Handing the decision to a convened public

Definition · Jürgen Habermas, Between Facts and Norms (1992)

This concept locates legitimacy in the exchange of reasons among equals: participants justify their views to each other, free of coercion, and remain able to change their minds. Counting votes or averaging preferences does not produce it, because the reasoning is what does the work.

Analysis

The collective-dialogue papers (NEW-E1, NEW-E2, NEW-S152) do put participants in iterative rounds where they read and react to statements other participants wrote, and NEW-S152 sustained that for four months across the hardest possible divide — that is more than a poll. But every procedure here decides by score rather than by reasons: agreement rates in AI-E3's Polis run, bridging-based ranking in the dialogue papers, per-side agreement percentages in NEW-S152, ideal-point distances in NEW-S151, and raw rating counts in NEW-S43, where the drawn panel rates outputs exactly as a convenience sample would. Nobody answers anybody's objection, no reason travels with a vote, and none of these papers reports that a participant revised a view. The half that is carried is co-presence and mutual visibility; the half that is missing is the reason-giving Habermas says does the actual legitimating work.

Definition · Jean-Jacques Rousseau, The Social Contract (1762)

This distinction separates what is good for a public in common from the sum of what its members privately want. Aggregating private wants, however fairly, does not produce the former, since the two can and often do diverge.

Analysis

Bridging-based ranking in the dialogue papers is a deliberate refusal to take the raw majority: a statement one faction rejects will not surface however many others endorse it, and NEW-S152 holds out for roughly 84% on each side rather than 84% overall. That distinguishes broadly-holdable positions from popular ones, which is a real step past summing. What is computed is still an intersection of privately held positions, though — NEW-S43 uses a lottery to make the sum representative of what the population privately wants, NEW-S151 aggregates submitted ideal points, and no step in any paper asks a participant what would be good for the body in common rather than what they personally endorse. Common ground among private wants is the object Rousseau distinguishes the general will from.

Definition · John Rawls, A Theory of Justice (1971)

This device requires that rules be chosen without knowledge of which position the chooser will occupy under them, e.g., rich or poor, majority or minority. Not knowing generally pushes the chooser to protect the worst-off position, since it may turn out to be their own.

Analysis

Every procedure in this cell has people act from a position they know and can see the consequences of: Polis participants vote on principles in AI-E3, panelists rate outputs in NEW-S43, community members submit their own ideal points in NEW-S151, and NEW-S152's participants are constituted by which side of the conflict they belong to. Sortition randomizes who does the choosing, not what the chooser knows; it blocks anyone from rigging the composition but leaves each drawn person fully informed of their own stake, which is the opposite of the device's operative condition. No step anywhere asks a participant to pick a rule without knowing where they would stand under it, and none of the built systems reweights toward a worst-off position.

Definition · T. M. Scanlon, What We Owe to Each Other (1998)

This test holds a principle justified only if no individual could reasonably reject it, and it is applied person by person rather than in aggregate. One sufficiently strong objection therefore outweighs many mild preferences, which is the case averaging handles wrongly.

Analysis

Bridging-based ranking and NEW-S152's per-side agreement floors give a bloc something close to a veto: a statement one side rejects fails no matter how enthusiastic the other side is, which is the anti-aggregative shape Scanlon's test has. The unit and the currency are both wrong for the concept, though — the veto belongs to a predefined group rather than to a person, it is triggered by headcount rather than by the strength or reasonableness of the objection, and a 16% minority inside a side is simply outvoted. NEW-S151 runs the other way explicitly: laws and constitution alike fall to a large enough supermajority, so nothing is placed beyond reach of numbers. Nothing at inference tests a candidate model action against whether any individual could reasonably reject it.

This supplies an end-to-end mechanism by which public input changes a model's weights, which no political tradition provides, since none of them contemplated training. The reservations above concern what passes through that mechanism rather than the mechanism itself.

Sortition is a way of constituting a body that will then deliberate. Applying it to a rater pool constitutes a body that will not deliberate at all — the democratic mechanism is used to fix the sample rather than to produce a judgment, which is a use citizens'-assembly practice has no counterpart for.