This episode features Professor Philip Mannheim presenting conformal gravity, a fourth-order derivative theory based on local conformal symmetry (Weyl tensor squared) that simultaneously addresses dark matter, dark energy, quantum gravity renormalizability, and the cosmological constant problem without introducing dark matter, dark energy, or string theory.
Einstein’s Relativity and Its Loopholes
Einstein’s theory combines two distinct achievements: general coordinate invariance (the geodesic equation, equivalence principle, metric as gravitational field, Riemann tensor as curvature) and the Einstein field equations (generalizing Poisson’s equation to curved spacetime).
The first set is generic and logically secure; the second is phenomenological — Einstein chose the second-order Poisson equation because it reproduced Newton’s 1/r potential in the solar system, not because it was uniquely derived.
A fourth-order Poisson equation (∇⁴φ = ρ) yields 1/r + r solutions; the extra r term is negligible in the solar system but becomes significant at galactic scales, where dark matter is invoked.
Eddington noted in 1920 that the Einstein-Hilbert action (Ricci scalar) is not unique — varying Ricci scalar squared gives fourth-order equations with extra terms. No fundamental principle forces the second-order choice.
Quantum Gravity and Renormalizability
Standard quantum gravity seeks a quantum theory whose low-energy limit is Einstein gravity, but Einstein gravity has a dimensionful coupling constant (Newton’s constant) and is non-renormalizable.
Renormalizable theories (QED, electroweak, QCD) have dimensionless couplings and conformal symmetry at the classical level; mass enters dynamically via spontaneous symmetry breaking in the vacuum, not in the Lagrangian.
Conformal gravity uses the Weyl tensor squared action (C²), giving a dimensionless coupling and fourth-order equations — renormalizable by power counting.
The theory was long rejected because fourth-order theories appear to have ghost states (negative norm) and violate unitarity.
Conformal Symmetry, Dynamical Mass, and the Higgs
Conformal symmetry (15 generators: Lorentz + translations + dilatation + special conformal transformations) is the full symmetry of the light cone; the Dirac spinor is its fundamental representation, requiring four-component fermions and predicting right-handed neutrinos.
Mass generation occurs dynamically at a renormalization group fixed point: when the fermion bilinear ⟨ψ̄ψ⟩ acquires anomalous dimension 2 (instead of 3), infrared divergences force spontaneous symmetry breaking, generating mass without fundamental scalars.
This solves the hierarchy problem — no elementary scalar means no quadratic divergences — and predicts the Higgs is a composite bound state, not an elementary field. The “God particle” becomes the “God vacuum.”
Solving Dark Matter: Fourth-Order Gravity and Cosmological Potentials
Mannheim and Kazanas (1980s) solved the static, spherically symmetric vacuum solution: V(r) = -β/r + γr/2. The 1/r term matches Newton/Einstein; the linear γr term is new.
For galaxies: the falling 1/r (Kepler) plus rising γr yields flat rotation curves. The rising term is not flat — it rises, and the sum with the falling term produces the observed flatness.
Two linear potentials emerge: one local (from the galaxy’s own stars, γlocal) and one global (from the rest of the universe, γglobal = √(-K), where K < 0 is the cosmological curvature). They compete, with opposite signs.
Fits to 138 galaxies (with O’Brien) used only luminous matter plus two universal linear potentials (γglobal, γlocal) and one cluster-scale parameter — 3 universal parameters total. ΛCDM requires two free parameters per galaxy (276 for 138 galaxies).
The data show a universal acceleration scale ~10⁻³⁰.⁵ cm⁻¹ ≈ cH₀, matching MOND’s a₀. This scale is written in the data, not imposed by the theory.
Cosmic Coincidence, Accelerating Universe, and No Big Bang
Conformal cosmology with negative curvature (K < 0) gives a Friedmann equation: ȧ² + K = -ρ (sign flip from standard). No singularity at a=0 — the universe is cyclic/eternal, no Big Bang, no horizon or flatness problem.
The cosmological constant is controlled by conformal symmetry: the trace of the energy-momentum tensor vanishes, so vacuum energy induced by electroweak symmetry breaking cannot exceed other energy densities — naturally explaining ω_Λ ≈ 0.7, ω_m ≈ 0.3 (cosmic coincidence).
Predicted deceleration parameter q₀ = -0.37 (between 0 and -1), fitting supernova data without fine-tuning. Universe accelerates at all epochs, not just recently.
Inflation is unnecessary: horizon problem solved by a(t) ~ tⁿ with n < 1 (finite ∫dt/a); flatness problem solved by the sign-flipped Friedmann equation.
PT Symmetry, Ghost Resolution, and Unitarity
Fourth-order theories have propagators ~ 1/k⁴ = 1/k² - 1/(k²+m²). The minus sign suggests negative-norm ghosts if one assumes the Dirac inner product (Hermitian Hamiltonian).
Bender & Mannheim showed the Hamiltonian is not Hermitian but PT-symmetric (parity × time reversal). PT symmetry is an antilinear symmetry; it guarantees real eigenvalues and unitary time evolution without Hermiticity.
The “ghost” arose from using the wrong Hilbert space: the physical states are not normalizable on the real axis but become normalizable on a complex contour (e.g., imaginary axis). The dual space is the PT-conjugate, not the Hermitian conjugate.
Probability conservation, not Hermiticity, is the fundamental requirement. CPT theorem is rederived from complex Lorentz invariance + probability conservation.
Decays/resonances require complex-conjugate energy pairs (E ± iΓ). The growing mode (e⁺Γᵗ) and decaying mode (e⁻Γᵗ) combine with a relative minus sign to give a single Breit-Wigner peak while preserving unitarity — the time delay and time advance cancel.
Gravitons, Wavefunction Collapse, and Fermions
Quantizing conformal gravity: Weinberg’s theorem forbids a massive spin-2 particle coupling non-Einstein, but the escape clause is positive-definite Hilbert space metric. Conformal gravity gives zero-norm gravitons — gravitational radiation exists classically, but no graviton particle emerges.
Speculation: wavefunction collapse may be caused by emission of unobservable zero-norm gravitons.
Right-handed neutrinos are required by conformal symmetry (Dirac spinor = fundamental rep of conformal group). They get Majorana masses, enabling seesaw mechanism for light neutrino masses and spontaneous parity breaking (SU(2)_L × SU(2)_R × U(1) → SU(2)_L × U(1)).
Tensions with ΛCDM and Remaining Challenges
Hubble tension (Planck 68 vs. SH0ES 73 km/s/Mpc) and DESI hints of redshift-dependent dark energy challenge ΛCDM. Redshift-dependent dark energy worsens the cosmological constant problem (why not 10¹²⁰?).
Galaxy finite size: quadratic potential from cosmological fluctuations eventually dominates and turns negative (v² < 0), naturally cutting off galaxies — unlike 1/r which never cuts off.
Gravitational lensing: standard formulas assume asymptotically flat geometry. Conformal gravity’s linear potential means non-asymptotic geometry; lensing by clusters requires full trajectory calculation, not a difference experiment like solar bending. This is an active area of work.
CMB fluctuations and large-scale structure: the theory must reproduce the acoustic peaks. Mannheim’s group (Phelps, Amarasinghe, Li, Norman) is working on this — the critical remaining test.
Personal Reflections and Philosophy
Tenure enabled pursuit of a non-mainstream theory despite persistent resistance (ghosts, non-Hermiticity, challenging Einstein).
“Nature will keep you honest” — data, not mathematical elegance, is the ultimate arbiter. The universal acceleration scale in galaxy data is a clue ΛCDM must explain.
Advice to students: don’t trust experts (Wigner: “an expert is someone who’s made the most mistakes in the field”); trust your judgment; keep an eye on data; don’t work on conformal gravity until you have tenure.
