Sylvie Delacroix

Transformative agency is not a fixed human endowment. The capacity to perceive that existing frameworks have become inadequate and to translate such perception into action depends on infrastructural conditions that can be eroded. This paper develops an account of how conversational infrastructure shapes the pre-reflective capacities on which transformative agency depends. Central to this account is structural uncertainty, an environmental quality that keeps open the question of whether established modes of engagement remain adequate. Among the settings where such uncertainty operates, sense-making conversations play a distinctive role. Characterised by productive uncertainty rather than efficient resolution, they constitute infrastructure through which societies maintain their capacity for ethical transformation. Such conversations are now increasingly mediated by large language models optimised to provide comprehensive responses. The very features that make these systems feel helpful may foreclose the exploratory space ethical inquiry requires. The paper traces how infrastructure shapes habits of attention, drawing on the ‘kind environment’ thesis while identifying its limits for understanding capacities that involve perceiving framework inadequacy. These limits become salient in institutional contexts where design choices shape conditions for transformative agency, from judicial concepts like mercy to professional settings that support exploratory dialogue. The analysis concludes by examining interventions that could preserve these conditions, including temporal designs that resist premature resolution and participatory architectures enabling collective shaping of conversational infrastructure.

For as long as knowledge asymmetry continues to be deemed the defining characteristic of the lay-professional relationship, the courts’ delineation of obligations meant to address lay vulnerability will too frequently end up compounding the layperson’s non-epistemic, ‘sense of self’ vulnerability. The proposed re-conceptualisation of professional responsibility calls for reform on several fronts: among these, an expanded ‘duty to consult’ (beyond do-not-resuscitate-orders) is uniquely placed as a justiciable criterion capable of addressing such a situational, ‘sense of self’ vulnerability.

On a narrow understanding, adjudication processes are all about assessing and containing uncertainties of various kinds. If one were to find ways of getting language models deployed in judicial settings to reliably quantify the many shades of uncertainty colouring their outputs, one could end up with a formidable support tool, or so the reasoning goes. This paper challenges this narrow understanding of uncertainty in LLM-augmented judicial practices. Rather than seeing uncertainty as something to be eliminated or quantified by judicial support tools, we argue that some non-quantifiable forms of uncertainty play a vital role in supporting moral perception. While the significance of moral perception to judicial processes has been amply demonstrated, less attention has been paid to how perceptual intelligence depends on the ways in which uncertainty is allowed to surface within institutional infrastructure. Contemporary research on LLMs’ communication of uncertainty offers a unique opportunity to examine this under-appreciated relationship in concrete terms. Can we design LLM-powered augmentation tools that express uncertainty in ways that enhance rather than diminish the relevant forms of moral perception? This question takes on particular urgency given the efforts invested in tackling the epistemic calibration challenges that stem from LLMs’ opaque ‘epistemic façade’. Most of these efforts center on developing methods to quantify and calibrate the reliability of LLM outputs. While such methods may help guard against unwarranted epistemic confidence, these quantitative approaches are unlikely to capture the types of uncertainty that play a key, enabling role within moral perception.

The integration of large language models (LLMs) into our conversational infrastructure presents a critical inflection point for democratic practice. While contemporary digital platforms systematically erode transitional conversational spaces—interfaces between private intuition and public deliberation where tentative thoughts can be explored—this paper argues that specialized LLM interfaces could potentially reconstruct these essential democratic environments. I propose a design framework for ‘transitional conversational spaces’ that leverages uncertainty expression not merely to prevent unwarranted epistemic confidence but to create communicative environments conducive to democratic capability development. Drawing on theories of democratic deliberation and moral perception, this paper distinguishes between epistemic uncertainty (addressable through additional information) and hermeneutic uncertainty (concerning the inherently contestable nature of interpretation). The proposed framework emphasizes ‘ensemble interfaces’ that make visible the contingent nature of value judgments by presenting outputs from multiple models trained on different datasets. The design principles outlined challenge tokenistic participation by advocating for substantive participatory infrastructure with features like ‘tinkerability’—enabling communities to experiment with system configurations—and mechanisms that counter designer-centric development models. These principles stand in contrast to conventional ‘participatory AI’ approaches that treat engagement as merely instrumental to system optimization rather than as constitutive of democratic practice. This paper does not claim to solve all challenges of democratic participation but rather identifies one valuable design direction that could potentially enhance our collective capacity for exploratory dialogue. Implementation would require institutional transformations that align technological development with democratic values beyond current procedural approaches to AI governance.

Just like other experts, members of the professions develop their craft thanks to a deep internalisation of both complex cognitive structures and a mix of habits and intuitive understandings. These non-cognitive aspects of expertise can be what distinguishes the merely competent from the truly brilliant. Yet habits can also be what makes us blind to important features of the world we inhabit. In the life of a professional, these features include the vulnerability of those seeking her services, which in turn grounds the professional's particular ethical responsibility. This book develops an in-depth account of habit to understand its impact upon the way moral decisions are made in a professional context. Its central thesis is the following: what most often stands in the way of a professional meeting her ethical responsibility is not so much stupidity (or character defects) but rather the deleterious aspects of habituation. This book calls for renewed attention to be paid to habits and their relationship to ethical agency. Mostly neglected in moral and legal theory, such an inquiry not only conditions an adequate understanding of the risks inherent in a legal system's institutional structure. It is also essential if we are to come to grips with the challenges raised by the professions' growing reliance upon automated systems."--Publisher description.

Lady Lovelace’s notes on Babbage’s Analytical Engine never refer to the concept of surprise. Having some pretension to ‘originate’ something—unlike the Analytical Engine—is neither necessary nor sufficient to being able to surprise someone. Turing nevertheless translates Lovelace’s ‘this machine is incapable of originating something’ in terms of a hypothetical ‘computers cannot take us by surprise’ objection to the idea that machines may be deemed capable of thinking. To understand the contemporary significance of what is missed in Turing’s ‘surprise’ translation of Lovelace’s insight, one needs to distinguish between trivial surprises and those events, propositions or encounters that lead us to question our understanding of ourselves or what surrounds us. Only some of these non-trivial surprises are the product of originality endeavours. Not only is it uncommon for surprises to track such endeavours, the type of autonomy that would be required on the part of ‘digital computers’ for originality and surprise to intersect in that way goes far beyond the operational autonomy that can be achieved by ‘learning machines’. This paper argues that a salient translation of Lovelace’s originality insight—for contemporary and future ‘learning machines’—is an upside-down version of Turing’s surprise question: can computers be surprised by us in a non-trivial, ‘co-produced’ way?

This essay aims to contribute robust grounds to question the Susskinds’ influential, consequentialist logic when it comes to the legitimacy of automation within the legal profession. It does so by questioning their minimalist understanding of the professions. If it is our commitment to moral equality that is at stake every time lawyers hail the specific vulnerability inherent in their professional relationship, the case for wholesale automation is turned on its head. One can no longer assume that, as a rule, wholesale automation is both legitimate and desirable, provided it improves the quality and accessibility of legal services. The assumption, instead, is firmly in favour of designing systems that better enable legal professionals to live up to their specific responsibility. The rest of the essay outlines key challenges in the design of such profession-specific, ‘ethics aware’ decision-support systems.

In his review of Bernard Williams' Ethics and the Limits of Philosophy, Hart eloquently formulated an apprehension that still haunts much of contemporary jurisprudence: if the moral 'I must' has to be 'seen as coming not from outside, but from what is most deeply inside us? the fear is that this will not be enough'. I argue that this fear is the byproduct of the dualist outlook within which Hart—and a significant part of contemporary legal theory—is confined: because of his bald naturalist premise, Hart could not conceive of moral objectivity except in terms presupposing an order of Reason resolutely distinct from the 'natural' world. This paper seeks to debunk this dualist outlook by engaging with the kind of 'non-bald' naturalism advocated in different ways by both McDowell and Blackburn. In considering contemporary efforts to draw a middle way between ethical scepticism and metaphysical rationalism, the paper draws on the pragmatic elements emerging from the confrontation between Habermas and Rawls.

In his review of Williams' Ethics and the limits of philosophy, Hart eloquently formulated an apprehension that still haunts much of contemporary jurisprudence: if the moral 'I must' has to be 'seen as coming not from outside, but from what is most deeply inside us [...] the fear is that this will not be enough.' I argue that this fear is the byproduct of the dualist outlook within which Hart - and a significant part of contemporary legal theory - is confined: because of his bald naturalist premisses, Hart could not conceive of moral objectivity except in terms presupposing an order of Reason resolutely distinct from the 'natural' world. In view of the heavy metaphysical assumptions needed to sustain such a rationalist outlook, Hart surrendered instead to a cautious agnosticism. This paper proceeds in two steps. A first it seeks to debunk this dualist outlook by engaging with the kind of 'non-bald' naturalism advocated in different ways by both McDowell and Blackburn. Subsequently considering contemporary efforts to draw a middle way between ethical skepticism and metaphysical rationalism, this paper draws on the pragmatic elements emerging from the confrontation between Habermas and Rawls. Together with the critique of 'physicalist' naturalism, I take these pragmatist insights to be key to conquering a traditional reluctance when it comes to accounting for law's axiological dimension.

Hart's and Kelsen's respective outlooks on the concept of normativity not only differ by the way they explain this concept but also, more importantly, in what they seek to achieve when endeavouring to account for the normative dimension of law. By examining Hart's and Kelsen's models in the light of Korsgaard's understanding of the “normativity problem,” my aim is to emphasise not only their contrasted perspectives, but also the common limit they impose on their theories by dismissing as inappropriate any question regarding the emergence of legal normativity. On the basis of my previous arguments, I shall explain why I deem Raz's analysis of the contrast between Hart's and Kelsen's conceptions of normativity to be misleading.

This book offers a 'genealogical' explanation of law's normativity. The term 'genealogical' conveys a commitment to a non-metaphysical type of enquiry. While it explains how law, as a normative phenomenon, comes about, it does not seek to ground law's normativity in anything but the context of social interaction giving rise to it. Legal normativity is brought about on a daily basis. Whether in revolutionary circumstances or in the quotidian need for judges, lawmakers or citizens to balance law's demands with those of morality or prudence, our ability to bind ourselves through law ultimately depends on our capacity to articulate a better way of living together, and to commit ourselves to it. These efforts of assessment and articulation depend, in turn, on our conception of normative agency. Assert the need to trace the truth of ethical judgments to some independent moral 'facts' conditioning their objectivity, and you will get a different understanding of what it is we are doing when we dispute law's authority in the name of moral values. Tracing the truth of moral judgements back to our own social practices not only affects the nature of disagreement; it also dramatically increases our responsibility when, as lawmakers, judges, or citizens we 'take the law into our own hands' and confront it with our moral expectations.

The aim of this paper is to underline the relevance of Schmitt's critique of Kelsenian normativism in the context of today's debate about the status of legal positivism. Schmitt's underlining of the limits which a certain kind of positivism imposes upon itself highlights a contemporary issue about what legal theory should aim at when accounting for the normative dimension of law. Schmitt's ultimate failure to take up the theoretical challenge he himself raised (with its well‐known consequences) is deemed to illustrate—negatively—the importance of providing a plausible account of the social practices which bring law into existence.

Can words – rather than a State – constitute a country? It may be made of land, rivers, forests or deserts – yet, without its inhabitants’ words, there would be no map to draw, no tale to sing, no country to speak of. Palestinian tales abound. They speak of departed lands, vanished homes, forfeited livelihoods. They lament internal wrangling, squeal occupational anger, seek to whisper away those quotidian checkpoint humiliations. Yet, they also speak of hope. If there ever were such a thing as “authoritative hope”, the ongoing Palestinian constitution drafting process may be it. But hope cannot be formalized, let alone authorized. And there is some danger in pretending otherwise.