Defeater Register¶
Register — instantiates Evidentiary Trace Warranting
Lists facts, conditions, and counterevidence that would weaken or defeat an evidence relation.
A Defeater Register is a standing, maintained list of the specific things that would knock down a given piece of evidence — the facts that, if true, would either rebut the claim or cut the trace's support for it. Its defining trait is that it inverts the usual bookkeeping: instead of cataloguing reasons to believe, it catalogues, in advance and by name, the reasons the relation might fail. Each entry is a concrete, checkable condition ("if the sensor was recalibrated after this reading, the reading no longer supports the claim") together with what its arrival would do to the relation. Because it is a register — a living artifact, not a one-time check — it doubles as a tripwire list: an entry that later comes true is a signal to re-open the evidence, not just a footnote.
Example¶
A reporter is building a story that a city official steered a contract to a friend's firm. The central trace is a set of emails showing the two arranged a dinner the week the bid closed. Before publishing, the editor has the reporter open a defeater register: a written list of everything that would sink or soften the story if it turned out to be true.
The entries are specific. If the dinner was a routine constituent event with a public calendar entry, the "secret coordination" reading collapses. If the firm was the only qualified bidder, "steering" loses its force even if they were friendly. If procurement was handled by a committee the official didn't sit on, the causal link is cut. The reporter also logs a negative-evidence entry: we have found no competing bids in the record — but have we actually searched the archive, or only failed to receive them? That distinction matters, because "no rival bids found" is only a defeater-of-a-defeater if the search had the power to find them. The register turns a gut sense of "what could go wrong" into a checklist the newsroom can actually work down before the story runs.
How it works¶
The register is a table with one row per defeater and a small, fixed set of columns:
- The defeater — a concrete fact or condition, stated so it is checkable, not a vague worry.
- Type — whether it rebuts the claim directly (the claim is false) or undercuts the trace (the trace no longer supports the claim), a distinction drawn from defeasible-reasoning theory.[n1]
- Effect if true — how much of the relation it removes: fatal, downgrading, or scoping.
- Status — open, checked-and-ruled-out, or triggered.
A distinct class of row is reserved for absence claims: places where the argument leans on something not being observed. Each such row must record whether a real search was done and whether that search had the power to detect the thing, so "we didn't see it" is never silently promoted to "it isn't there." Maintaining the register means revisiting statuses as facts come in and flagging any entry that flips to triggered.
Tuning parameters¶
- Enumeration effort — how hard you hunt for defeaters up front. Exhaustive lists catch more but cost time and can breed paralysis; thin lists are fast but leave blind spots.
- Rebut/undercut discipline — whether entries must be typed. Typing sharpens the response (a rebuttal needs a counter-fact; an undercutter needs a provenance fix) but adds overhead.
- Detection-power bar for absences — how strong a search must be before a "not observed" row counts as ruled out. A high bar guards against false absence at the cost of investigation effort.
- Refresh trigger — whether the register is reviewed on a schedule, on new information, or only before a decision.
- Ownership — whether one skeptic keeps it or the whole team contributes; adversarial ownership finds more defeaters but can turn combative.
When it helps, and when it misleads¶
Its strength is that it operationalizes disconfirmation. Confirmation bias makes people collect support and ignore the exits; a register makes the exits an explicit, assignable to-do list, and its absence rows directly counter the "no signal means no effect" error by forcing a detection-power question. It is most valuable exactly when a case feels strong, because that is when nobody looks for what would break it.
Its failure mode is that a register lists defeaters but does not weigh them: a long column of ruled-out worries can lend false confidence, since the ones you never thought to write down are the ones that actually sink you. The classic misuse is treating an unfilled register as proof of robustness — "we couldn't think of any defeaters" is a statement about imagination, not about the world. The guarding discipline is to keep the register adversarial (invite someone whose job is to break the case) and to treat every triggered row as a mandatory re-open rather than a note to revisit later.
How it implements the components¶
defeater_and_alternative_hypothesis_map— the register is this map made concrete: the enumerated, typed list of conditions and rival explanations that would defeat or reduce the relation.negative_evidence_detector— its absence-claim rows flag every place the argument rests on something unobserved and force a detection-power check before absence is read as evidence.
It does not test whether a live trace actually discriminates among hypotheses (relevance_rule — that's Relevance and Alternative Explanation Check) nor assign the surviving relation a strength label (scoped_evidential_weight — that's Evidence Strength Ladder); the register catalogues what could break the relation, it does not grade what remains.
Related¶
- Instantiates: Evidentiary Trace Warranting — the register supplies the archetype's defeater-and-alternatives slot as a maintained artifact.
- Consumes: Evidence Relation Matrix — the register attaches to the relations recorded there, populating their defeater column.
- Sibling mechanisms: Admissibility or Relevance Gate · Claim-Evidence-Reasoning Card · Evidence Relation Matrix · Evidence Strength Ladder · Evidence Update Review · Relevance and Alternative Explanation Check · Trace-to-Claim Diagram
Editorial Notes¶
Form Classification¶
Form family: Record, Log & Register
Rationale: Defeater Register operates as a durable record, ledger, register, or trace whose value depends on preserving actual state or history because it lists facts, conditions, and counterevidence that would weaken or defeat an evidence relation.
Independent corroboration: The frozen evidence defines Defeater Register as 'Lists facts, conditions, and counterevidence that would weaken or defeat an evidence relation', so its operative form is Record, Log & Register.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Philosophy
Origin pattern: Single lineage
Present-day reach: Multi-domain
Rationale: Epistemology and defeasible-reasoning theory cohered rebutting and undercutting defeaters as distinct ways evidence support can fail.
Related originating lineages:
- Law & Governance — Legal evidence practice developed parallel registers of rebuttal and challenges to evidentiary connection.
Review resolution: Epistemology and defeasible-reasoning theory cohered rebutting and undercutting defeaters as distinct ways evidence support can fail. The retained alternate lineages materially shaped the mechanism's form.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] In the philosophy of defeasible reasoning (notably John Pollock's work), a rebutting defeater is evidence that the conclusion is false, while an undercutting defeater attacks the connection between the evidence and the conclusion without showing the conclusion false. The register's Type column preserves this distinction because the two demand different responses. ↩