Skip to main content

Hal & Carla

Hal, a diabetic, loses his insulin. Before collapsing, he enters Carla's house and uses some of her insulin. Carla is another diabetic, not home. Should Hal be punished?

— Trevor Bench-Capon, introducing value-based argumentation frameworks (2003). A widely-used worked example for reasoning about values.

The arguments

Hal & Carla — value-based attacks
attacksdrag nodes · scroll to pan
H1 and C1 attack each other symmetrically. C2 neutralises Hal's compensation argument H2 by showing Carla was also endangered. H2 in turn attacks C1 — if Hal cannot pay, property-rights-as-remedy dissolves.

Why values matter

Pure Dung semantics can't resolve this — symmetric attacks give you multiple extensions. But you can attach values to arguments: H1 promotes life, C1 promotes property. Different audiences with different value orderings reach different stable positions rationally:

  • An audience that ranks life > property → H1 accepted, C1 rejected.
  • An audience that ranks property > life → the opposite.

Bench-Capon's VAF framework is the machinery that makes "different audiences, same framework, different conclusions" formally precise.

In our library

Value-based argumentation is on the roadmap for a future crate (argumentation-values). Currently our argumentation-schemes supports a related mechanism: PracticalReasoning schemes carry a value dimension in their bindings, and encounter-argumentation's StateActionScorer can be composed with a value-aware inner scorer to produce audience-conditioned outcomes.

Further reading