First-Principles Model¶
Method — instantiates Core Model First
Builds the initial model from fundamental relations, constraints, or causal claims rather than from accumulated details.
Most models are assembled upward — you gather everything known about a system and try to compress it. A First-Principles Model does the opposite. It starts from a handful of relations you are willing to treat as bedrock — a conservation law, a budget identity, a definitional constraint, a causal claim you can defend — and derives the core downward from them. The variables it carries are the ones the fundamentals force it to carry; its central relation is whatever those fundamentals entail. Nothing enters because it was measured or because it looked important; it enters because the premises demand it. That is the one idea that makes it this mechanism: the core is reached by derivation from invariants, not by cataloguing detail.
Example¶
An engineering team is asked whether a proposed electric cargo drone can lift a 5 kg payload for a useful flight. Rather than collecting rotor spec sheets, they reason from fundamentals. Hovering flight requires thrust equal to total weight; the power to produce that thrust scales roughly with weight raised to the three-halves power divided by the square root of rotor disc area; and usable flight time is battery energy (specific energy times battery mass) divided by that power. Three fundamental relations, and a governing link between payload, rotor size, and endurance falls out. Plugging in illustrative numbers, the derivation reports that the machine can hover with the 5 kg load, but only for about six minutes, and — the useful part — that the binding constraint is disc loading, not battery capacity: doubling endurance needs a roughly 40% larger rotor, and a bigger battery barely helps. That conclusion came from the physics, before a single component was priced.
How it works¶
- Name the bedrock. State the laws, identities, and defensible causal claims you will treat as given, and refuse to smuggle in anything else.
- Let the fundamentals pick the variables. Carry only the quantities the premises force; discard the rest.
- Derive the relation. Write what the premises entail as an equation or constraint, and check that its units and dimensions balance.
- Clear the explanatory floor. Confirm the derivation reproduces the known governing behavior of at least one real case before any detail is considered.
- Log every premise. Each idealization becomes an entry future work can challenge.
Tuning parameters¶
- Depth of derivation — how far down toward first principles you push. Deeper is more defensible and more transferable, but slower and more brittle if a premise turns out wrong.
- Choice of primitives — which relations you treat as bedrock versus derived. Fewer primitives make a cleaner model but risk assuming away the very effect that matters.
- Idealization aggressiveness — how much you strip (frictionless, well-mixed, rational actors). More idealization yields a crisper derivation and a wider gap from reality.
- Dimensional-check strictness — how rigorously units are policed at each step; the cheapest error-catcher available.
When it helps, and when it misleads¶
Its strength is a core you can defend without data: it exposes the binding constraint, and because it is built from invariants it transfers to cases no dataset covered. But a derivation is only ever as sound as the premises fed into it. An elegant argument from a wrong or over-idealized starting point is confidently wrong — the physicist's self-mocking "spherical cow."[n1] The classic misuse is to mistake rigor for validation: because the math is clean, the estimate is trusted and never checked against a real case. The guarding discipline is that every derivation must clear an explanatory floor against at least one observed instance, and every premise stays in the assumption register where it can be argued with.
How it implements the components¶
First-Principles Model fills the origin side of the archetype — deriving the spine, not testing or bounding it:
core_variable— the variables are exactly those the fundamentals force the model to carry.core_relation— the central relation is derived as what the premises entail, stated as an equation or constraint.assumption_register— each fundamental premise and idealization is logged explicitly so later work knows what was taken for granted.explanatory_floor— the derivation must reproduce the known governing behavior before any refinement begins.
It does not validate the core against a reference dataset — that belongs to Baseline Model — nor fix the frame and scope of what is covered, which Minimal Causal Diagram supplies.
Related¶
- Instantiates: Core Model First — First-Principles Model is one way to originate the validated core the archetype refines from.
- Sibling mechanisms: Baseline Model · Minimal Causal Diagram · Simple Prototype · Stripped-Down Simulation · Minimum Viable Explanation · Core Architecture Sketch · Toy Model
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: First-Principles Model operates as a computation, comparison, model, or analytic representation used to infer, estimate, or choose because it builds the initial model from fundamental relations, constraints, or causal claims rather than from accumulated details.
Independent corroboration: The frozen evidence defines First-Principles Model as 'Builds the initial model from fundamental relations, constraints, or causal claims rather than from accumulated details', so its operative form is Analysis, Modeling & Optimization.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Physics
Origin pattern: Convergent development
Present-day reach: Universal
Rationale: Building quantitative models from fundamental laws is characteristic of theoretical and computational physics.
Related originating lineages:
- Engineering & Design — Engineering independently uses physics-based models rather than empirical detail accumulation.
- Philosophy — Reasoning from first principles has an older philosophical lineage in foundational argument.
Review resolution: Both reviewers agree that physics is primary. I retain engineering_design, philosophy only as formative origin lineage(s), without treating every later application as an origin. convergent is appropriate because the same operational structure arose through materially independent professional lineages. Reach is universal as a separate applicability judgment: it does not widen or narrow the recorded provenance. Encyclopedia synthesis is false because the artifact is already established enough that encyclopedia-specific synthesis is not required. The secondary differences are reconciled with no unresolved primary-provenance ambiguity.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
First-principles derivation and a data-driven Baseline Model are complementary origins for the same core, not rivals: one reasons downward from invariants, the other fits the simplest defensible pattern upward from observations. A core that both routes agree on is unusually trustworthy; a disagreement between them is itself a refinement trigger worth chasing.
[n1] The "spherical cow" is physicists' name for a model idealized so far that its premises no longer resemble the system it describes — a standing reminder that a flawless derivation inherits every flaw in the fundamentals it started from. ↩