Rules vs. Principles in POSI

In recent days, several signatories to the Principles on Open Scholarly Infrastructure have taken to performing self-audits of their compliance with the principles. Of course, holding oneself to account in this way is a welcome development. Without some form of self-appraisal it is not possible to know how close one is to fulfilling the goals of POSI.

However, I wanted to gesture, in this post, towards a distinction between "rules" and "principles" that I believe are fundamental to the systems that POSI wishes to implement. People have begun treating the example specifications in POSI as though they were hard-and-fast rules. For example, the wind-down contingency clause specifies "12 months" – and people can become very anxious about strictly fulfilling this criterion.

But the truth is, taking this example, that some infrastructures will need a year to wind down, while others will only require 6 months. The principles actually go on to state: "12 months in most cases". But some might even take longer than a year. This wind-down period is not just about closing down the organisation, but is also concerned with giving the community that uses the infrastructure enough time to migrate. The point is that the principle here is: you build and set aside sufficient contingency funding to ensure an orderly wind-down. However, some people are treating this as though it were a rule with a hard and fast metric of "1 year".

While we will be adjusting the POSI document to v1.1 with clarification on this particular point about a year's surplus, there is a simple way to judge how each clause applies to you: aim for the principle not any rule.

Another example of this is the licensing of open data. Some have read the suggestion of a CC0 public domain dedication as the rule that POSI signatories must follow. There are very good reasons for this license/dedication (using a more restrictive data license will inhibit forkability and make it much, much harder for third-parties to combine your dataset with others in research projects). However, in reality you should be looking at the principle of the clause. What is it trying to achieve? The answer is: we want other organizations to be able to replicate infrastructures and to be able to re-use their data. The question then is: which licenses or public domain declarations will achieve that. We hope that you will conclude that CC0 is appropriate and useful. But you might have a good reason for a different dedication or license.

Another reason to think of principles rather than rules is that we don't want POSI to transmute into a box-ticking audit exercise. The term "principle" derives, etymologically, from the Latin principium, meaning "a beginning, commencement, origin, first part," or "foundation". To apply the principles from POSI to an organization means embedding them from the start, from the very base or foundation. It does not mean grafting a superficial set of rules over your organization, doing the bare minimum to tick each point off the list.

This is why I am sceptical of self-audit cultures around POSI. We do not want principles to be lost to the rules. Instead, we want organizations that operate, from the ground up, according to the ethics and principles that the document embodies. This requires, of course, interpretation. If the rules are flexible, then we must interpret them according to each context (which may vary between organizations). This may appear intolerable to those who want to be told precisely what to do or to those who believe that digital "law" should be so strictly codified as to not need legal intervention. But ethics and morality – principles – are rarely so clear cut. The rules in POSI are a guide that require interpretation. The principles are what should guide us.

Rules vs. Principles in POSI was originally published by Martin Paul Eve at eve.gd: Martin Paul Eve on July 28, 2023.