The Core Idea

One function. Three parameters. Every scale from Planck to cosmic.

Think of it as a dimmer switch from sparse/independent to dense/collective. Feed it the local density of a system, and it outputs a number between 0 (sparse/independent) and 1 (dense/collective). The same switch, the same scale, works for a single atom and a galaxy cluster.

C(ρ) = tanh(γ · ln(ρ/ρcrit + 1))

tanh, in one picture. tanh is just an S-shaped curve — flat, then steep, then flat again. Nothing happens at low density, a small change flips the behavior in the middle, and past the knee more density changes almost nothing. That's the whole “dimmer switch.”

01coherence Cdensity ρ (log scale →)ρcrit — C ≈ 0.88 here: a saturation knee, not a midpointC = 0.5 at ≈ 0.32·ρcritdim: sparse, independent (C ≈ 0)bright: dense, collective (C ≈ 1)the switch

Drawn with γ = 2 (the galaxy-scale pin — refuted on SPARC data, see the caveat below) on a logarithmic density axis. Note the two honest markers: the curve's true midpoint (C = 0.5) sits near 0.32·ρcrit, and at ρcrit itself C ≈ 0.88 — ρcrit is a saturation knee, not a critical point.

ρ
Presence: local density — how many relevant elements are packed in a given region (e.g. stars per cubic light-year, atoms per unit volume)
C
Coherence: 0 = sparse/independent, 1 = dense/collective. ⚠ Not quantum coherence — superconductors score low here.
γ
2/√Ncorr: transition sharpness⚠ This card used to read “coupling strength” (corrected 2026-07-27). That is the wrong kind of quantity: the Hill identity proved below makes γ a Hill coefficient — a log-log slope, with Hill index n = 2γ — not a coupling. Also note the map 2/√Ncorr is audited sign-inverted, and the galaxy pin γ = 2 inverts to Ncorr = 1 — the ideal-gas value.
ρcrit
A × Vflat²: saturation knee (not a critical point)This scaling is refuted, and had been sitting here uncaveated (added 2026-07-27). The BTFR forces ρcrit ∝ V−2, the opposite sign to the V+2 written above: at the knee GM/r² = a₀ gives r = √(GM/a₀), so ρ ~ M/r³ ∝ M−1/2 ∝ V−2 using M ∝ V⁴. See the locality no-go. The constant A is separately unanchored — its stated first-principles formula is off by ~600× from the value in use.

Why These Specific Choices?

Why tanh?

tanh is an S-shaped curve that smoothly transitions from 0 to 1 — think of it as a dimmer switch from sparse/independent to dense/collective. The function must be bounded [0, 1], monotonic, and smooth. tanh's form recalls the Ising model self-consistency equation m = tanh(βJzm), but the resemblance is only visual: C(ρ) has no feedback loop (ρ goes in, C comes out), so the Ising analogy carries no motivating force. Other sigmoids (logistic, error function, Hill) satisfy the same four constraints. tanh is a phenomenological choice, not a derived result. Note: C(ρ) is real-analytic everywhere on its domain — no singularity, no free-energy functional, and no length scale anywhere in the equation — so it has no critical point and therefore no critical exponents (β, ν) to compare. Asking for them is a category error, not a near-miss with Landau theory; the analogy is motivational only.

Need an analogy first? · Why tanh? (motivation, not derivation) →

Why γ = 2/√Ncorr?

The 1/√Ncorr dependence resembles central-limit-theorem scaling (fluctuations ~ 1/√N), which is generic statistics for correlated ensembles. Ncorr (number of correlated particle units) is the physically measurable quantity. The factor of 2 is motivated by phase-space arguments (6D contracted to 3 effective) but should be understood as a motivated ansatz rather than a rigorous derivation.

Why 2/√Ncorr? (motivation, not derivation) →

Why log? (Honest answer: it's cosmetic)

The traditional motivation was compression: density spans 80+ orders of magnitude (from interstellar gas at 10−24 g/cm³ to neutron stars at 1014 g/cm³), and the logarithm squeezes that range into something the tanh can work with. But an exact identity retires that story. With x = ρ/ρcrit:

tanh(γ·ln(1+x)) ≡ [(1+x) − 1] / [(1+x) + 1] — exactly.

The equation is an ordinary rational (Hill-type) saturation function of (1+x); the tanh∘ln decomposition is notation, not mechanism. Any saturating rational function “spans 80 orders of magnitude” for free — the ln does no work that the exponent 2γ doesn't undo. So the Hill functions and Naka–Rushton curves named below are not merely similar alternatives; the flagship equation is one of them, exactly. (Identity stated 2026-07-10. Provenance: derived independently by two successive internal review passes, 2026-07-09 and 2026-07-10 — the audit machinery caught it; the identity then took a day to reach this page. No external derivation is on record.)

What It Predicts

Audit note: 6 of 6 badges previously labeled “Validated” on this site have been demoted to Reparametrization on closer review. These regime cards have not yet been independently audited — treat all current badges as Reparametrization-pending-audit until shown otherwise. See honest assessment.

γ « 1: Collective Regime

Many correlated particles (large Ncorr → small γ = 2/√Ncorr). BECs, superconductors. Despite being quantum systems in the real-world sense, these score C ≈ 0 here — the S-curve is so flat that typical densities leave C near zero. This is why “coherence” in Synchronism is not quantum coherence.

Reparametrization — regime definition

γ ≈ 1: The Boundary

Phase transitions, chemistry, catalysis, biology. Where the S-curve has intermediate steepness. 1,703 phenomena cluster here at 89% boundary-consistency rate.

Pending null model: density-monotonic targets produce r > 0.95 from trivial polynomial fits. Chemistry correlations may reflect monotonic data structure, not framework specificity.

Reparametrization — null model pending

γ » 1: Independent Regime

Few correlated particles (Ncorr ≈ 1 → large γ). Ideal gas, independent stars in a galaxy. Steepest S-curve — C rises quickly with density. Galaxy dynamics is here (stars treated as independent, Ncorr=1, γ=2).

Reparametrization — regime definition
What C(ρ) actually is: A logarithmic compander — not merely in the family of μ-law encoders, Hill functions, and Naka–Rushton curves, but algebraically identical to a Hill-type rational saturation function: tanh(γ·ln(1+x)) ≡ [(1+x)−1]/[(1+x)+1] exactly (see “Why log?” above). The “Ising motivation” is conceptual: C(ρ) is not a self-consistency equation (ρ goes in, C comes out, no feedback). C(ρ) is analytic on its whole domain, with no singularity and no length scale in the equation — so it has no critical point, and critical exponents (β, ν) are undefined by construction, not “off by 2×.” (The apparent “2×” is βeff = 1 from the small-x expansion vs. mean-field ½ — but βeff = 1 is a tautology for any analytic function vanishing at the origin, not a measurement.) That is the diagnostic result that rules out C(ρ) as a Landau-theory continuum order parameter — a category error, not a near miss. The cost of this dismissal, stated where it is made (added 2026-07-23): abandoning criticality means abandoning universality classes entirely — so every “phase transition” phrasing elsewhere on this site is decorative vocabulary, not physics. The compander reframing absorbs this honestly, but the trade should be visible: no critical point ⇒ no universality ⇒ no Landau machinery to borrow authority from. See honest assessment.

The data say the same thing, independently (added 2026-08-08). Written as a Hill function the exponent is the Hill index n = 2γ. Fit freely to SPARC, the galaxy sector prefers γ ≈ 0.489, i.e. n ≈ 0.98 ≈ 1 — and n = 1 is the non-cooperative limit of a Hill function, the value that means no collective effect. The framework's whole distinctive claim is that density drives a collective transition. Its own best fit sets cooperativity to zero. That is the criticality retraction above arriving a second time, through the data rather than through the algebra. At exactly γ = 1/2 the identity closes: C(x) = x/(x+2) = μsimple(x/2), MOND's simple interpolating function — so “the fit converges to MOND” is not an empirical coincidence but algebra (see the RAR entry and For Researchers).

Choose Your Path

Related Concepts

FundamentalsThe four foundations — everything else flows from theseThe Coherence FunctionC(ρ) = tanh(γ ln(ρ/ρ_crit + 1))The γ Parameterγ = 2/√N_corr: why 2, why √NScale InvarianceFrom Planck to cosmic: 80 orders of magnitude