Use Case · Discourse

Discourse as a measurable flow.

Public discourse currently runs on engagement metrics. Clicks, watch-time, reactions. Those metrics optimize for two regimes: the Comfort regime on the safe side, and the Battlefield regime on the loud side. They optimize for neither sustained amplification of durable ideas nor low degradation of shared reality.

The Proposal

Discourse is a flow. Each utterance has measurable Amplification (constructive content emission, idea-flourishing, the kind of contribution that increases an audience's capacity to reason) and measurable Degradation (contempt, ad hominem, falsehood, the kind of contribution that reduces audience trust or distorts shared reality). The same M = τF + A accounting that the framework applies to interpersonal flows applies to communicative ones.

Two axes give four named regions. Top-left, the Cathedral: high amplification, low degradation. Top-right, the Battlefield: high amplification, high degradation. Bottom-right, the Toxic: low amplification, high degradation. Bottom-left, the Empty (or Comfort): low amplification, low degradation, the surface-civility regime where the absence of friction is mistaken for the presence of substance.

Topologies (Battlefield, Cathedral, Marketplace, Hearth, Tribunal) reweight the same utterance differently. A claim that would be heard as confrontation in a Hearth topology may be heard as ordinary contention in a Battlefield topology. The framework's reporting modes distinguish the topology-weighted M from the unweighted base measurement.

Attribution

Discourse-topology measurement is the work of the CCII Research Collective (paper in progress). This use-case page summarizes the public-facing implication of that work. The framework primitives (F, A, D) are inherited from Transmutarianism core; the discourse extension adds the topology weighting and the explicit Degradation term.

The Tool

The discourse evaluator at /discourse-evaluator/ implements a minimal public-facing version. Eight historical exemplar debates are plotted on the Amplification by Degradation quadrant. Click any dot for the excerpt, scoring, and rationale. Paste your own text to score it.

The v1 scorer is a transparent keyword-and-structure heuristic: ad hominem terms increase D, sentence-length variance increases A. The keyword list and weights are visible in the page source. The scorer does not understand context, sarcasm, or rhetorical irony. A future version may use a calibrated model.

Open the discourse evaluator

What This Requires

• Shared scoring conventions. The v1 keyword heuristic is one possible operationalization. Robust deployment needs scoring conventions that survive cross-cultural and cross-genre transfer. The CCII Research Collective's forthcoming paper specifies the operational form.
• Open transcript datasets. Public-domain transcripts (pre-1928 in the US) are usable directly. Post-1928 excerpts require documented fair-use rationale for short quotation. The eight exemplars on the evaluator follow this convention.
• A public calibration set. A shared set of texts with consensus scoring is the basis for any future calibrated model. The eight exemplars are a starting point, not a finished calibration.
• Refusal of Comfort. The framework's most important guardrail against discourse-quality optimization: a discourse environment that reads as low-D because it does not measure D is not safe. It is unmeasured. The Comfort state must be distinguishable from the verified-safety state. This is the same guardrail the framework imports into smart-city accounting.

Next

Try the evaluator, read the next case, or return to the framework.