Laboratorio di teoria del diritto e scienze cognitive

What is the Legal Theory and Cognitive Science Laboratory?

The Legal Theory and Cognitive Science Laboratory is a project developed within the Interdepartmental Research Centre Alma-AI at the University of Bologna. Its aim is to conduct experiments in order to empirically investigate the cognitive foundations of law and legal institutions. It is run by a research group consisting of two legal theorists (Corrado Roversi and Michele Ubertone) and three cognitive psychologists (Luisa Lugli, Stefania D’Ascenzo, Caterina Villani) and can count on the collaboration of Anna Borghi at the Sapienza University of Rome and the Institute of Cognitive Sciences and Technologies of the Italian National Research Council. The lab is currently involved in a European project called RECOGNISE, along with research groups from the Universities of Palermo, Maastricht, Krakow, Alicante, and Ljubljana, but it was also supported in the past by the ALMA-IDEA research funding of the University of Bologna.

“Villani et. al. 2021. Is Justice Grounded? How Expertise Shapes Conceptual Representation of Institutional Concepts. Psychological Research”

What kind of experiments have been conducted so far?

The focus of our research so far was on embodied cognition and legal concepts. We found evidence in support of the hypothesis that institutional concepts have interesting features of their own, distinguishing them from both purely abstract and purely concrete concepts. We have explored two sub-hypotheses. First, institutional concepts come into two distinct and cognitively salient variants: pure institutional concepts and meta-institutional concepts, the latter being a prerequisite for the processing of the former. Second, institutional concepts are typically processed and represented differently depending on the subjects’ expertise: non-experts rely on experts in the processing of institutional notions; this is a consequence of a broader division of labor between experts and non-experts on which institutional reality is grounded. We tested these hypotheses by making two types of behavioral measurements: 1) questionnaires asking subjects to rate words according to various criteria and dimensions (e.g. “how much do you think this word is related to visual experience?”; “how much do you rely on others to identify what this word stands for?” etc.); 2) measurements of reaction times in the processing of different concept kinds primed by different contexts. In previous works, we have also used property-generation tasks to assess the relation between institutional concepts and artefact concepts.

What’s the relevance of this kind of experiments for the understanding of human conceptualisation?

The experiments seek to further the state of the art on embodied cognition by looking at a usually neglected form of conceptualisation, that of institutional notions, which is particularly interesting considering the existing embodied cognition literature on concrete and abstract concepts. Studies show that while concrete concepts mainly activate perceptual properties of words referent, abstract concepts elicit higher proportions of complex and rich experiences and dimensions, involving episodes and situational relations, emotions, introspection and interoceptive states and are probably more affected by linguistic, cultural and individual variability. However, it has been shown that a clear-cut dichotomy between the two categories is untenable. Institutional concepts are especially interesting because they seem to be somehow intermediate between totally concrete and totally abstract ones.

What’s the relevance of this kind of experiments for the understanding of law?

Traditionally, both legal scholars and policy-makers have had the tendency to presuppose, in their understanding of law and institutions, a totally naïve and implausible notion of how the mind works, one which we could call the legislator’s default theory of mind. Just think of the notion of contract conceived as the “meeting” of two or more minds on an identical propositional content. There is a default presupposition that the people will usually apprehend the content of norms directly and homogeneously in virtue of a set of a common shared representation of institutional notions. This is almost surely not true, based on what we know already about conceptualisation, for at least two reasons. Firstly, because the complex conceptual machinery that allows a community to operate as a legal system is characterised by a division of cognitive labor. In significantly technical fields such as law, different people are competent in the use of same words in virtue of different cognitive abilities. Experts of law have a direct understanding of institutional notions whereas non-experts, even when competent in the use of the corresponding words, usually have a deferential or indirect understanding of them. Secondly, the mastery of an institutional concept —or any concept really —is not simply a matter of entertaining a mental representation as a sort of if-then algorithm, containing a propositional description of the conditions that should be met for something to fall into its extension. The mastery of a concept is also a form of know-how, which is part of a more general ability to interact with one’s social and physical environment. This research seeks to provide an evidence-based description of the division of labour between experts and non-experts involved in institutional conceptualisation as well as of the role of embodied cognition in this domain.

How do the aims of the Lab fit into the institutional aims of ALMA-AI?

AI is bringing about disruptive changes in the way we think generally and in the way we think about institutions specifically. This change has not always been for the better. The Internet has sometimes favoured the creation of vicious epistemic environments which constitute a threat for the rule of law. Internet algorithms are sometimes said to be partially culpable for the spreading of conspiracy theories and the delegitimisation of institutional role and rules. The Cambridge Analytica scandal has been instructive in showing us how the use of profiling algorithms can poison the epistemic environment in which democratic institutions can flourish and work. We believe that the only way to design safer technologies and put in place rules able to protect us from these threats is having a better understanding of the cognitive underpinnings of institutions. What makes people recognise institutions as legitimate? What kind of cognitive abilities are required to think about institutions critically and understand the value of the rule of law? Have experts and non-experts in law different ways of thinking about what it is for something to be legal? Are institutions realised by way of a cognitive division of labour? These are all questions that should be answered if we want to be able to govern and understand the changes institutions will undergo in the age of AI.