The Laws of Black-Hole Mechanics

The laws of black-hole mechanics organize four different kinds of classical statement: constancy on a stationary horizon, a variational identity between nearby solutions, a monotonicity theorem for evolving event horizons, and a historical obstruction to attaining extremality whose unrestricted form is false but whose connected-horizon restriction survives. Their resemblance to the laws of thermodynamics is mathematically fruitful, but it is not itself a proof of temperature or statistical entropy. This chapter develops the geometric and field-equation content first, states every global and energy hypothesis, and then tests which conclusions survive beyond vacuum Einstein gravity.

Helpful background. Exact Solutions of Einstein’s Equations supplies Schwarzschild and Kerr–Newman checks; Causality and Global Structure supplies event horizons, generator completeness, and predictability; The Einstein Equation connects curvature to matter; and Flat-Background Hilbert Stress-Energy from Metric Variation supplies the classical flux and energy conditions.

Four laws with four logical forms

For a connected stationary Killing horizon, the zeroth law states that the surface gravity $\kappa$ is constant over the horizon. This is a theorem only after one chooses a route: a bifurcate-horizon argument uses smooth Killing geometry, while a field-equation argument uses Einstein’s equation plus an energy or matter hypothesis. Constancy never removes the normalization freedom of the horizon generator. Once the asymptotic time is normalized, Schwarzschild gives $\kappa=1/(4GM)$.

The equilibrium first law is a one-form on the space of admissible solutions. For asymptotically flat four-dimensional Einstein–Maxwell black holes,

$$\delta M =\frac{\kappa}{8\pi G}\delta A +\Omega_H\delta J+\Phi_H\delta Q.$$

The coefficients are evaluated on the unperturbed stationary solution. The variation satisfies the linearized equations and boundary conditions, the asymptotic generators and couplings are fixed, and gauge potentials are referenced consistently. Bardeen, Carter, and Hawking present the original four-law structure and stationary first law in Bardeen, Carter, and Hawking 1973, pp. 161–170, Open PDF.

The second law in classical Einstein gravity is Hawking’s area theorem. It is not a local conservation law: null focusing gives the contradiction mechanism, but global predictability or suitable completeness prevents a negatively expanding event-horizon generator from ending harmlessly. Under those hypotheses and null convergence, later cross-sections have no smaller area. Hawking’s original argument is given in Hawking 1972, §3, pp. 156–160.

The third law has the narrowest stable scope. Israel claimed that bounded stress-energy satisfying the weak energy condition near an apparent horizon prevents a continuous process from reaching $\kappa=0$ in finite advanced time Israel 1986, pp. 397–399. A spherical proposition survives when the trapped region is nonempty and the outermost apparent horizon is connected. The unrestricted claim does not: a peer-reviewed Einstein–Maxwell–charged-scalar construction satisfies the dominant energy condition yet changes from an early Schwarzschild apparent-horizon component to a late extremal Reissner–Nordström event horizon through a disconnected outermost apparent horizon Kehle and Unger 2025, Theorem 1 and §1.4.3, pp. 3 and 11–13, Open PDF. This is neither the algebraic Kerr–Newman extremality bound nor evidence that extremal formation is generic.

Equilibrium identity and physical process

Two derivations produce the same first-order formula but answer different questions. The Equilibrium First Law is a Hamiltonian identity on a stationary background. It separates a smooth-bifurcation theorem for arbitrary admissible linearized perturbations from a future-horizon calculation along stationary solution families; their Maxwell work terms appear at different boundaries but give the same complete law. The detailed charge construction and stationary Einstein–Maxwell calculation are developed in Conserved Charges, Smarr Relations, and First Laws. Iyer and Wald prove the smooth-field result in Iyer and Wald 1994, §VI, Theorem 6.1 and equations (92)–(102), Open PDF, pp. 17–18.

The Physical-Process First Law instead follows a weak influx through a nearly stationary horizon. On the background, expansion and shear vanish. Linearized Raychaudhuri reduces to

$$\frac{d(\delta\theta)}{d\lambda} =-8\pi G\,\delta T_{ab}k^ak^b.$$

Future settling supplies the boundary condition, and $\chi^a=\kappa\lambda k^a$ converts the integrated focusing into Killing- energy flux. The result is

$$\frac{\kappa}{8\pi G}\delta A =\delta M-\Omega_H\delta J-\Phi_H\delta Q.$$

This calculation excludes caustics, newly joining generators, and effects quadratic in the perturbation. In particular, shear squared contributes at second order. A violent merger is not a large version of this linear proof.

Where the hypotheses enter

The same short list of words does not govern every law:

Result	Local input	Global or boundary input	Primary failure mode
Zeroth law	Killing identity and field equations, or bifurcate symmetry	Connected horizon and fixed generator normalization	Dropping the matter or symmetry hypothesis; comparing differently normalized generators
Equilibrium first law	Linearized equations and symplectic identity	Stationary background, integrable charges, fixed boundary data and one consistent gauge/inner-section choice	Varying outside the solution phase space or mixing charge conventions or boundary expressions from different Maxwell representatives
Area theorem	Raychaudhuri and null convergence	Event horizon plus completeness or predictability and regular cross-sections	Energy-condition violation, generator endpoints, or using a foliation-dependent horizon
Physical-process first law	Linearized focusing and conserved flux	Initial near-stationarity, future settling, no caustics	Strong local influx, added generators, or higher-order backreaction
Surviving third-law proposition	Raychaudhuri, weak energy condition, and a nonempty trapped region in spherical symmetry	Connected outermost apparent horizon and Israel’s finite-time trapped-surface definition	A disconnected outermost horizon can change branches while regularity and the energy condition hold

This separation prevents a common circularity. The area theorem cannot be proved merely by citing the first law and assuming positive mass flux, because the equilibrium first law does not describe finite evolution. Conversely, the physical-process calculation does not establish the global area theorem, because it assumes a perturbatively controlled horizon and future settling.

Routes through the chapter

For the shortest stationary route, read Zeroth Law and Constant Surface Gravity and then Equilibrium First Law. Together they explain which horizon potentials are constant and why their coefficients appear in a Hamiltonian variation.

For the dynamical route, read Hawking Area Theorem before Physical-Process First Law. The first emphasizes global causal control; the second performs the near-equilibrium flux calculation. Comparing them is the clearest way to see why both use Raychaudhuri without proving the same theorem.

For extremal and modified-gravity questions, continue through Third Law and Unattainability of Extremality and Validity Beyond Einstein Vacuum Gravity. These pages separate conditional theorems, rigorous counterexamples, stationary Noether-charge identities, linearized monotonicity, and unresolved nonlinear regimes.

Classical mechanics is not yet thermodynamics

Every result in this chapter can be stated within classical field theory. The quantity $A/(4G)$ is useful because it makes the first law read $(\kappa/2\pi)\delta[A/(4G)]$, and higher-curvature theories replace it by a Noether-charge functional. That notation does not by itself establish a Hawking temperature, count microscopic states, include exterior matter entropy, or prove a generalized second law.

Those additional claims require quantum fields, state choices, renormalized stress-energy, and generalized entropy. They belong to semiclassical gravity. Keeping the layers distinct also clarifies failure: classical negative null energy can invalidate the area theorem, while a semiclassical generalized- entropy theorem, where available, concerns a different quantity under different assumptions.

Review the chapter

A researcher observes a small positive Killing-energy flux $E$ through a Schwarzschild horizon and argues: “The equilibrium first law proves that the area grows by $8\pi GE/\kappa$ during any process; therefore the full area theorem follows.” Identify the valid calculation and the invalid inference.

Solution

For a weak, quasi-stationary influx that creates no caustics and settles to a stationary black hole, the physical-process derivation gives $\delta A=8\pi GE/\kappa$. Since $\kappa=1/(4GM)$, this is $32\pi G^2ME$, matching the first-order expansion of $A(M+E)-A(M)$.

The equilibrium first law alone is a variational identity about a stationary background; it does not control arbitrary time evolution. The general area theorem instead requires null convergence, event-horizon achronality, regularity, and completeness or predictability to exclude negative expansion and generator endpoints. Strong processes may also contain caustics and higher-order shear terms, so the linear flux formula cannot be promoted to the global theorem.

Stationary Classification, Rigidity, and No-Hair Results next uses stationarity, horizon regularity, topology, and the field equations to ask which black-hole geometries can realize these mechanics data.

References

• Bardeen, J. M., B. Carter, and S. W. Hawking. “The Four Laws of Black Hole Mechanics.” Communications in Mathematical Physics 31 (1973): 161–170. DOI.
• Hawking, Stephen W. “Black Holes in General Relativity.” Communications in Mathematical Physics 25 (1972): 152–166. DOI.
• Israel, Werner. “Third Law of Black-Hole Dynamics: A Formulation and Proof.” Physical Review Letters 57 (1986): 397–399. DOI.
• Iyer, Vivek, and Robert M. Wald. “Some Properties of Noether Charge and a Proposal for Dynamical Black Hole Entropy.” Physical Review D 50 (1994): 846–864. DOI.
• Kehle, Christoph, and Ryan Unger. “Gravitational Collapse to Extremal Black Holes and the Third Law of Black Hole Thermodynamics.” Journal of the European Mathematical Society (2025), published online first. DOI. Open PDF, arXiv:2211.15742v2.