing well-established research traditions, and exacerbating the already large epistemic and social divides separating research domains and locations As denounced by a number of critics in science and science studies, there is a substantive risk of some OS policies – despite their good intentions and progressive slant - acting as a reactionary force which reinforces conservatism, discrimination, commodification and inequality in research, thus ultimately closing down opportunities for inquiry in a disastrous reversal of what they set out to achieve.

I won’t go any further into the distinction between the object-sharing and judicous connection approaches to open science. Leonelli already does an excellent job of that in her book. Instead, I want to highlight five quotes that are critical of the standard object-sharing approach, because these are motivating towards a consideration of her alternative judicious connection approach.

(1) Standard Open Science Privileges a Homogeneous View of Science

Many of the more institutionalised OS initiatives tend to privilege a homogenous, universally applicable understanding of the scientific method over a pluralistic and situated one.

Similar points have been made by several others. For example, Malich and Rehmann-Sutter (2022) pointed out that:

Metascience often construes science and psychology from a homogenizing perspective that omits the plurality and complexity of both scientific methods and psychological approaches. This omission entails several epistemic risks because the reforms that metascience initiates or accompanies can reinforce these homogenizing tendencies

(2) Open Science Should Avoid Mindless Transparency

Open Methods is not a matter of recording and sharing every detail of a research procedure, but rather a reflection on which research components and techniques are most salient to the outcomes, and should thus be accessible and reproducible.

I totally agree. As I argue near the end of this presentation, I’m opposed to what I call “mindless transparency.”

By the way, I think Popper (2002, p. 230) would also agree with this point:

When the judge tells a witness that he should speak 'The truth, the whole truth, and nothing but the truth', then what he looks for is as much of the relevant truth as the witness may be able to offer. A witness who likes to wander off into irrelevancies is unsatisfactory as a witness, even though these irrelevancies may be truisms, and thus part of 'the whole truth'.

(3) Reproducibility Should Not be an Unconditional Demarcation Strategy

Taking a monolithic understanding of reproducibility as a demarcation strategy for the whole of science side-steps the precious plurality of methods developed to suit specific goals, concepts and target objects.

Again, this point has been made by several others (e.g., Devezer et al., 2019, 2021; Feest, 2019; Haig, 2022). Nonetheless, replication is often portrayed as a “defining” or “essential” aspect of science.

(4) Skepticism Against Qualitative Research is Unwarranted

A case in point is the skepticism towards qualitative research often displayed within debates on reproducibility, which sometimes portray qualitative methods as hopelessly subjective and devoid of rigorous forms of data collection and verification....Assuming that qualitative research traditions in anthropology or sociology are incapable of rigorous research means drastically reducing the diversity of systems regarded as exemplars of good practice.

Yes! We need to be very careful that open science does not reopen “‘sectarian’ divides between quantitative and qualitative” research, as Dunleavy (2022) has cautioned.

(5) Open Science Tech and Procedures Should Consider Diversity and Inclusion

To date, some parts of the OS movement – particularly its institutionalized, top-down incarnations - have paid too much attention to designing procedures and technologies for sharing, and this has come at the expense of strategies, training and procedures to assess who is included and excluded from such apparatus, understand why and with which implications, and mitigate eventual instances of epistemic injustice.

Another important point! A good example of this lack of attention to inclusion comes from Prosser et al. (2022) in their article “When open data closes the door.” These researchers found surveyed 261 social psychology journals and found that journal guidelines for sharing data were exclusionist with regards to qualitative data:

Guidelines explicitly addressing open qualitative data were very rare, provided by just three journals (2.1% of all relevant journals). All other journals that made reference to 'open data' did so without making a distinction between qualitative and quantitative data.”

Further Information

Leonelli, S. (2023). Philosophy of open science. [Preprint] http://philsci-archive.pitt.edu/21986/

Order the published print version here.

The University of Exeter’s PHIL_OS project (2021–2025): “Aims to develop an empirically grounded philosophy of Open Science [OS] that emphasises the diversity of research environments around the world and articulates the conditions under which OS can leverage such diversity to promote good research practice.