The snows of yesteryear

Marco Almada

Marco Almada

— 7 min read
A photo of an abandoned stone building, with its entrances sealed, in a sunny day.
Photo by Dann Zepeda / Unsplash

Hello, dear reader, and welcome to another issue of AI, Law, and Otter Things! In today's issue, I wrap up my initial reflections on legal research programmes and the production of knowledge with some thoughts on whether the notion of a 'degenerate research programme' might help us put a name to the intuition that some research agendas just...fade away. After that, as usual, there are reading recommendations, job and event opportunities, and a cute otter.

When I engaged with the notion of a research programme in an earlier issue, I did so in an anthropophagic mood. I was not so much worried with doing justice to Lakatos's picture of science or the problems that he and other philosophers of science were trying to address, focusing instead on whether (my reading of) their outputs might be somehow relevant to the law.[1] For today's flight of fancy, however, I will need to add some context: what they were looking at.

If we go to Lakatos's Falsification and the Methodology of Scientific Research Programmes,[2] we can see from the outset that the notion of scientific research programmes was introduced as an attempt to explain science as a rational practice (or at least as something that can be reconstructed as one). Doing so requires, on the one hand, distinguish scientific from non-scientific practices[3] and, on the other hand, explain why scientific knowledge seems to grow. It is in this context that the distinction between the 'hard core' and the 'protective belt' of a scientific research programme emerges: to show why scientists might be rationally justified in sticking with a particular set of theories even in light of discoveries that cannot be properly explained by them.

So, when would rationality compel the abandonment of a particular research programme? To answer this question, Lakatos distinguishes between progressive and degenerating research programmes. In short, a progressive research programme is one in which abandoned theories are replaced by new theories that predict some facts that were not expected under the previous theories (this is what makes the programme theoretically progressive) and these new theories lead to new empirical discoveries (thus being empirically progressive). If a research programme fails to have both problems, it becomes a degenerating programme, as it is no longer yielding new knowledge about the world. But it can retain a claim to scientific-ness if it is, at the very least, theoretically progressive.

The account given above of scientific progress does not speak to many pressing concerns of legal scholarship. There is, indeed, a need to demarcate law from non-law, but, in legal philosophy, that tends to have more with the nature and identification of political authority than with facts above the world.[4] But the distinction between progressive and degenerating programmes might nonetheless be of some use to explain why some research programmes, once popular, end up losing steam.

When extending such an account to legal scholarship, we once again face the problem that its main object of interest is not directed at empirical facts.[5] To the extent that one is interested in the normativity of the law, one is not looking at facts, but at the counterfactual expectations that the law may produce -- such as the idea that somebody who has been harmed in a certain way has the right to compensation. Therefore, one would have to specify a certain object of interest that would play the role that empirical facts play with regard to scientific research programmes.

Multiple constructions of this type might be possible, leading to different kinds of insights. For example, a legal positivistic research programme might be assessed in terms of what it tells us about the norms that belong to a legal system. It would be theoretically progressive to the extent that the succession of theories belonging to it offers explanation to a growing set of phenomena: for example, by accounting to recent developments such as the relevance of technical standardization and other forms of private orderings within state law. And it would be empirically progressive to the extent that the theoretical predictions offer a good account of law "as it is". As the latter framing suggests, a lot of this evaluation would eventually hinge on value judgments about research agendas that might themselves be contestable.

A more concrete example might illustrate this. A bit over a decade ago, the legal scholar Anu Bradford introduced the term Brussels Effect to label her theory of how the EU manages to set global standards through unilateral action. This theory, refined by Bradford and others along the years, proposes certain conditions that would be necessary and sufficient for the propagation of EU unilateral norm-setting, and identifies some examples where it would be arguably successful. As such, the early successes of the theory illustrate both elements of a progressive research programme. Theoretical progressiveness can be seen in the successive refinements of the theory, while the wealth of instances of the Effect would suggest a degree of empirical progressiveness. It comes as no surprise, under these circumstances, that the Brussels Effect became part of the imaginary not just of scholars but also of the EU institutions themselves, as seen in the AI Act's declared aims of leveraging this effect towards EU leadership in the regulation of AI technologies.

In the last two years or so, however, one might point towards some indicia of degeneration in both aspects of the research programme. From an empirical perspective, one might point out to the absence of EU influence in cases where it would be expected: for example, AI regulation around the world has pursued approaches that deviate from the EU template, even when the theoretical conditions for a Brussels Effect are met. To accommodate these deviations, some authors have proposed adjustments to the policy-making mechanisms as well as supplementary effects that would address some of the perceived weaknesses of the model. If those adjustments fail to expand the reach of predictions offered by current theories of the Brussels Effect, the research programme might eventually become theoretically degenerative. At that point, it would be irrational to stick to it as a tool for understanding whether and how EU-set norms spread around the world.

To fully explain why research programmes in law are abandoned, one cannot restrict oneself to purely epistemic reasons. Macro-political factors will surely come into play, as one can see, for instance, in the rise (and perhaps fall?) of the originalist fetish in US legal academia. Politics at a smaller scale can also be influential, as disputes between schools of thought shape the availability of academic jobs and what kinds of scholarship get published where. I suspect we even might get something out of looking at this problem from the perspective of aesthetic fads among scholars.

But, unless we are fully sceptical about the epistemic value of legal scholarship, it does not seem much of a stretch to believe that epistemic factors play at least some role in why some approaches to the study of the law are no longer seen as fruitful. If what a particular set of theories tells us about the law no longer matches our construction of reality, we might no longer see that as particular informative. And, to the extent that rejection is guided by epistemic grounds, looking at the programmes of yesteryear as something akin to degenerating research programmes might tell us something about how legal programmes generate knowledge about the law.

  1. Of course, were I to develop this beyond this early-stage reflection, I would need to properly reconstruct the concept and explain where and why I deviate from its 'proper' formulation. I will not do so at the moment, but please bear with me. ↩︎

  2. Here, I am relying on the version collected in Lakatos, The Methodology of Scientific Research Programmes: Volume 1: Philosophical Papers (John Worrall and Gregory Currie, eds.) (OUP, 1999). ↩︎

  3. Such as astrology, or whatever it is that your colleague next door is invoking as a 'scientific' basis for legal scholarship. ↩︎

  4. With apologies to my legal philosopher friends, even those of a hedgehog persuasion. ↩︎

  5. Lakatos himself would not see much of a problem in applying the idea of a research programme to this kind of a normative object. In p. 92 of the book quoted in [^fn-fmrsp], he explicitly mentions that this approach could be extended to non-empirical research programmes, as he does himself in his treatment of logic. ↩︎

Things you might want to read

Opportunities

Disclaimer: as usual, I am gathering these links purely for convenience and because I think they might be of interest to readers of this newsletter. Unless I explicitly say otherwise, I am not involved with any of the selection processes indicated below.

The DSA and Platform Regulation Conference 2026 invites abstracts by 30 September 2025, with the event taking place in 16-17 February 2026.

The Law department at the University of Surrey is looking for a Lecturer B in Law, with particular interest in scholars whose research is relevant to some aspect of defence or national security policy. Applications are due by 2 October 2025.

In 8 October 2025, Ben Farrand (Newcastle University) will speak at a webinar about Cybersecurity, Weaponised Interdependence, and the AI Race: The Geopolitical Commission at a Crossroads. Free registration, and make sure to join and get a sneak peek at his upcoming book!

The Institute of Security and Global Affairs at Leiden University is looking for two PhD candidates (4 years) and a post-doc (2.5 years) for a project on best practices in organizational cybersecurity. Applications are due by 24 October 2025.

Vassilis Galanos and Maarten Hillebrandt are putting together a special issue on Disinformation after Generative AI & Synthetic Data for Information Polity. Abstracts are due by 14 November 2025, with full papers expected by early March 2026 for publication in 2027.

And now, the otter!

a couple of otters standing on top of a log
Photo by Theo Jores / Unsplash

Hope you found something interesting above, and please consider subscribing if you haven’t done so already:

Subscribe

Thanks for your attention! Do not hesitate to hit “reply” to this email or contact me elsewhere to discuss some topic I raise in the newsletter. Likewise, let me know if there is a job opening, event, or publication that might be of interest to me or to the readers of this newsletter. Hope to see you next time!

Read more