Contents
A theory with a similar name
The name of constitutive pluralism resembles that of constitutional pluralism, a theory in legal scholarship of how several orders that each claim constitutional authority, such as the European Union and its member states, relate to one another. The two are separate theories.
Political philosophy
The right to justification: Rainer Forst
Constitutive pluralism starts from the premise that those who make decisions that bind people owe reasons to the people who must comply with them. This premise is close to the political philosopher Rainer Forst’s account of a basic right to justification. Both treat each person as someone who can demand and examine the reasons for a decision.
Non-domination: Philip Pettit
The pillar of non-domination extends the conception of non-domination developed by the political philosopher Philip Pettit, which asks whether those subject to an intervention can effectively control how it is carried out. Constitutive pluralism applies this requirement to the field of relationships between people, and it requires that no agent be able to read, rearrange, or intervene in that field arbitrarily. A complaint channel is not enough. Those affected need a route that can lead to an examination of the grounds, a change of the decision, or, where necessary, a halt.
Basic liberties: John Rawls
The philosopher John Rawls’s account of the basic liberties, such as liberty of conscience and the political liberties, addresses the capacity to form, revise, and pursue a conception of the good, that is, a view of what makes a life worthwhile. It also treats the resources and institutional conditions of that capacity as relevant. The protection of how people form what they value therefore already belongs to this tradition. The question that constitutive pluralism adds is how such demands operate when evidence, categories of eligibility, and the time needed to contest decisions depend on several interacting systems.
Complex equality: Michael Walzer
Non-aggregation shares a concern with the political philosopher Michael Walzer’s theory of complex equality. On this theory, goods form spheres, each with criteria of its own. Advantage in one sphere should not command goods in another, as when money buys political power. Non-aggregation forbids using a comprehensive evaluation across contexts to decide a person’s entitlements and treatment. Such a use is a wrong of the same kind, because it decides a person’s claim in one context by a ranking of the person across contexts instead of by the purposes of that context.
Value pluralism: Isaiah Berlin
The political philosopher Isaiah Berlin rejected grading all values on one scale. Constitutive pluralism resembles this view in rejecting a single ranking of values fixed for a whole society by one measure. It rejects such a ranking because the ranking removes the conditions under which those affected can question and choose again.
Agonistic pluralism: Chantal Mouffe
The political theorist Chantal Mouffe’s agonistic pluralism holds that confrontation between adversaries, who accept each other’s right to contend, rather than enemies, who seek to eliminate each other, is a condition of democracy’s existence. It also holds that every consensus is a temporary stabilization of power that entails some exclusion. Constitutive pluralism permits a reasoned final decision in a particular case without unanimity among those affected, and it assesses that decision in part by whether they could be heard and can still obtain warranted revision. Mouffe’s concern extends to decisions about whose objections count, which are themselves exercises of power. Constitutive pluralism treats those decisions as open to challenge. This responds to exclusion, but Mouffe’s concern remains, because challenge does not remove that power.
Information and evaluation
Contextual integrity: Helen Nissenbaum
The information ethicist Helen Nissenbaum’s contextual integrity judges a flow of information by the norms of the context in which it occurs, and it already addresses aggregation, profiling, and power in information flows. Within that scope, non-aggregation singles out one use for prohibition, which is a comprehensive evaluation of a person across contexts that decides how the person is treated.
Governance and institutional design
Spontaneous order: Friedrich Hayek
The economist and social philosopher Friedrich Hayek argued that no central planner can gather the knowledge dispersed through a society, and he defended spontaneous order, which arises from people’s interactions without a designer of the whole. Constitutive pluralism draws on this argument. A principle that prescribes institutions may therefore seem to repeat the deliberate design of social order as a whole that Hayek warned against. However, the separation of evaluation across domains, veto rights, approval of exceptions, and rules that lapse unless renewed are boundary conditions that preserve what cannot be planned, rather than a plan for society as a whole. Whether a concrete design of these conditions turns into comprehensive control depends on how it is carried out. Once AI can aggregate dispersed information, protecting spontaneous order requires designing these conditions explicitly.
Polycentric governance: Elinor Ostrom
The political economist Elinor Ostrom studied polycentric governance, in which several centers of decision interact, participants help make and modify rules, monitoring is accountable, and dispute resolution is accessible. For the age of AI, constitutive pluralism translates this order into the requirement that no single AI stack, which combines a foundation model, training data, and a design philosophy, monopolize the medium through which people collaborate. Such a monopoly is objectionable where it removes the effective control of those affected or their grounds for objection.
Democratic experimentalism: Charles Sabel and William Simon
In the democratic experimentalism of the legal scholars Charles Sabel and William Simon, goals, performance measures, and decision procedures are revised through local experimentation, reporting, and review. Constituencies that emerge during implementation can also take part and raise challenges. Revising goals and admitting new participants therefore already belong to these theories. Three questions remain. The first concerns burdens of revision that accumulate across institutions. The second concerns people’s need for time outside an institution’s program of continual revision. The third concerns conflicts between continuing and changing an arrangement. Constitutive pluralism answers the first and third through autonomy of pace and the second through slack. Where existing theories answer them through their own commitments, its contribution to these protections narrows to synthesis and application.
Research and debate on AI
Gradual disempowerment: Jan Kulveit and colleagues
The AI researcher Jan Kulveit and colleagues examine gradual disempowerment, a process through which humans lose influence over institutions. They separate the local compliance of AI systems with human instructions from the preservation of human influence across economic, cultural, and political institutions. Constitutive pluralism shares this separation and supplies its own normative reasons for it.
The legitimacy of algorithmic decisions: John Danaher
The philosopher of law and technology John Danaher argues that reliance on algorithmic systems threatens the legitimacy of public decision-making, that is, whether people have reason to accept those decisions, because these systems constrain human participation in, and comprehension of, those decisions. He weighs the costs of opacity against the benefits that the systems bring. Constitutive pluralism shares the concern with participation. The form of participation in collective decisions and any specific power to decide need further reasons. The core of standing, however, which consists of protection, reasons, and a route to independent review, is not weighed against ordinary benefits.
Contestable AI: Kars Alfrink and colleagues, and Margot Kaminski and Jennifer Urban
The design researcher Kars Alfrink and colleagues argue that AI systems should be contestable by design, responsive to human intervention throughout their lifecycle. The legal scholars Margot Kaminski and Jennifer Urban argue for an individual right to contest AI decisions, modeled on due process, the principle that people facing an adverse decision are entitled to procedural protections such as notice and a hearing. This right can extend across contexts. Constitutive pluralism shares with Alfrink and colleagues the view that contestation should reach design. Alongside these accounts, it examines burdens that accumulate on people across several institutions, none of which is defective when assessed alone, without claiming that either account omits them.
Pluralistic alignment: Taylor Sorensen and colleagues
Alignment research seeks to make the behavior of AI systems accord with human values and intentions. In a roadmap to pluralistic alignment, the AI researcher Taylor Sorensen and colleagues distinguish aims such as representing a range of reasonable answers and responding from specified perspectives. The roadmap does not purport to settle whom or what an AI system should serve. Even when an AI system can present many perspectives, a single institution may still hold the power to choose its objectives, appoint representatives, and decide which objections count. Non-domination concerns that power.
Plurality and civic AI: E. Glen Weyl, Audrey Tang, and colleagues
In the book Plurality: The Future of Collaborative Technology and Democracy (2024), its authors, E. Glen Weyl, Audrey Tang, and the Plurality Community, describe the idea behind its title as “technology for collaboration across social difference.” Constitutive pluralism asks what standing the participants in such collaboration, and others affected by it, should have.
Tang and Caroline Green propose a framework called civic AI, which seeks “AI that the people it affects can check, correct and switch off.” Its refusal to blend complementary checks into one score shares a concern with reducing evaluation to a single measure. Non-aggregation more specifically forbids using a comprehensive evaluation of a person across contexts to decide entitlements or treatment. The two frameworks also converge in their concern with keeping AI correctable. Civic AI affirms the right of communities to fork its tools, that is, to copy them and develop their own versions, and to rebuild elsewhere, and the people served by an AI model for their place agree on a date for ending or review and on who decides what follows. Constitutive pluralism explicitly extends the demand for reasons and review to any continuation that binds others or brings about unavoidable losses of what matters, including the continuation of a community itself. The two frameworks can therefore be read together.
First published: 12 November 2026. Last updated: 12 November 2026.