← Face C2: Universal Subjective Convergence Implies Objectivity
Face C3: The Steelman Is the Test (Draft)
Steelman first. Test what survives.
Before you reject a position, be able to state it so accurately that its defenders would recognize it. A claim that survives the strongest available opposition is strengthened by having survived. A claim that doesn’t should be revised or abandoned. The goal is a framework more credible after scrutiny, not one that requires looking away from the weak points.
Most argument is wasted because people argue against the weakest version of the opposing view. The steelman requirement is different: before evaluating any major claim, construct the strongest version of the opposing argument — not the version you can most easily defeat, but the version that most pressures your own position. Then evaluate your view against that.
This is harder than it sounds. Most people mistake understanding the surface of a position for understanding the position. The steelman test reveals the difference: can you state the opposing view so well that someone who holds it would say, yes, that’s exactly what I mean? If not, you haven’t engaged the position — you’ve argued with something constructed to be defeated.
The steelman is also self-applying. Before defending your own position, apply the strongest available criticism to it. A position that survives this pass is stronger for having survived it. The willingness to revise is not weakness — it is the difference between a framework actually trying to be true and one merely trying to be consistent.
In practice:
You and your partner are arguing about whether to move cities. You’ve been making the same points for thirty minutes and neither of you is listening.
Stop. Before your next response, state their case in a way they would recognize as accurate — not as a rhetorical move, but as a genuine attempt to understand what they actually believe. Not “you think your career matters more than mine” — that’s the uncharitable version. What do they actually think? What’s the real concern?
When you can state their position so well that they say yes, that’s it — then respond. What you’ll find, often, is that the argument you’ve been having isn’t the real disagreement. The real disagreement is smaller, or different, or addressable in a way the surface argument wasn’t. The steelman clears the underbrush.
Formal Statement (Concordius Framework)
Worst-case warrant. Let 𝒜 be the set of admissible counter-arguments to a claim C — conditioning operators that lower its warrant. Define the robust (steelman) warrant ρ(C) = inf_{a∈𝒜} τ(C | a). A strawman substitutes a weak a′ and reports τ(C | a′) ≥ ρ(C), overstating warrant; the steelman evaluates at a* = argmin, returning the guaranteed lower bound. “Test what survives” is then exact: ρ(C) is the worst case, so a claim with ρ(C) above threshold holds no matter which admissible opponent is raised. The framework reads the discipline as noise-floor reduction before reception (Appendix F): the evaluator must be tuned finely enough to detect the opponent’s actual lower-constraint content P_Φ a*, not its H₄₈ surface — otherwise the infimum is taken over a misread set and ρ is meaningless.
Tier: criterion — rigorous as min-max (distributional) robustness; the open part is specifying 𝒜, the true strongest opposition, which is interpretive. Shares the operator-A caveat (Face A0).