In my research, I have focused on the problem raised by members of the life sciences community over the existing state of the research data ecosystem. The ecosystem is messy and disordered. This is because scientists have adopted different practices to manage data. This makes it difficult to navigate the ecosystem via computational means and find, as well as process what is needed. Thus, to declutter this ecosystem and bring order to it, scientists are being asked to collaborate and standardise how they manage their research data. Yet, prior efforts to standardise indicate that scientists still differ in how they manage data. The general assumption as to why they do so is that data management is a resource intensive task and scientists prefer to direct their resources towards doing research.

I take the stand that there is another reason for different practices to manage data: scientists have different definitions of ‘purity’ and ‘pollution’. Here, the concepts of purity and pollution are the difference between order and disorder. While order maintains the purity of data, disorder pollutes the data. Differences in how data is managed indicates a difference in what one values as order. To standardise their research data management practices, scientists may have to adjust their local distinctions between order and disorder. That is, they may have to accept what, to them, is polluted or reject what, to them, is pure. As this entails accepting some clutter and disorder in their neatly ordered data, this compromise also serves as a legitimate reason for non-compliance with standards. Thus, amongst other recommendations, I propose that in their efforts to standardise these practices, scientists should collectively engage in a dialogue to unpack the different conceptions of purity and pollution.

• Centre for Digital Life, Norway
• Druglogics
• Crossover Research 2.0
• ELSA Norway