Skip to content

First-Principles Statement

Reasoning artifact — instantiates Regress Termination Rule

Names the base premises a group will reason up from for one specific decision, and flags what those premises leave uncertain, so the chain of 'why' bottoms out at a chosen, visible foundation.

When a group keeps asking "but why do we believe that?", the debate can descend forever. A First-Principles Statement ends the descent by writing down, for one specific claim, the handful of premises the group agrees to reason up from — the point below which, for this decision, they will not dig. Unlike an axiom set it is informal and non-generative: it supplies starting premises for human reasoning that still has to do the connecting work, rather than a system from which conclusions fall out mechanically. Its defining discipline is that it pairs each base premise with an honest note of what that premise leaves unresolved — which is precisely what keeps it from collapsing into a mere statement of belief dressed up as bedrock.

Example

A public-health foundation is deadlocked over whether to fund insecticide-treated bed nets or deworming, and every argument bottoms out in a deeper "why" — why that outcome measure, why trust that trial, why weight future lives that way. Rather than keep chasing the regress, they write a first-principles statement for this allocation: (1) we optimize expected health per dollar, measured in QALYs; (2) we treat well-conducted randomized-trial effect sizes as our evidence base; (3) we assume our marginal dollar is not simply displacing another funder's. Each premise carries its residue: "(1) sidesteps equity across regions; (3) is weakly evidenced and could flip the ranking." With the base named, the team can finally reason upward to a decision — and anyone who disagrees now has a precise target ("I reject premise 1") instead of an endless why.

How it works

  • Fix the claim first. Name the exact decision the premises are for; a first-principles statement drifts into philosophy the moment it forgets which question it is grounding.
  • Strip to the base. Push each supporting reason downward only until you reach premises the group is willing to simply accept here, and stop.
  • State premises as chosen, not proven. Each is written as a stipulation the group adopts, phrased so it can be pointed at and rejected.
  • Attach the residue. Every premise carries a short note of what it leaves uncertain — the move that distinguishes this artifact from an assertion of faith.

Tuning parameters

  • Decomposition depth — how far down you push before declaring a premise "base." Deeper stripping exposes more hidden commitments but costs time and can over-intellectualize a small decision.
  • Premise count — a few load-bearing premises versus an exhaustive list. Fewer is legible and forces prioritization; more is thorough but dilutes what actually carries the decision.
  • Uncertainty granularity — how finely the residue is tagged. Coarse tags are quick; fine ones show exactly which premise is fragile enough to reopen the whole chain.
  • Consensus requirement — whether every party must accept each premise, or whether contested premises are simply marked as such and carried forward.

When it helps, and when it misleads

Its strength is turning a bottomless "why" into a short, contestable list: it surfaces the premises a group was reasoning from unawares and gives dissenters a precise place to push. It is at its best early in a decision, before positions harden around conclusions.

Its signature misuse is reverse-engineering — writing the "first principles" after the preferred answer, so the statement rationalizes a conclusion instead of grounding one. The related failure is mistaking chosen premises for necessary truths and dropping the residue, at which point it becomes exactly the statement of belief it was meant to replace. First principles are starting points selected for a purpose, not self-evident foundations,[1] and the discipline that keeps the artifact honest is to keep every premise labeled as adopted-and-contestable with its uncertainty still attached.

How it implements the components

  • claim_under_justification — it opens by naming the exact decision the premises are meant to ground, anchoring the regress to a live question.
  • foundational_assumption — the listed premises are the provisional base: visible, named, and offered as contestable stipulations.
  • uncertainty_residue — each premise carries what it leaves unresolved, so action proceeds without pretending the deeper questions vanished.

It does not distinguish which kind of support each demand needs (link_type_distinction → Burden-of-Proof Rule), calibrate an evidence threshold (sufficiency_threshold → Evidence Sufficiency Rubric), or fence a formal system (decision_scope_boundary → Axiom Set).

  • Instantiates: Regress Termination Rule — this is the informal, premise-naming instance of the pattern for a live human decision.
  • Sibling mechanisms: Axiom Set · Decision Closure Criteria · Evidence Sufficiency Rubric · Burden-of-Proof Rule · Five Whys with Stop Rule · Governance Authority Chain · Research Stopping Rule · Review Trigger Register · Timeboxed Inquiry Record · Assumption Log

Notes

A First-Principles Statement only terminates the regress when paired with reopening logic; on its own it fixes a base but says nothing about what would unfix it. In practice it is most trustworthy when its premises are logged for future review — the job of Assumption Log — and its residue is wired to a Review Trigger Register so a fragile premise can be revisited when evidence lands.

References

[1] In Aristotle's usage, archai ("first principles") are the starting points of demonstration — accepted so that reasoning can begin — not conclusions reached by it. Naming them as chosen starting points, rather than discovered truths, is what keeps the statement contestable. ↩