Burden-of-Proof Rule¶

Allocation rule — instantiates Regress Termination Rule

Allocates who must justify what, and to what standard, before a claim is accepted or reopened — so a support chain closes by a fixed rule of responsibility rather than by whoever argues longest.

A regress of "prove it" can run indefinitely because nobody has said whose job it is to prove and what kind of proof settles it. A Burden-of-Proof Rule ends that by allocating responsibility in advance: it names which party must make the case, what would discharge that duty, and — crucially — first sorts the demand into its type, since a demand for evidence, a demand for authority, and a demand for a value justification are answered by different parties in different ways. It is not the mechanism that sets how much is enough (a threshold) nor who holds final authority; it is the mechanism that decides who carries the load so the chain doesn't close by exhaustion or volume.

Example¶

A medical-device company is adding a feature that nudges patients to adjust dosing. The review board could argue forever about safety, so it adopts a burden-of-proof rule instead. First it classifies the open questions: is this dose safe? is evidentiary (routes to the clinical-evidence standard); should the device advise dosing at all? is a value question (routes to the ethics committee); are we allowed to ship it? is procedural (routes to regulatory sign-off). Then it fixes the default: the proposing team bears the burden of showing safety to the defined standard before launch — the reviewers are not obliged to prove harm. It also sets the far side of the chain: once shipped, the burden flips — to reopen, a challenger must bring new adverse-event signal, and closure is only valid once the notified-body reviewer (the counterparty) accepts it. The result is that no single meeting decides by fatigue; the rule decides who must do what before the question can close.

How it works¶

Sort the demand by type. Is this call for prior evidence, higher authority, a clearer definition, or a deeper value justification? The rule refuses to let one type stand in for another (authority is not evidence; a cause is not an ought).
Assign the default burden. Name which party must carry the case — proposer or objector — and make the choice explicit rather than ambient.
State what discharges it. Point to the standard that, once met, closes the demand (and hand the sizing of that standard to the sufficiency mechanisms).
Set the reopening burden and the acceptor. Fix how much a challenger must bring to reopen, and whose sign-off makes a closure stick.

Tuning parameters¶

Default allocation — who holds the burden absent argument. A precautionary default (proposer must prove safe) protects against downside but can freeze innovation; a permissive default (objector must prove harm) does the reverse.
Discharge standard — how strong the required showing is. Raising it makes acceptance harder and reopening easier; the two move together.
Reopening-burden height — how much a challenger must produce to force reconsideration. High prevents relitigation but risks entrenchment; low keeps things live but invites churn.
Counterparty set — whose acceptance is required for closure — regulator, client, affected party, or none. Wider sets add legitimacy and slow the close.

When it helps, and when it misleads¶

Its strength is neutralizing "whoever argues longest": by fixing responsibility ahead of the dispute, it prevents both endless demands and quiet rubber-stamping, and by typing the demand it stops a group from answering the wrong question. It shines wherever closure is adversarial or accountable — safety cases, legal and regulatory review, contested approvals.

Its central abuse is that the allocation is itself a lever: placing an impossibly high burden on one side, or an impossibly low one on the other, can predetermine the outcome while looking procedurally clean — the burden set after you know who you want to win. Where a party bears the burden is a value choice, not a neutral fact — the precautionary principle and the presumption of innocence are exactly such choices about who must prove what.^[1] The discipline that keeps it honest is to fix the allocation and its standard before the specific dispute, keep it symmetric to the stakes, and make the choice reviewable rather than tacit.

How it implements the components¶

link_type_distinction — its first move is to type the demand (evidence / authority / definition / value) so the right party and the right kind of answer are engaged.
burden_of_reopening — it fixes who must provide what to challenge a closure after the decision has moved on, guarding against both relitigation and rigid refusal.
counterparty_acceptance_test — it names whose sign-off a closure requires to hold, binding the terminus to an external acceptor rather than the deciding party alone.

It does not set how much support is enough (stopping_criterion → Decision Closure Criteria; sufficiency_threshold → Evidence Sufficiency Rubric) or define which real-world events reopen a decision (reopening_condition → Review Trigger Register).

Instantiates: Regress Termination Rule — it supplies the who-must-prove-what rule the rest of the appraisal leans on.
Sibling mechanisms: Decision Closure Criteria · Governance Authority Chain · Evidence Sufficiency Rubric · Axiom Set · First-Principles Statement · Five Whys with Stop Rule · Research Stopping Rule · Review Trigger Register · Timeboxed Inquiry Record · Assumption Log

Notes¶

Keep it distinct from a Governance Authority Chain: the authority chain says who may finally close a question, while the burden-of-proof rule says who must do the justifying before it can be closed. The two compose — a well-run approval uses both — but conflating them lets an authority end a chain by fiat while appearing to have discharged a burden.

References¶

[1] Legal systems make the burden allocation explicit and deliberately asymmetric — the presumption of innocence puts the burden on the accuser; the precautionary principle puts it on the proponent of a risky action. Both illustrate that where the burden sits is a chosen value commitment, not a neutral default. ↩