I really enjoyed reading the article “The Value of Openness in Open Science” by Santana (2024), which I accidentally stumbled across on bluesy (Connect with me there: @mirjamglessmer.bsky.social). Santana (2014) make the point that maximum openness, while maybe well-intended, is not necessarily optimal.

Openness in science is a value that is not necessarily 100% aligned with other, possibly equally important values. There are several issues that can arise if there is too much openness without considering competing values:

1. “Standards standardize”: While standards for some things might be good, they also work against epistemic diversity, which in itself is good for science. Open Science, as standard, can disadvantage researchers from less wealthy institutions or regions because they can more easily be scooped, and because publishing open access is often a lot more expensive. Pre-registration of studies as standard might push people towards confirmatory rather than exploratory research. Implementing standards before they are well thought-through can create a big mess. Standardizing how data is represented might lead to a loss of contextual data because not all context can be represented in a standardized form.
2. What makes science better on the individual level might not make it better on larger scale. Here, the author gives an example of a lab tech messing up (bad), which leads to better protocols and checks in the group (good), which might lead to a loss of diversity and redundancy with other groups in the field (bad), which might lead to more reproduction of results and thus better science (good) but less overall knowledge production (bad), and so on. You cannot scale up what exactly makes “good science”.
3. “Scientists are experts”. Expert judgement cannot always be made transparent in a way that everybody can understand it. When experts “dumb down” their thought process, they necessarily misrepresent it to some degree. To avoid having to do that, they might restrict what they work on to what is more easily explainable. And they might, like non-experts looking at the experts’ work from the outside, start relying on metrics to judge the quality of work rather than using their expert judgement.
4. “Context matters”. Just because some piece of code is available to everybody means that everybody can understand the assumptions and decisions that went into it and interpret the output accordingly. Same for data, protocols, anything else.
5. “Misinformation”. Results can be used too early or out of context.

This article is not only super well-written and very entertaining to read. I really liked how it changed my view on openness in Open Science — it is way more complicated than just making everything more open and then everything will get better, and there are so many facets to consider! While Santana (2024) provides a really helpful way to talk about openness as one value that needs to be balanced against other values, and it becomes very clear that the sweet spot will look different in different contexts, and that we need to learn to be ok with that.

Santana, C. (2024). The Value of Openness in Open Science. Canadian Journal of Philosophy, 1–15. doi:10.1017/can.2024.44