The Idea of Justice
The Idea of Justice is Sen's 2009 treatise on political philosophy, advancing a comparative approach to justice as an alternative to the transcendental institutionalism of John Rawls and his successors. Rather than asking 'what would a perfectly just society look like?', Sen argues that the urgent question is 'what injustices are currently remediable?' — a question that can be addressed through public reasoning and comparative judgment without requiring agreement on ultimate ideals. The book extends the capability approach into political philosophy and provides the analytical foundations for applying capability theory to AI governance.
In the AI Story
The book's central methodological move is the distinction between transcendental and comparative approaches to justice. Transcendental approaches — Rawls's most prominent — attempt to specify the ideal just society and judge actual arrangements by their distance from the ideal. Comparative approaches ask which of two or more actual arrangements is more just, without requiring specification of an ultimate ideal. Sen argues that the comparative approach is both more tractable and more useful: it focuses attention on injustices that can actually be removed, rather than on ideals that may never be achievable and may not even be uniquely specifiable.
The book develops the concept of public reasoning as the mechanism through which societies arrive at comparative judgments about justice. Public reasoning requires that evaluations be made through open, inclusive, informed deliberation — not by experts in isolation, not by markets in aggregate, but by citizens engaged in the difficult work of arguing about what matters and how to achieve it. The commitment to public reasoning is a feature of Sen's framework, not a limitation: it means that the selection of relevant capabilities, the identification of priority injustices, and the design of remediation strategies belong to democratic deliberation.
The book explicitly connects the capability approach to the broader tradition of political philosophy, engaging with Adam Smith, Condorcet, Mary Wollstonecraft, Karl Marx, and John Stuart Mill. Sen situates his framework within what he calls the 'realization-focused' tradition of thinking about justice — a tradition concerned with the actual lives people lead rather than with the abstract design of ideal institutions.
Applied to AI, the book's framework produces a specific methodological orientation. The question is not 'what would a perfectly just AI-governed society look like?' — a question that may be impossible to answer and that delays action while debate continues. The question is 'what AI-related injustices are currently remediable, and how?' This orientation produces tractable policy objectives: expand access to conversion factors, build transparency infrastructure, strengthen democratic deliberation over AI deployment, provide protective security for those displaced by the transition. Each objective is testable, comparable, and within the reach of public reasoning.
Origin
The book developed from Sen's long engagement with Rawlsian theory of justice and from lectures delivered at New York University, Harvard, and elsewhere. Published by Harvard University Press in 2009.
Key Ideas
Comparative over transcendental. Focus on removing remediable injustices rather than specifying the perfectly just society.
Public reasoning. The mechanism through which societies arrive at collective judgments about justice.
Realization-focused justice. Evaluation of actual lives lived rather than ideal institutional designs.
Plural impartiality. Justice requires considering perspectives from outside one's own community, not just within it.
Appears in the Orange Pill Cycle
Further reading
- Amartya Sen, The Idea of Justice (Harvard University Press, 2009)
- John Rawls, A Theory of Justice (Harvard University Press, 1971)