No-Hair Theorem and Black Hole Uniqueness

The no-hair theorem answers a simple question with one of the deepest results in classical gravity: when a star collapses into a stationary black hole, how much of the star survives in the geometry outside? A star carries an astronomical number of independent parameters — density profile, composition, magnetic structure, an arbitrary set of lumps. The theorem says the horizon keeps almost none of it. Everything gravitational and electromagnetic waves can carry is radiated to infinity or swallowed, and what remains is exactly the long-range charges protected by Gauss laws: mass $M$ , angular momentum $J$ , and electric charge $Q$ . John Wheeler's slogan "black holes have no hair" names this collapse of infinite detail onto a three-parameter family. This is not one theorem but a chain, assembled over two decades. Hawking's rigidity theorem shows a stationary black hole is either static or axisymmetric. Israel's theorem forces the static case to be Schwarzschild in vacuum, Reissner–Nordström with charge. The Carter–Robinson theorem reduces the rotating case to a boundary-value problem whose only regular solution is Kerr, and adding a Maxwell field gives Kerr–Newman; Mazur and Bunting closed the argument with a positivity identity built on the hidden sigma-model structure of the field equations. The endpoint is clean: a regular, stationary, asymptotically flat, four-dimensional electrovacuum black hole is the Kerr–Newman solution, fixed by $(M,J,Q)$ alone. Every higher multipole — quadrupole, octupole, the entire infinite tower — is a determined function of those three numbers, not an independent degree of freedom. This article develops the machinery, verifies the key relations, and shows how ringdown spectroscopy and horizon imaging now test the statement directly.

Key takeaways

The no-hair theorem is a family of uniqueness results: a stationary, asymptotically flat, four-dimensional electrovacuum black hole with a regular non-degenerate horizon is completely characterized by three externally observable numbers — mass $M$ , angular momentum $J$ , and electric charge $Q$ .
The endpoint is the Kerr–Newman family; it degenerates to Kerr ( $Q=0$ ), Reissner–Nordström ( $J=0$ ), and Schwarzschild ( $J=Q=0$ ) as the parameters are switched off.
The proof chain is Hawking rigidity (stationary $\Rightarrow$ static or axisymmetric) $\to$ Israel (static $\Rightarrow$ Schwarzschild/RN) $\to$ Carter–Robinson + Mazur–Bunting (axisymmetric $\Rightarrow$ Kerr/Kerr–Newman).
"Hair" means any independent field configuration or charge beyond $(M,J,Q)$ ; the Kerr multipoles obey $M_\ell + \mathrm{i}S_\ell = M(\mathrm{i}a)^\ell$ , so the quadrupole $M_2=-J^2/M$ is fixed, not free.
The theorem has explicit hypotheses — four dimensions, stationarity, asymptotic flatness, electrovacuum matter, a non-degenerate horizon — and relaxing any one admits hairy or non-unique black holes (colored, Skyrmion, AdS-scalar, higher-dimensional black rings).
It is tested observationally by black-hole spectroscopy (Kerr quasinormal modes depend only on $M,J$ ) and by Event Horizon Telescope shadow geometry.
Key idea: a stationary, asymptotically flat, four-dimensional electrovacuum black hole with a regular non-degenerate horizon is uniquely the Kerr–Newman solution, fixed entirely by $(M,J,Q)$ — all other multipole moments are determined functions of these three, and no additional independent field hair survives.

Where this lives today

The no-hair theorem stopped being a purely mathematical statement the moment gravitational-wave detectors began recording the ringdown of merger remnants. Black-hole spectroscopy is now a live program: the remnant settles to Kerr and rings in quasinormal modes whose complex frequencies depend only on $(M,J)$ , so checking that one $(M,J)$ describes several fitted modes is a direct test. LIGO–Virgo–KAGRA analyses of GW150914 and the growing catalog have tested the dominant $(2,2,0)$ mode against overtones and, tentatively, the subdominant $(3,3,0)$ mode; consistency so far supports the Kerr hypothesis, and the planned space detector LISA aims to measure multiple modes to percent-level precision. On the imaging side, the Event Horizon Telescope's pictures of M87 and Sgr A probe the photon ring and shadow, whose size and shape are fixed by Kerr through $(M,J)$ and inclination; an anomalous deformation would betray extra multipole hair, and current images are Kerr-consistent within uncertainties. In parallel, theorists build "bumpy" and hairy metrics — parametrized deviations from Kerr, dilatonic and Gauss–Bonnet holes, boson stars — so data can bound how much hair nature permits. The open frontier is the near-extremal regime, higher-dimensional and AdS settings relevant to holography, and whether Hawking evaporation restores the information the classical theorem appears to erase.

The physical picture

Collapse is violent, and violence is exactly what erases detail. A stationary exterior — one that no longer changes in time — cannot store most of a star's parameters. The only way for the geometry to settle down is to shed every deformation a wave can carry: any bump in the density, any tilt of the magnetic field, any excess in the mass quadrupole becomes a burst of gravitational and electromagnetic radiation that escapes to infinity or falls through the horizon. What cannot be radiated is precisely what a conservation law protects at infinity: total energy (mass), total angular momentum, and total electric charge. Everything else has, in Wheeler's phrase, no place to live on a stationary horizon.

Figure

Gravitational collapse radiates away every deformation a wave can carry — density profile, shape, magnetic structure, higher multipoles — until the exterior geometry is fixed by only three Gauss-law charges: mass M, angular momentum J, and electric charge Q.

The cleanest way to see that "everything else" is truly gone is the multipole moments. A generic spinning star has an arbitrary mass quadrupole, squashable independently of its mass and spin. A Kerr black hole cannot: horizon regularity over-determines the higher moments, locking the quadrupole to $M_2=-J^2/M$ and the whole tower to the single complex rule $M_\ell+\mathrm{i}S_\ell=M(\mathrm{i}a)^\ell$ . Any attempted quadrupole bump is non-stationary and rings down. The second figure shows the flip side — the entire solution family is one two-dimensional region, a point per allowed $(a/M,\,Q/M)$ , bounded by the extremal curve where the horizon degenerates.

Figure

The whole family of regular stationary electrovacuum black holes is one quarter-disk: axes a/M and Q/M, bounded by the extremal circle where (a/M)² + (Q/M)² = 1 and the surface gravity κ vanishes. Inside, every point is a Kerr–Newman hole with horizon r₊/M = 1 + √(1 − (a/M)² − (Q/M)²); outside there is no horizon, only a naked singularity.

The deep engine behind uniqueness is that the stationary axisymmetric Einstein–Maxwell equations are secretly a nonlinear sigma model. Packaged into the Ernst potentials, the metric and Maxwell data become harmonic maps from a two-dimensional half-plane into a symmetric target space of non-positive curvature — and such maps are rigid: two with the same boundary values must coincide. Two candidate black holes with the same horizon, axis, and charges are therefore the same spacetime. That single geometric fact, made quantitative by the Mazur identity, turns "collapse erases detail" from a slogan into a theorem — and it is why the principle is matter-specific: linear massless scalars obey their own no-hair lemma, but nonlinear gauge fields have no such rigidity and can carry genuine hair.

The core equations

The exterior of any regular stationary electrovacuum black hole is one metric. In Boyer–Lindquist coordinates and geometrized units $G=c=1$ , the Kerr–Newman line element packages mass, spin, and charge into two functions $\rho^2$ and $\Delta$ .

\mathrm{d}s^2 = -\frac{\Delta}{\rho^2}\big(\mathrm{d}t - a\sin^2\theta\,\mathrm{d}\phi\big)^2 + \frac{\sin^2\theta}{\rho^2}\big[(r^2+a^2)\mathrm{d}\phi - a\,\mathrm{d}t\big]^2 + \frac{\rho^2}{\Delta}\mathrm{d}r^2 + \rho^2\,\mathrm{d}\theta^2 \qquad (1)

where: $\rho^2 = r^2 + a^2\cos^2\theta$ ; $\Delta = r^2 - 2Mr + a^2 + Q^2$ ; $M$ — ADM mass; $a=J/M$ — specific angular momentum; $Q$ — electric charge. Reading: every symbol that can appear is one of $M$ , $a$ , $Q$ — no slot for a fourth parameter. Setting $Q=0$ gives Kerr, $a=0$ gives Reissner–Nordström, both zero gives Schwarzschild. The horizon is where $\Delta=0$ . Its two roots set the outer event horizon and the inner Cauchy horizon, and the condition that they be real is the extremality bound that draws the boundary of the uniqueness disk.

r_\pm = M \pm \sqrt{M^2 - a^2 - Q^2}\,,\qquad M^2 \ge a^2 + Q^2 \qquad (2)

where: $r_+$ — event horizon (outer root); $r_-$ — Cauchy horizon (inner root); the inequality $M^2\ge a^2+Q^2$ is required for a regular horizon rather than a naked singularity. Reading: saturation $M^2=a^2+Q^2$ is the extremal (degenerate) case with $r_+=r_-$ and zero surface gravity — the red boundary of the disk, which the standard uniqueness theorems handle separately. That all higher structure is fixed follows from the multipole moments. The Geroch–Hansen moments of Kerr collapse into a single complex relation — the sharpest quantitative statement of "no hair."

M_\ell + \mathrm{i}\,S_\ell = M\,(\mathrm{i}a)^\ell \qquad (3)

where: $M_\ell$ — mass multipole of order $\ell$ ; $S_\ell$ — current (angular-momentum) multipole; $a=J/M$ . Reading: $M_0=M$ , $S_1=Ma=J$ , and $M_2=-Ma^2=-J^2/M$ . The quadrupole is not a free hair; it and every higher moment are fixed polynomials in $(M,J)$ . Rigidity supplies the symmetry that makes the whole program tractable. Hawking's theorem shows the horizon of a stationary black hole is a Killing horizon: there is a Killing field that is null on it, built from the stationarity and axial fields.

\chi^\mu = \xi^\mu + \Omega_H\,\psi^\mu \qquad (4)

where: $\xi^\mu$ — asymptotically timelike (stationarity) Killing field; $\psi^\mu$ — axial (axisymmetry) Killing field; $\Omega_H$ — horizon angular velocity, a constant over the horizon; $\chi^\mu$ — null on and tangent to the horizon. Reading: the horizon rotates rigidly at one rate $\Omega_H$ , and $\Omega_H=0$ recovers the static case, splitting the proof into two branches. The reason axisymmetric uniqueness works is a hidden algebraic structure. The Ernst potentials repackage the metric and Maxwell fields into two complex scalars valued in a symmetric space.

\mathcal{E} = f - |\Phi|^2 + \mathrm{i}\,\chi,\qquad \Phi = A_t + \mathrm{i}\,\tilde A \qquad (5)

where: $f=-\xi^\mu\xi_\mu$ — norm of the timelike Killing field; $\chi$ — twist potential encoding rotation; $\Phi$ — complex electromagnetic Ernst potential; $A_t,\tilde A$ — electric and magnetic scalar potentials. Reading: quotienting by the two Killing fields turns the field equations into elliptic equations for $(\mathcal{E},\Phi)$ on a 2D half-plane, valued in the non-positively curved coset $SU(2,1)/S(U(1,1)\times U(1))$ — the curvature that makes the maps rigid. The horizon carries two constants that appear throughout black-hole thermodynamics and label the endpoint of ringdown.

\kappa = \frac{r_+ - r_-}{2(r_+^2 + a^2)},\qquad \Omega_H = \frac{a}{r_+^2 + a^2} \qquad (6)

where: $\kappa$ — surface gravity, constant over the horizon by the zeroth law; $\Omega_H$ — horizon angular velocity; $r_\pm$ — outer/inner radii. Reading: $\kappa\to0$ at extremality ( $r_+\to r_-$ ), the degenerate limit excluded from the standard theorems, and both quantities are again functions of $(M,J,Q)$ only. Finally, the matter-specific edge of the theorem is captured by the scalar no-hair identity, which shows a static scalar field cannot cling to the hole.

\Box\,\phi = 0 \;\Rightarrow\; \int_\Sigma (\nabla\phi)^2\,\sqrt{h}\,\mathrm{d}^3x = 0 \qquad (7)

where: $\phi$ — static (massless or massive) scalar field; $\Sigma$ — exterior static slice; $h$ — induced spatial metric. Reading: multiplying the wave equation by $\phi$ and integrating by parts kills the boundary terms at horizon and infinity, leaving a non-negative integral that can only vanish if $\nabla\phi=0$ , so $\phi$ is constant — no scalar hair. The same integrand's positivity is what fails for nonlinear Yang–Mills fields, which is why colored hair can exist.

Derivation AIsrael static vacuum uniqueness by level-set foliation

The goal is to show that a static, asymptotically flat vacuum black hole with a regular non-degenerate horizon must be spherically symmetric — hence Schwarzschild — thereby fixing the entire static branch of the uniqueness program with mass as the only label. The strategy is Israel's original one: foliate the exterior by the level surfaces of the Killing norm, write the vacuum equations as elliptic equations for that norm, and use integral identities together with the Gauss–Bonnet theorem to force the equipotential surfaces to be round spheres. We work on a static slice with metric $\mathrm{d}s^2=-V^2\mathrm{d}t^2+g_{ij}\mathrm{d}x^i\mathrm{d}x^j$ , where $V^2=-\xi^\mu\xi_\mu$ is the lapse.

Step 1. Reduce the vacuum equations on the static slice. Staticity kills all time derivatives, and the four-dimensional vacuum equations collapse to the pair $V R_{ij}=\nabla_i\nabla_j V$ and $\nabla^2 V=0$ , where $R_{ij}$ is the Ricci tensor of the three-metric $g_{ij}$ . Why: this converts a spacetime problem into an elliptic problem for a single scalar $V$ on a 3-manifold — the arena where positivity arguments live.

Step 2. Use $V$ itself as a radial coordinate. It increases monotonically from $V=0$ on the horizon to $V=1$ at infinity, so the surfaces $V=\text{const}$ foliate $\Sigma$ with unit normal $n_i=\nabla_i V/|\nabla V|$ . Why: adopting the potential as a coordinate turns the geometry into data on a one-parameter family of closed surfaces the field equations now control.

Step 3. Integrate the field-equation combinations over the leaves. They produce integrals of the extrinsic curvature and of $|\nabla V|$ with non-negative integrands, giving inequalities saturated only when each leaf is totally umbilic and $|\nabla V|$ is constant on it — the signature of spherical symmetry. Why: the positivity does the real work, the analytic version of "any deviation from roundness costs curvature the equations cannot supply."

Step 4. Close with Gauss–Bonnet. Each equipotential is a closed 2-surface, so $\oint_{V=\text{const}} K\,\mathrm{d}A = 8\pi$ , and the saturated inequalities force $K$ constant — a round sphere of area increasing outward. Why: topology plus saturation upgrades "umbilic" to "metric sphere," pinning the geometry.

Step 5. Identify the solution. A static vacuum geometry foliated by round spheres with $\nabla^2 V=0$ is Schwarzschild, lapse $V=\sqrt{1-2M/r}$ , with mass $M$ the only free constant. Why: the reduced problem has a unique regular solution — no hair beyond $M$ . Robinson, and later Bunting–Masood-ul-Alam with a conformal positive-mass argument, extended this to electrovacuum, giving Reissner–Nordström labelled by $(M,Q)$ . The reduction $V R_{ij}=\nabla_i\nabla_j V$ , $\nabla^2 V=0$ is the standard static vacuum system, and $V=\sqrt{1-2M/r}$ is the Schwarzschild lapse.

Go deeper · Derivation B · 1 minMultipole rigidity fixes the Kerr quadrupole

Derivation BMultipole rigidity fixes the Kerr quadrupole

The goal is to demonstrate quantitatively that a Kerr black hole cannot carry an independent quadrupole moment — that $M_2$ is the fixed function $-J^2/M$ of the first two moments — the sharp, testable form of "no hair for multipoles" and the basis of black-hole spectroscopy. The method is to read the Geroch–Hansen moments off the asymptotic expansion of the Kerr Ernst potential and track the first three orders, showing the quadrupole is pinned the moment $M$ and $J$ are chosen.

Step 1. Write the Kerr solution in Ernst form and pass to the potential $\xi=(1-\mathcal{E})/(1+\mathcal{E})$ , which admits an expansion about spatial infinity in suitable Weyl-type coordinates whose coefficients are, by the Geroch–Hansen prescription, the multipole moments $M_\ell+\mathrm{i}S_\ell$ . Why: the moments are defined covariantly as tensors at infinity, but for a stationary axisymmetric metric they reduce to expansion coefficients of a single complex potential, making them directly computable.

Step 2. Read off the two leading coefficients: $M_0=M$ and $S_1=Ma=J$ . Why: the monopole must be the ADM mass and the current dipole the total angular momentum — these are the conserved charges, and they fix the meaning of $M$ and $a$ before any higher moment appears.

Step 3. Read off the next nonvanishing coefficient: $M_2=-Ma^2$ . Substituting $a=J/M$ gives $M_2=-J^2/M$ . Why: no new integration constant enters at quadrupole order — the horizon-regular Kerr solution has already used up its freedom on $M$ and $a$ , so the quadrupole is forced.

Step 4. Confirm the pattern for the full tower: $M_{2k}=(-1)^k M a^{2k}$ and $S_{2k+1}=(-1)^k M a^{2k+1}$ , with all odd $M_\ell$ and even $S_\ell$ vanishing by reflection symmetry — the single relation $M_\ell+\mathrm{i}S_\ell=M(\mathrm{i}a)^\ell$ . Why: every moment is a fixed polynomial in $(M,J)$ , so a black hole has an infinite multipole tower but zero independent hair beyond the first two — a generic spinning star, by contrast, can set its quadrupole freely. *The Geroch–Hansen relation $M_\ell+\mathrm{i}S_\ell=M(\mathrm{i}a)^\ell$ for Kerr gives $M_2=-Ma^2=-J^2/M$ directly by substitution.*

Go deeper · Derivation C · 1 minThe Mazur identity and stationary axisymmetric uniqueness

Derivation CThe Mazur identity and stationary axisymmetric uniqueness

The goal is to show that two stationary axisymmetric electrovacuum black holes with identical boundary data — same horizon, same axis, same charges — are the same spacetime, completing the rotating branch and identifying it as Kerr–Newman. The engine is the sigma-model structure of the reduced Einstein–Maxwell equations: encode the Ernst potentials in a Hermitian matrix, form a positive scalar measuring the difference of two solutions, and integrate a divergence identity whose boundary terms vanish, forcing the interior difference to vanish too.

Step 1. Recast the reduced equations as a matrix sigma model $\nabla\cdot(\rho\,\Phi^{-1}\nabla\Phi)=0$ , where $\Phi$ is the Hermitian matrix built from the Ernst potentials, valued in $SU(2,1)/S(U(1,1)\times U(1))$ , and $\rho$ is the Weyl cylindrical radius. Why: this exposes the geometry — the field equations are harmonic-map equations into a non-positively curved target, the setting where uniqueness is a rigidity theorem.

Step 2. Given two candidate solutions $\Phi_{(1)}, \Phi_{(2)}$ , form the deviation $\sigma = \mathrm{tr}(\Phi_{(2)}\Phi_{(1)}^{-1}) - n$ , with $n$ the matrix dimension. Because the target is a symmetric space, $\sigma\ge0$ with equality iff $\Phi_{(1)}=\Phi_{(2)}$ . Why: $\sigma$ is a genuine non-negative "distance" between the two solutions, so proving $\sigma\equiv0$ proves they coincide.

Step 3. Establish the Mazur identity $\nabla^2\sigma = \rho^{-1}\,\|\,\Phi_{(1)}^{-1}(J_{(2)}-J_{(1)})\Phi_{(2)}^{1/2}\,\|^2 \ge 0$ , with $J_{(i)}=\Phi_{(i)}^{-1}\nabla\Phi_{(i)}$ the matrix currents. The right side is a manifest Frobenius norm, so $\sigma$ is subharmonic. Why: a Laplacian with a definite sign forbids interior maxima and ties the interior of $\sigma$ to its boundary flux.

Step 4. Evaluate the boundary integral $\oint_\partial \rho\,\partial_\mu\sigma\,\mathrm{d}S^\mu$ over horizon, axis, and infinity. Once $(M,J,Q)$ and regularity are fixed, the two solutions share the same horizon rod length, charges, and a regular axis, so each contribution vanishes. Why: identical boundary data is the hypothesis of the theorem, and it makes the only source of $\sigma$ vanish.

Step 5. Combine: a non-negative volume integral of $\nabla^2\sigma$ equals a vanishing boundary integral, so $\nabla^2\sigma=0$ and the norm on the right must vanish everywhere, giving $\Phi_{(1)}=\Phi_{(2)}$ . Why: the currents agree, so the potentials agree, so the metrics agree — the unique solution with the given data is Kerr–Newman.

Limiting cases

Switching the hairs off one at a time walks the Kerr–Newman family down to its famous special cases; pushing the parameters to the boundary of the disk reaches the extremal limit the standard theorems exclude.

Case	Result	Meaning
No spin, no charge	$J=0,\;Q=0 \Rightarrow$ Schwarzschild	Israel static vacuum uniqueness: the only static asymptotically flat vacuum black hole, labelled by $M$ alone.
Charged, no spin	$J=0,\;Q\neq0 \Rightarrow$ Reissner–Nordström	Israel / Bunting–Masood-ul-Alam static electrovacuum uniqueness, parameters $(M,Q)$ .
Spinning, uncharged	$Q=0,\;J\neq0 \Rightarrow$ Kerr	Carter–Robinson stationary axisymmetric vacuum uniqueness, parameters $(M,J)$ with $\lvert J\rvert\le M^2$ .
Extremal boundary	$M^2 = a^2 + Q^2 \Rightarrow \kappa=0,\;r_+=r_-$	Degenerate horizon on the red circle; zero surface gravity, treated by separate extremal uniqueness results.
Beyond the disk	$M^2 < a^2+Q^2 \Rightarrow$ no horizon	Naked singularity — excluded by cosmic censorship; not a black hole, outside the theorem.

Common trapNo-hair is a universal law of nature that forbids all black-hole hair.

It is not a law of nature; it is a theorem with explicit hypotheses, and every one can be broken. The rigorous statement assumes four spacetime dimensions, stationarity, asymptotic flatness, electrovacuum matter (Einstein–Maxwell only), a regular non-degenerate horizon, and standard energy conditions. Relax any one and hair appears: five-dimensional black rings are non-unique for given $(M,J)$ ; anti–de Sitter scalar condensates give hairy black holes; Einstein–Yang–Mills "colored" holes and Skyrmion hair evade the scalar argument; Gauss–Bonnet–dilaton and other modified-gravity holes carry scalar hair. A subtler trap: the surviving $(M,J,Q)$ are global Gauss-law charges, and "no hair" forbids independent localized field profiles, not the charge itself — a black hole certainly can carry charge $Q$ ; what it cannot carry is an extra, independent field configuration on top of it.

Worked examples

1 · Why astrophysical black holes are Kerr, not Kerr–Newman (fully worked)

Take a black hole of one solar mass, $M=M_\odot=1.989\times10^{30}\ \mathrm{kg}$ . How much spin and how much charge would it take to reach the extremal boundary of the uniqueness disk, and what does the answer say about real black holes?

Step 1. Set the length scale. The gravitational radius is $GM/c^2 = (6.674\times10^{-11})(1.989\times10^{30})/(2.998\times10^8)^2 = \mathbf{1.48\ km}$ , so a solar-mass Schwarzschild horizon sits at $r_s=2GM/c^2\approx 2.95$ km.

Step 2. Extremal spin. The bound $\lvert J\rvert\le GM^2/c$ gives $J_{\max}=GM^2/c = (6.674\times10^{-11})(1.989\times10^{30})^2/(2.998\times10^8) = \mathbf{8.8\times10^{41}\ kg\,m^2/s}$ . Astrophysical black holes routinely reach a large fraction of this — measured spins run up to $a/M\approx0.9$ – $0.998$ — so spin is a real, populated axis of the disk.

Step 3. Extremal charge. Saturating $M^2\ge Q^2$ in SI units gives $Q_{\max}=\sqrt{4\pi\varepsilon_0 G}\,M = (8.62\times10^{-11}\ \mathrm{C/kg})(1.989\times10^{30}\ \mathrm{kg}) = \mathbf{1.7\times10^{20}\ C}$ . That is a colossal charge — about $10^{39}$ elementary charges.

Step 4. Interpret. A real black hole sits in a plasma of free charges; any net charge is neutralized almost instantly, because once $Q$ is even a tiny fraction of $Q_{\max}$ the electrostatic force on an infalling proton overwhelms gravity by a factor $\sim10^{18}$ . So although the theorem allows charge, nature drives $Q\to0$ , collapsing the disk to its spin axis. Astrophysical black holes are Kerr to superb accuracy — the charge axis exists in the theorem but is empty in the sky.

2 · Ringdown frequency of the GW150914 remnant

The first detected black-hole merger left a remnant of mass $M_f\approx62\,M_\odot$ and dimensionless spin $a/M\approx0.67$ . Estimate the frequency of its dominant quasinormal ringdown mode, and state why this is a no-hair test.

Step 1. No-hair says the remnant is Kerr, so its quasinormal spectrum $\omega_{\ell m n}$ depends only on $(M_f,J_f)$ . Set the frequency scale $c^3/(GM_f) = (2.998\times10^8)^3/[(6.674\times10^{-11})(62)(1.989\times10^{30})] = \mathbf{3.27\times10^{3}\ s^{-1}}$ .

Step 2. The dominant $(\ell,m,n)=(2,2,0)$ mode of a non-spinning hole has $\mathrm{Re}(M\omega)=0.3737$ , giving $f=\mathrm{Re}(\omega)/2\pi = 0.3737\times3.27\times10^3/(2\pi) = \mathbf{195\ Hz}$ .

Step 3. Spin raises the real frequency; for $a/M\approx0.67$ the $(2,2,0)$ mode has $\mathrm{Re}(M\omega)\approx0.49$ , giving $f\approx0.49\times3.27\times10^3/(2\pi)=\mathbf{255\ Hz}$ — matching the observed ringdown near 250 Hz.

Step 4. The damping time is set by the imaginary part; $\mathrm{Im}(M\omega)\approx0.09$ gives $\tau\sim GM_f/(0.09\,c^3)\approx\mathbf{3\text{–}4\ ms}$ . The no-hair test is consistency: fit $(M,J)$ from the dominant mode, predict the frequencies of the overtone and subdominant modes, and check that a single $(M,J)$ describes them all — a mismatch would signal hair.

3 · No static scalar hair, in one line

Show that a static, regular, massless scalar field around a Schwarzschild black hole must be constant.

Step 1. Multiply $\Box\phi=0$ by $\phi$ and integrate over the static exterior slice: $\int_\Sigma \phi\,\Box\phi\,\sqrt{h}\,\mathrm{d}^3x = 0$ . Staticity removes time derivatives, leaving the spatial Laplacian.

Step 2. Integrate by parts: $\oint_{\partial\Sigma}\phi\,\nabla^\mu\phi\,\mathrm{d}S_\mu - \int_\Sigma(\nabla\phi)^2\sqrt{h}\,\mathrm{d}^3x = 0$ . Regularity keeps $\nabla\phi$ finite at the bifurcation surface and asymptotic falloff kills the infinity term, so the boundary integral vanishes.

Step 3. What remains is $\int_\Sigma(\nabla\phi)^2\sqrt{h}\,\mathrm{d}^3x=0$ with a non-negative integrand, forcing $\nabla\phi=0$ . Hence $\phi=$ constant — no scalar hair. A mass term adds a further non-negative $\int m^2\phi^2$ , reinforcing it; nonlinear Yang–Mills fields lack this positive-definite identity, which is how colored hair sneaks in.

Physics of the portrait

The no-hair theorem is not the work of one hand but a relay, and its opening leg was run by a physicist who took a physicists' folk belief and turned it into a proof.

Origins · 1 minWerner Israel and how this was discovered

Werner Israel

1931–2022 · founder of the black-hole uniqueness program

Werner Israel, born in Berlin in 1931 and raised in South Africa before settling in Canada, changed black-hole physics in 1967 with a single theorem: a static, asymptotically flat vacuum black hole with a regular horizon must be spherically symmetric, hence exactly Schwarzschild. Before that paper, Schwarzschild was widely seen as an idealized special case — a perfectly round star's exterior, unlikely to describe the messy remnant of real collapse. Israel's foliation argument showed the opposite: roundness is forced, not assumed. A year later he extended the result to charged holes, delivering Reissner–Nordström. John Wheeler's slogan, coined around 1970 with Remo Ruffini, captured the reaction: "a black hole has no hair." The rotating case fell to others: Brandon Carter in 1971 reduced it to a two-parameter boundary-value problem, Stephen Hawking's 1972 rigidity theorem supplied the crucial axisymmetry, David Robinson completed the Kerr uniqueness in 1975, and Pawel Mazur and Gary Bunting independently closed the electrovacuum case in the early 1980s with a positivity identity drawn from the equations' hidden sigma-model symmetry; Jacob Bekenstein meanwhile proved that scalar and other linear fields leave no hair. It is Israel's 1967 theorem — the first rigorous statement that collapse erases almost everything — that opened the door the whole field walked through.

History

1967

— Israel proves that static vacuum black holes must be Schwarzschild

1971

— Carter reduces the rotating case to a two-parameter boundary-value problem, and Ruffini and Wheeler popularize the phrase a black hole has no hair

1972

— Hawking's rigidity theorem supplies axisymmetry, and Bekenstein proves scalar fields leave no hair

1975

— Robinson completes the vacuum Kerr uniqueness proof

1982

— Mazur and Bunting independently prove electrovacuum uniqueness via the sigma-model identity

Going further

Texts. Misner, Thorne & Wheeler, Gravitation (1973), Box 33.1 and §33.7 · Wald, General Relativity (1984), §12.3–12.5 · Carroll, Spacetime and Geometry (2004), §6.5 · Hartle, Gravity (2003), Ch. 15 · Heusler, Black Hole Uniqueness Theorems (Cambridge, 1996) · Chruściel, Lopes Costa & Heusler, Stationary Black Holes: Uniqueness and Beyond, Living Rev. Relativity 15 (2012) 7. The key relations used throughout this article are: $\Delta=r^2-2Mr+a^2+Q^2$ ; $r_\pm=M\pm\sqrt{M^2-a^2-Q^2}$ ; the static vacuum reduction $VR_{ij}=\nabla_i\nabla_j V$ , $\nabla^2V=0$ with Schwarzschild lapse $V=\sqrt{1-2M/r}$ ; the scalar no-hair integration-by-parts; $\Omega_H=a/(r_+^2+a^2)$ ; $\kappa=(r_+-r_-)/[2(r_+^2+a^2)]$ ; and the Geroch–Hansen relation $M_\ell+\mathrm{i}S_\ell=M(\mathrm{i}a)^\ell$ , giving $M_2=-J^2/M$ . Worked anchor numbers: $GM_\odot/c^2=1.48$ km; $Q_{\max}\approx1.7\times10^{20}$ C; GW150914 remnant ringdown $f\approx250$ Hz.

Prerequisites. Kerr Metric and Rotating Black Holes · Killing Horizons and Surface Gravity · Schwarzschild Black Hole Structure · Reissner–Nordström Charged Black Hole · Event Horizon and Apparent Horizon. Leads to. Kerr–Newman Solution · Four Laws of Black Hole Mechanics · Hawking Area Theorem · Penrose Process and Energy Extraction.

Ask the tutor: ask me to walk through why a spinning black hole's quadrupole is locked to $-J^2/M$ , or how a single mismatched ringdown frequency would betray hair.

Does "no hair" mean a black hole has literally no features at all?

No — it means no independent features beyond three. A Kerr–Newman black hole is a rich object: it has an ergosphere, inner and outer horizons, an infinite tower of multipoles, a surface gravity, and an angular velocity. The point is that every one of these is a determined function of the same three numbers $(M,J,Q)$ ; fix those and everything else follows with no remaining freedom. "No hair" is shorthand for "no independent parameters," not "no structure." Two black holes with the same $(M,J,Q)$ are identical spacetimes even if one formed from a collapsing star and the other from colliding dust clouds.

Why can't a black hole keep a magnetic field or a baryon number?

Because neither is a long-range charge protected by a Gauss law in the vacuum exterior. A magnetic field threading a horizon is not sourced by a magnetic monopole charge (which does not exist in standard electromagnetism), so a stationary hole cannot anchor an intrinsic dipole field of its own — any nearby field is held there by external plasma currents, not by the hole. Baryon number is worse off: no long-range field couples to it at all, so nothing outside the horizon can measure it. When baryons fall in, only their energy, momentum, and electric charge leave a trace — which is exactly why the information puzzle is sharp.

Is the no-hair theorem actually proven, or is it still a conjecture?

Both, depending on how you state it. The precise electrovacuum version — four dimensions, stationary, asymptotically flat, regular non-degenerate horizon, Einstein–Maxwell matter — is a rigorous theorem, assembled from Israel, Carter, Hawking, Robinson, Mazur, and Bunting, with later work weakening the original analyticity assumption. What remains conjectural is the broad folk claim that "all black holes have no hair for any matter." That is simply false: colored, Skyrmion, dilatonic, and AdS-scalar hairy black holes are explicit counterexamples. So no-hair is a proven theorem for a specific, physically central class and a useful but defeasible principle everywhere else.

What exactly counts as "hair"?

Hair is any field degree of freedom outside the horizon not fixed by the conserved charges $(M,J,Q)$ . An independent quadrupole bump you could tune without changing mass or spin would be hair; a static scalar cloud with its own amplitude would be hair; a Yang–Mills configuration labelled by a discrete "color" index is hair. The test is independence — can you vary the feature while holding the Gauss-law charges fixed and still have a regular stationary black hole? For Kerr–Newman the answer is always no, which is what "bald" means; for colored black holes it is yes, so they are hairy. The word evokes something you could grab to tell two otherwise-identical holes apart.

Does Hawking radiation or the information paradox contradict no-hair?

Not directly, but it sharpens the tension. The no-hair theorem is classical: the exterior geometry forgets everything except $(M,J,Q)$ . Hawking's semiclassical calculation then says the hole radiates a nearly thermal spectrum set by those same few numbers, so the radiation also seems to carry no memory of the progenitor. If both are literally true, the infalling information is lost — conflicting with quantum unitarity, the information paradox. Most resolutions keep the classical theorem but argue quantum gravity supplies subtle "quantum hair," or that information leaks out in correlations of the Hawking radiation. No-hair is not overturned; its quantum fate is the open question.

How is no-hair actually tested with gravitational waves?

Through black-hole spectroscopy. A merger remnant settles by ringing in quasinormal modes — damped sinusoids whose frequencies and decay times are, for a Kerr hole, fixed functions of $(M,J)$ alone. The test is a consistency check: infer $(M,J)$ from the loudest mode, predict where the overtones and weaker modes must lie, and see whether the data agree with a single $(M,J)$ . A mode at a frequency the Kerr spectrum forbids would be evidence of hair or a non-Kerr object. Current LIGO–Virgo results are Kerr-consistent; louder future ringdowns (LISA) will push the precision to where small hair would show.