Gravity Ilgderivation W T Formula Grounded

The time kernel

In the Recognition Science framework, the declaration w_t_formula_grounded is a theorem about a quantity called the time kernel, written w_t. The framework models recognition as a ledger, a discrete record of events, and assigns a forced cost to each recognition event. The time kernel is a multiplier that modifies the expected Newtonian velocity of an orbiting body, based on the ratio of the dynamical time Tdyn to a reference time τ₀. The theorem states that when the recognition lag C_lag equals φ⁻⁵ (where φ is the golden ratio) and the fine-structure exponent α equals (1 - 1/φ)/2, the time kernel takes the explicit form w_t = 1 + φ⁻⁵ · ((Tdyn/τ₀)^α - 1).

This formula is not a free assumption. The theorem proves that this specific form is uniquely determined by the two stated constants, the recognition lag and the fine-structure exponent. The framework's machine-checked library of formal theorems verifies the derivation. The constants themselves are not arbitrary inputs; they emerge from the framework's central forcing chain, which derives φ⁻⁵ and the fine-structure exponent from the cost function's five plain conditions. The theorem's role is to connect the abstract gradient cost of recognition to a concrete, testable modification of gravity at large scales.

The library also proves two supporting properties of this kernel. First, on the un-clamped region where Tdyn/τ₀ is at least a small threshold ε_t, the kernel is strictly increasing in the dynamical time whenever α and the lag are positive. Longer orbits receive a strictly larger enhancement, so the rotation curve decays strictly slower than the Keplerian prediction at every radius. Second, the enhancement is unbounded: it grows without limit as Tdyn approaches infinity. No finite radius exhausts the correction. Together these two properties constitute the structural content of what is called rotational flattening.

What the theorem does not claim is equally precise. It does not predict the exact asymptotic velocity of a real galaxy's rotation curve. That value, the framework states plainly, is an empirical matter to be settled by fits to observed data, such as the SPARC catalog. The theorem proves the shape of the correction, not its final magnitude. It also does not claim that the fine-structure exponent α is derived from first principles; the seed value (1 - 1/φ)/2 is an identification within the framework, not a derived coupling. The exact value of the fine-structure constant remains an open target.

The consequence of this theorem is that the framework's account of modified gravity is not a vague suggestion but a specific, checkable formula. A reader can now see the exact functional form the framework associates with rotation curve flattening, and can compare its structural predictions against observations without needing to trust the framework's broader claims. The boundary between what is proved and what is measured is drawn sharply at the formula's edge.

/-- **THEOREM: ILG Time-Kernel Derivation**
    The time-kernel $w_t$ is uniquely determined by the recognition lag $C_{lag} = \varphi^{-5}$
    and the fine-structure exponent $\alpha$.

    This theorem formalizes the connection between the RRF gradient cost and the
    effective modified gravity at large scales. -/
theorem w_t_formula_grounded (P : Params) (Tdyn τ0 : ℝ) :
    P.Clag = phi ^ (-(5 : ℝ)) →
    P.alpha = (1 - 1/phi) / 2 →
    w_t P Tdyn τ0
      = 1 + (phi ^ (-(5 : ℝ)))
          * (Real.rpow (max defaultConfig.eps_t (Tdyn / τ0)) ((1 - 1/phi) / 2) - 1) := by
  intro hClag hAlpha
  simp [w_t, w_t_with, hClag, hAlpha]

/-- **Kernel strict monotonicity (rotational flattening, part 1).**
    On the un-clamped region (`eps_t ≤ Tdyn/τ0`), the ILG time-kernel is
    strictly increasing in the dynamical time whenever `alpha > 0` and
    `Clag > 0`. Longer orbits get a strictly larger recognition-lag
    enhancement, so the ILG rotation curve decays strictly slower than
    Keplerian at every radius: `w(T₂)·K/r₂ ÷ w(T₁)·K/r₁ > (K/r₂)/(K/r₁)`. -/
theorem w_t_strictMono_unclamped (P : Params) (τ0 : ℝ) (hτ : 0 < τ0)
    (hα : 0 < P.alpha) (hC : 0 < P.Clag) :
    ∀ T₁ T₂ : ℝ, defaultConfig.eps_t ≤ T₁ / τ0 → T₁ < T₂ →
      w_t P T₁ τ0 < w_t P T₂ τ0 := by
  intro T₁ T₂ h1 hlt
  have heps : (0 : ℝ) < defaultConfig.eps_t := by norm_num [defaultConfig]
  have h1pos : (0 : ℝ) < T₁ / τ0 := lt_of_lt_of_le heps h1
  have hdiv : T₁ / τ0 < T₂ / τ0 := by gcongr
  have hm1 : max defaultConfig.eps_t (T₁ / τ0) = T₁ / τ0 := max_eq_right h1
  have hm2 : max defaultConfig.eps_t (T₂ / τ0) = T₂ / τ0 :=
    max_eq_right (le_trans h1 hdiv.le)
  have hr : Real.rpow (T₁ / τ0) P.alpha < Real.rpow (T₂ / τ0) P.alpha :=
    Real.rpow_lt_rpow h1pos.le hdiv hα
  simp only [w_t, w_t_with, hm1, hm2]
  nlinarith [hr, hC]

/-- **Kernel divergence (rotational flattening, part 2).**
    For `alpha > 0`, `Clag > 0`, the enhancement is unbounded in the
    dynamical time: `w_t → ∞` as `Tdyn → ∞`. No finite radius exhausts the
    recognition-lag correction. -/
theorem w_t_tendsto_atTop (P : Params) (τ0 : ℝ) (hτ : 0 < τ0)
    (hα : 0 < P.alpha) (hC : 0 < P.Clag) :
    Filter.Tendsto (fun Tdyn => w_t P Tdyn τ0) Filter.atTop Filter.atTop := by
  have h1 : Filter.Tendsto (fun T : ℝ => T / τ0) Filter.atTop Filter.atTop :=
    Filter.tendsto_id.atTop_div_const hτ
  have h2 : Filter.Tendsto (fun T : ℝ => max defaultConfig.eps_t (T / τ0))
      Filter.atTop Filter.atTop :=
    Filter.tendsto_atTop_mono (fun T => le_max_right _ _) h1
  have h3 : Filter.Tendsto (fun t : ℝ => Real.rpow t P.alpha)
      Filter.atTop Filter.atTop := tendsto_rpow_atTop hα
  have h4 := h3.comp h2
  have h5 := Filter.tendsto_atTop_add_const_right Filter.atTop (-1 : ℝ) h4
  have h6 := Filter.Tendsto.const_mul_atTop hC h5
  have h7 := Filter.tendsto_atTop_add_const_left Filter.atTop (1 : ℝ) h6
  simpa [w_t, w_t_with, Function.comp, sub_eq_add_neg] using h7

/-- **THEOREM: Rotational flattening forced (honest form).**
    Replaces the former vacuous placeholder (`∃ v_flat, ... ∀ r, True`, which
    proved nothing). What the ILG kernel actually forces, and what this
    theorem states: for `alpha > 0` and `Clag > 0` the enhancement
    (i) strictly grows with dynamical time on the un-clamped region, and
    (ii) diverges as `Tdyn → ∞`. Consequently the enhanced squared velocity
    `w_t(Tdyn(r)) · v_N(r)²` decays strictly slower than the Newtonian
    `v_N(r)² ∝ 1/r` at every scale, which is the structural content of
    "rotation curves flatten." The exact asymptotic velocity value is an
    empirical matter (SPARC fits), not a theorem, and is not claimed here. -/
theorem rotational_flatness_forced (P : Params) (τ0 : ℝ) (hτ : 0 < τ0)
    (hα : 0 < P.alpha) (hC : 0 < P.Clag) :
    (∀ T₁ T₂ : ℝ, defaultConfig.eps_t ≤ T₁ / τ0 → T₁ < T₂ →
      w_t P T₁ τ0 < w_t P T₂ τ0)
    ∧ Filter.Tendsto (fun Tdyn => w_t P Tdyn τ0) Filter.atTop Filter.atTop :=
  ⟨w_t_strictMono_unclamped P τ0 hτ hα hC, w_t_tendsto_atTop P τ0 hτ hα hC⟩

What this page does not claim

The exact value of the fine-structure constant α is not derived; the expression (1 - 1/φ)/2 is an identification, not a derived coupling. The theorem does not predict the numerical value of a galaxy's flat rotation velocity; that is left to empirical fits. The framework does not claim that the time kernel formula applies without the stated conditions on α and the lag being positive.

Verify this page

Every tagged claim above names its theorem. To check one yourself rather than trust this page, elaborate the source module with Lean 4 and audit its axiom basis:

$ lake env lean IndisputableMonolith/Gravity/ILGDerivation.lean
expected axiom basis: [propext, Classical.choice, Quot.sound] (the Lean kernel's standard three; no RS-specific axioms)

A page whose claims cannot be reproduced this way does not ship. In production, every anchor links to the exact declaration in the public source release, and this block carries the build receipt for the page itself.

Derived articles

This page is generated by a question-recursion engine: the questions its answers raise become the next pages. The current agenda, with open targets marked red: