Lumen › Physics track › Chapter 6

Dark Sector & Cosmology  Empirical

Most of the universe is missing, in the precise sense that most of what gravitates and most of what drives the expansion is nothing we have ever caught in a detector. Theories of everything are obliged to say something about this darkness, and the easy move is to invent new substances to fill it. This theory does not add anything to the inventory. It says the two darks are not stuff at all but the two ways a single map fails — and that reading delivers the best number the whole framework produces, alongside its most useful lesson in self-correction.

The two darks as the two failures of one map

Return to the master identity, $C \circ P = I$. It is tempting to read it as saying the map is invertible, that you can always undo it, but that is exactly what it does not say. The composite is what mathematicians call a split idempotent: it is the identity on its own canonical part, but globally it loses information — it is not invertible, and it was never meant to be. Any such map has two distinct ways of being imperfect. It can send things to nothing — that collection is its kernel. And there can be things in the target it simply cannot reach — that is its cokernel. The theory’s claim is as economical as it is bold: the kernel is dark matter, the things that are there but do not couple to light, and the cokernel is dark energy, the part of the space the map cannot fill. Two darks, one map, two defects. And it explains their famous asymmetry — why dark matter clumps and dark energy is smooth — as the natural difference between a kernel, which carries dynamics, and a cokernel, which is just a measure-theoretic leftover.

The corner of a box

The leftover has a value, and it is the kind of thing that stops you when you first see it. Picture the four-ball sitting snugly inside the smallest box that contains it — a four-dimensional cube. The ball is round, the box has corners, and the ball cannot reach into the corners. Work out what fraction of the box the ball fills and you get $\pi^2/32$, about thirty-one percent; the corners it leaves untouched are the rest:

$\Lambda_0 = 1 - \dfrac{\pi^2}{32} = 0.6916.$

The measured fraction of the universe that is dark energy is $0.6847 \pm 0.0073$. The geometric corner volume and the cosmological measurement agree to within a single standard deviation, with absolutely nothing tuned. This is the theory’s strongest contact with reality — a pure number, the empty corners of a box, landing on the dark-energy density of the cosmos. And it is specific to four dimensions: the same construction gives the wrong answer in three or in six. Two independent ways of counting the structure — one through the faces of the folded box, one through the octonion units — arrive at the same single leftover. The verifier computes the corner volume and stands it next to the data.

The Hubble constant, and a theory correcting itself

Not every confrontation goes so well, and the most instructive one is where the theory caught itself in an error. It predicts a value for the Hubble constant, the present expansion rate, of about $75.8$ — and that sits above every measurement, the local ones near $73$, the cosmic ones near $67$. Taken at face value this looks like a serious failure, and an earlier version of the theory’s own scorecard recorded it as a three-sigma tension. But the scorecard had made a mistake, and the mistake is worth understanding because the fix is a model of honesty. The prediction is not a bare number; it is pinned to that one measured anchor, the galactic acceleration scale, and the relationship is $H_0 \propto \sqrt{a_0}$. The anchor itself is only known to about twenty-two percent. So the prediction inherits that uncertainty — halved by the square root, about eleven percent — which means it is not $75.8$ but $75.8 \pm 8$. The original scorecard had used that twenty-two-percent error to pass the anchor while forgetting to apply it to the Hubble constant that depends on the very same anchor. Apply it consistently and the “three-sigma tension” melts to about a third of a sigma from the local measurement. The number is not in conflict with the data; it is simply anchor-limited, and it will become a sharp test only when someone measures the acceleration scale more precisely. A theory eager to look impressive would have kept the dramatic tension. This one corrected itself downward, against its own interest, because the bookkeeping demanded it.

Gravity, and the galaxies that anchor everything

Gravity, in this picture, is the inward half of the breath — the surface lensing itself, the geometry pulling back toward the bulk. It rides the same density as electromagnetism, just weighted toward the interior, and the resulting formula for the enormous ratio between the Planck mass and the electron mass lands at $51.530$ against the observed $51.527$ — five thousandths of a percent. That number, though, is better than its footing: the spectral mechanism meant to derive it was later refuted, so the Gravity & Quantum Limits chapter files it as a matching formula, not a secured prediction. Run the same structure down to galactic scales and the dark-matter halos come out with a cored shape that produces flat rotation curves as an output rather than a fit, obeying the empirical Tully–Fisher law with a coefficient the theory derives. Tested against a hundred and seventy-five galaxies, it reproduces the universal acceleration scale and a single characteristic length of about a kiloparsec, measured three independent ways that agree to within ten percent with no per-galaxy freedom, while the obvious competing hypothesis is excluded at better than thirteen sigma. That kiloparsec is the light-crossing distance of one system tick — the same anchor the whole cosmology hangs on, which is precisely why the Hubble constant downstream of it inherits its uncertainty.

First-Edition sources & full derivations: P32 · P38 · P39 · P17

Verify this chapter

A standalone Python script (numpy only — no network, no corpus dependency) recomputes this chapter’s quantities from first principles and compares each to the measured value. The table below is its actual output. View the script · run all chapters.

7/7 checks passed. Method: exact arithmetic + standard error propagation (H0 inherits the a0 anchor uncertainty).

Measurement sources: Planck 2018 (Omega_L=0.6847+/-0.0073); SH0ES Riess+2022 (73.04+/-1.04); Planck (67.36+/-0.54); DESI 2024 (68.52+/-0.62); SPARC (a0=1.20+/-0.26e-10).

Independently pinned in the First Edition by: verify_P032.py, verify_P038.py, verify_P393.py, verify_P017.py — trace any number back to the full archive.

========================================================================================================
  LUMEN — Dark Sector & Cosmology
  the cokernel void Lambda_0, the anchor-limited H0, and the gravity ratio
  method: exact arithmetic + standard error propagation (H0 inherits the a0 anchor uncertainty)
========================================================================================================
  Quantity                      Lumen           Measured / target   Source          Residual          
  ----------------------------------------------------------------------------------------------------
  filled fraction vol(B^4)/vol( 0.308425        (geometric)         geometric       geometric   PASS  
  Lambda_0 = 1 - pi^2/32        0.691575        0.6847              Planck 2018     +0.94 sigma PASS  
  H0 prediction (anchor-limited 75.8 km/s/Mpc   (geometric)         2 alpha^2/T_bre geometric   PASS  
  H0 vs SH0ES                   75.8 km/s/Mpc   73.04 km/s/Mpc      SH0ES           +0.33 sigma PASS  
  H0 vs DESI                    75.8 km/s/Mpc   68.52 km/s/Mpc      DESI            +0.88 sigma PASS  
  H0 vs Planck                  75.8 km/s/Mpc   67.36 km/s/Mpc      Planck          +1.03 sigma PASS  
  log(M_Pl/m_e) = (3pi/20) mu1/ 51.5299         51.5278             CODATA masses   +39.03 ppm  PASS  
  ----------------------------------------------------------------------------------------------------
  7/7 checks passed.
  Measurement sources: Planck 2018 (Omega_L=0.6847+/-0.0073); SH0ES Riess+2022 (73.04+/-1.04); Planck (67.36+/-0.54); DESI 2024 (68.52+/-0.62); SPARC (a0=1.20+/-0.26e-10)
  Independently pinned by First-Edition verifiers: verify_P032.py, verify_P038.py, verify_P393.py, verify_P017.py
========================================================================================================