Helicity and Spin-2 Nature of Gravitons

Every article in this corner of the chapter has circled one number: how many independent ways a gravitational wave can shake spacetime, and what that count says about gravity's quantum. This is where the strands meet. From aq-gr-transverse-traceless-gauge we inherit that a wave along a fixed direction is a symmetric, transverse, traceless 2×2 block of the metric perturbation; from aq-gr-plus-cross-polarizations, that this block is spanned by exactly two amplitudes, $h_+$ and $h_\times$ . The synthesis is a single question with a sharp answer: how do those two amplitudes transform when you rotate your frame about the line of sight? Under such a rotation by angle $\psi$ the complex combinations $h_+ \pm i h_\times$ acquire a phase $e^{\mp 2i\psi}$ . That factor of two — double the rotation angle — is the geometric fingerprint of a rank-2 tensor, and it is precisely what it means to say gravity's quantum carries helicity $\pm 2$ and is a massless spin-2 field. The identification is not a quantization of anything; it is representation theory of the Poincaré group, read directly off the classical field, where masslessness leaves only the two extreme helicities. The rest of the chapter is downstream of this one structure: the helicity-2 fingerprint forces gravitational radiation to begin at the quadrupole (aq-gr-quadrupole-formula), organizes the multipole hierarchy (aq-gr-multipole-expansion-gravitational-radiation), and fixes the energy the two states carry away (aq-gr-isaacson-stress-energy). Spin-2 is the seed; the radiation theory is the tree.

Key takeaways

A gravitational wave along $\hat z$ has only two physical degrees of freedom — the plus and cross polarizations — in the transverse-traceless part of $h_{\mu\nu}$ .
Under a rotation by $\psi$ about the propagation axis, $h_+ \pm i h_\times$ transform with phase $e^{\mp 2i\psi}$ , so the field carries helicity $\lambda = \pm 2$ : gravity's quantum is a massless spin-2 particle, the graviton. The factor of two is the tensor's rank, so the pattern repeats every $180°$ , not $360°$ (contrast the photon, $e^{\mp i\psi}$ ).
Masslessness collapses the naive $2s+1 = 5$ states to just $2$ : the intermediate helicities $0, \pm 1$ are pure gauge, killed by residual diffeomorphisms — the count is $10 - 4 - 4 = 2$ .
Spin-2 dictates quadrupole-dominated radiation, since a massless spin- $s$ field radiates only through multipoles with $\ell \ge s$ ; the same structure is the spin-weight $-2$ Weyl scalar $\psi_4$ of numerical relativity.
The exactly-massless prediction is now tested: LIGO–Virgo bounds put $m_g \lesssim 1.27\times10^{-23}\ \mathrm{eV}/c^2$ , and network polarization tests find pure tensor modes.
Key idea: In the TT gauge a wave along $\hat z$ is a $2\times2$ symmetric traceless tensor in the transverse plane; under a rotation by $\psi$ about $\hat z$ the combinations $h_+\pm i h_\times$ acquire phase $e^{\mp 2i\psi}$ , so the gravitational field carries helicity $\pm 2$ — it is a massless spin-2 field.

Where this lives today

The spin-2 prediction is now a live experimental target on three fronts. First, polarization tests with detector networks: general relativity permits only the two transverse-traceless (helicity $\pm 2$ ) modes, while alternative theories can add scalar "breathing" (helicity $0$ ) or vector (helicity $\pm 1$ ) modes — up to six in general. A network of differently-oriented detectors reads the antenna-pattern response and constrains non-tensor content; the three-detector event GW170814 already favoured pure tensor over pure scalar or vector. Second, waveform modelling in numerical relativity: binary-black-hole simulations extract radiation through the spin-weight $-2$ Weyl scalar $\psi_4$ , expanded in ${}_{-2}Y_{\ell m}$ and dominated by the $(2,2)$ mode — so effective-one-body and surrogate models are organized around these multipoles. Third, graviton-mass bounds: the LIGO–Virgo–KAGRA combined bound $m_g \lesssim 1.27\times10^{-23}\ \mathrm{eV}/c^2$ is fully consistent with an exactly-massless spin-2 field. Direct detection of a single graviton being impractical, the spin-2 nature will keep being tested through the classical field — exactly as it was first derived.

The physical picture

Forget quantization for a moment and watch what a wave does. A ring of freely-falling test masses in the transverse plane is squeezed and stretched as the wave passes: the plus mode $h_+$ elongates the ring along $x$ while pinching $y$ , then reverses; the cross mode $h_\times$ does the same along the diagonals at $45°$ . The ring's area never changes (traceless) and nothing moves along the line of sight (transverse). Two patterns, two amplitudes — and the whole spin story hides in how they turn into each other when you rotate your head.

Figure

Two free-mass rings under the plus and cross polarizations across one wave period: h₊ stretches along x and squeezes along y, h× does the same along the 45° diagonals, so a 45° frame rotation turns + into × while the whole pattern returns to itself only after 180° — the geometric signature of spin 2, h₊ ± i h× → e^∓2iψ (h₊ ± i h×)

Rotate your frame by $45°$ and a pure-plus pattern becomes a pure-cross pattern; rotate by a full $180°$ and it is back where it started. A configuration that returns to itself every half-turn is the signature of a spin-2 object, just as electric field lines (spin-1) return every full turn and a scalar (spin-0) never changes. Helicity is just the winding rate of the polarization phase — once per revolution for light, twice for gravity. The second figure lays out this ladder and shows why masslessness leaves only the two extreme rungs.

Figure

The massless helicity ladder: with mass a spin-s field would carry 2s+1 states, but masslessness leaves only the two extreme helicities λ = ±s. A scalar keeps 0, the photon ±1, the graviton ±2; the intermediate helicities are pure gauge, and the reduction 10 − 4 − 4 = 2 fixes the graviton's two physical polarizations

Why only two states and not five? A massive spin-2 particle can be brought to rest, where the full rotation group $SO(3)$ acts and all $2s+1 = 5$ orientations are physical. A massless particle has no rest frame; only the little group of a null momentum — $\mathrm{ISO}(2)$ of the transverse plane — acts, and gauge invariance removes everything except the two extreme helicities. The graviton's masslessness is not a detail: it is why gravity has exactly two polarizations, not six.

The core equations

The starting point is inherited: in the TT gauge a wave along $\hat z$ collapses to a single $2\times2$ block carrying two amplitudes. Everything else is how that block rotates.

h_{ij}^{\mathrm{TT}} = \begin{pmatrix} h_+ & h_\times & 0 \\ h_\times & -h_+ & 0 \\ 0 & 0 & 0 \end{pmatrix} \cos[\omega(t-z)] \qquad (1)

where: $h_{ij}^{\mathrm{TT}}$ — transverse-traceless metric perturbation; $h_+,\,h_\times$ — plus and cross amplitudes; $\omega$ — angular frequency; $c=1$ along $+\hat z$ . The zero row and column are transversality; the equal-and-opposite diagonal is tracelessness. The two amplitudes multiply two fixed basis tensors built from the transverse axes.

e^{+}_{ij} = \hat x_i\hat x_j - \hat y_i\hat y_j,\qquad e^{\times}_{ij} = \hat x_i\hat y_j + \hat y_i\hat x_j \qquad (2)

where: $e^+_{ij}$ — plus basis tensor (stretch $x$ , squeeze $y$ ); $e^\times_{ij}$ — cross basis tensor (the same rotated $45°$ ); $\hat x,\hat y$ — orthonormal axes transverse to $\hat z$ . The natural basis for rotations is not plus and cross but their complex circular combinations, which a rotation multiplies by a pure phase rather than mixing.

e^{R/L}_{ij} = \frac{1}{\sqrt{2}}\left(e^+_{ij} \mp i\,e^\times_{ij}\right) \qquad (3)

where: $e^{R}_{ij}$ — right-handed tensor, helicity $+2$ ; $e^{L}_{ij}$ — left-handed tensor, helicity $-2$ ; these are the eigenvectors of the rotation generator about $\hat z$ . The central result is how the physical amplitudes transform under a rotation $R(\psi)$ about the propagation axis — this one line is the entire spin-2 identification.

(h_+ \pm i h_\times) \xrightarrow{\ R(\psi)\ } e^{\mp 2i\psi}\,(h_+ \pm i h_\times) \qquad (4)

where: $\psi$ — rotation angle about $\hat z$ ; the exponent $\mp 2$ is the helicity eigenvalue $\lambda = \pm 2$ — the field winds twice per turn of the frame. Helicity is the projection of angular momentum onto the direction of motion; for a massless particle it is a Lorentz invariant taking only two values.

\lambda = \frac{\vec J\cdot\vec p}{|\vec p|}, \qquad \lambda \in \{+2,\,-2\} \qquad (5)

where: $\vec J$ — angular momentum; $\vec p$ — propagation direction; $\lambda$ — spin projection. Because a massless particle cannot be overtaken, no boost flips $\lambda$ , and the two values label distinct particles. That only two survive out of ten is a gauge-counting statement.

N_{\mathrm{phys}} = \underbrace{10}_{h_{\mu\nu}} - \underbrace{4}_{\text{Lorenz}} - \underbrace{4}_{\text{residual}} = 2 \qquad (6)

where: $10$ — components of symmetric $h_{\mu\nu}$ ; $4$ — Lorenz conditions $\partial^\mu \bar h_{\mu\nu}=0$ ; $4$ — residual gauge reaching TT gauge. The last four exist only because the field is massless. The helicity-2 structure resurfaces operationally as the Weyl scalar $\psi_4$ , a spin-weight $-2$ field expanded in harmonics that start at $\ell=2$ .

\psi_4 = -\ddot h_+ + i\,\ddot h_\times = \sum_{\ell\ge 2}\sum_{m=-\ell}^{\ell} \psi_4^{\ell m}\,{}_{-2}Y_{\ell m}(\theta,\phi) \qquad (7)

where: $\psi_4$ — outgoing-radiation Weyl scalar, a helicity $-2$ field; ${}_{-2}Y_{\ell m}$ — spin-weighted spherical harmonics of spin weight $-2$ ; the sum starts at $\ell\ge 2$ as a direct consequence of spin 2. Finally, only these two states carry energy: the Isaacson effective stress-energy (owned by aq-gr-isaacson-stress-energy) averages over many wavelengths.

\langle t_{00}\rangle = \frac{1}{32\pi G}\,\big\langle \dot h_+^2 + \dot h_\times^2\big\rangle \qquad (8)

where: $t_{00}$ — Isaacson effective energy density; $G$ — Newton's constant; $\dot h_{+,\times}$ — the amplitudes' time derivatives. Two terms, two helicities; there is no third.

Derivation AFrom the transverse-traceless tensor to helicity ±2

The goal is to compute, from scratch, how the two polarization amplitudes transform when the observer rotates their frame about the propagation axis, and to show the result is a pure phase $e^{\mp 2i\psi}$ whose exponent — the number two — is forced by the metric perturbation being a rank-2 tensor rather than a vector or scalar. That single number is the whole content of "spin 2", so the derivation is short but load-bearing.

Step 1. Work in the transverse plane spanned by $\hat x,\hat y$ . The TT conditions collapse the perturbation to $h^{\mathrm{TT}}_{ij}=h_+\,e^+_{ij}+h_\times\,e^\times_{ij}$ , a symmetric traceless $2\times2$ block with exactly two amplitudes. Why: transversality and tracelessness are the constraints inherited from aq-gr-transverse-traceless-gauge, leaving two amplitudes rather than the six of a general symmetric spatial tensor.

Step 2. Apply an active rotation about $\hat z$ : as a rank-2 tensor, $h_{ij}$ is rotated on both indices, $h'_{ij}=R_{ik}R_{jl}\,h^{\mathrm{TT}}_{kl}$ , with $R(\psi)$ acting as $\hat x\to\cos\psi\,\hat x+\sin\psi\,\hat y$ . Why: this is the one place tensor rank enters — a vector is hit by one factor of $R$ , a scalar by none; the two factors here manufacture the double angle.

Step 3. Carry out the contraction. The two factors of $R$ combine trigonometrically, and with the trace and longitudinal parts absent no other structures appear: $h'_+=\cos2\psi\,h_+ +\sin2\psi\,h_\times$ , $h'_\times=-\sin2\psi\,h_+ +\cos2\psi\,h_\times$ . Why: the products $\cos^2\psi-\sin^2\psi$ and $2\sin\psi\cos\psi$ are the double-angle identities — the "2" is born here, from multiplying two single-angle rotations.

Step 4. Diagonalize with the complex combinations: $h'_+ \pm i h'_\times = (\cos2\psi \mp i\sin2\psi)(h_+\pm i h_\times)=e^{\mp 2i\psi}(h_+\pm i h_\times)$ . Why: mixing under rotation becomes a pure phase in this basis — these are eigenstates of the rotation generator, and the phase is the eigenvalue.

Step 5. Read off the helicity. Writing $R(\psi)=e^{-i\psi J_z}$ , the eigenvalue $e^{\mp 2i\psi}$ means $J_z(h_+\pm i h_\times)=\pm 2\,(h_+\pm i h_\times)$ , i.e. $\lambda=\pm 2$ . Why: this is helicity applied to the classical field, identifying the graviton as massless spin-2 without any creation operator. *A numeric check confirms $(1,0)$ maps to $(0,-1)$ at $\psi=45°$ and returns at $\psi=180°$ — period $\pi$ , the hallmark of spin 2.*

Go deeper · Derivation B · 1 minWhy only two polarizations survive — the gauge count

Derivation BWhy only two polarizations survive — the gauge count

The goal is to trace the reduction of the ten independent components of the symmetric perturbation $h_{\mu\nu}$ down to two physical helicity states, making explicit which constraint removes which components and — crucially — showing that the last four removals exist only because the field is massless. The same arithmetic on a massive spin-2 field would leave five states; the gap of three is the fingerprint of gauge invariance.

Step 1. Start from the trace-reversed perturbation $\bar h_{\mu\nu}=h_{\mu\nu}-\tfrac12\eta_{\mu\nu}h$ , with field equation $\Box \bar h_{\mu\nu}=-16\pi G\,T_{\mu\nu}$ . Imposing the Lorenz gauge $\partial^\mu \bar h_{\mu\nu}=0$ supplies four conditions, cutting ten components to six. Why: the Lorenz gauge is the linearized shadow of coordinate freedom — recalled from aq-gr-gauge-freedom-linearized-gravity.

Step 2. Look for a vacuum plane wave $\bar h_{\mu\nu}=A_{\mu\nu}e^{ik\cdot x}$ . The field equation forces $k^2=0$ (masslessness) and the gauge forces transversality $k^\mu A_{\mu\nu}=0$ . Why: the null condition $k^2=0$ does quiet but essential work — it is what permits the residual gauge freedom of the next step; a massive field has none.

Step 3. Exploit the residual gauge: a shift $x^\mu\to x^\mu+\xi^\mu$ with harmonic $\Box\xi^\mu=0$ preserves the Lorenz condition yet changes $A_{\mu\nu}\to A_{\mu\nu}-i(k_\mu\varepsilon_\nu+k_\nu\varepsilon_\mu-\eta_{\mu\nu}k_\lambda\varepsilon^\lambda)$ . Four functions $\varepsilon^\mu$ remove four more components. Why: these four parameters exist only because $k^2=0$ admits nontrivial harmonic $\xi^\mu$ — masslessness is again the enabling ingredient.

Step 4. Spend that freedom to reach the TT gauge, $A_{0\mu}=0$ and $A^i{}_i=0$ , killing the timelike row and spatial trace. What remains is exactly the $2\times2$ block $A_+$ , $A_\times$ . Why: this canonical choice isolates the physical, gauge-invariant content — the block whose rotation we diagonalized above.

Step 5. Count: $10-4-4=2$ . Had the field been massive, the residual freedom of Step 3 would not exist, and $10-4=6$ minus the trace constraint would leave the five states of the spin-2 multiplet. Why: the gap between two and five is exactly the three would-be helicity $0,\pm1$ modes — pure gauge for a massless graviton, physical for a massive one. *The count $10-4-4=2$ is convention-independent.*

Go deeper · Derivation C · 1 minHow spin 2 forces the quadrupole

Derivation CHow spin 2 forces the quadrupole

The goal is to show that the spin-2 nature is not a labelling convention but has a hard dynamical consequence: gravitational radiation must begin at the mass quadrupole, with no monopole or dipole. The argument runs on two independent tracks that agree — a helicity/spin-weight argument forbidding low multipoles kinematically, and a conservation-law argument forbidding them physically — both landing on $\ell_{\min}=s=2$ , which is why the radiation theory of the next chapter opens at the quadrupole formula.

Step 1. A helicity- $s$ radiation field is expanded in spin-weighted harmonics ${}_{-s}Y_{\ell m}$ , which do not exist for $\ell<|s|$ . For the graviton $s=2$ , so only $\ell\ge 2$ can appear — there is no $\ell=0$ or $\ell=1$ basis function to carry a monopole or dipole. Why: this is representation theory on the sphere, independent of any source detail, and it is why $\psi_4$ decompositions start at $(2,2)$ .

Step 2. Independently, conservation laws forbid the low poles: monopole radiation would need $\dot M\ne 0$ , but mass-energy is conserved; dipole radiation would need $\ddot{\vec d}\ne0$ , but $\dot{\vec d}=\vec P$ is the conserved total momentum. Why: mass and momentum conservation mirror the kinematic ban of Step 1 — two routes to the same missing terms, the mark of a robust result.

Step 3. The leading surviving moment is therefore the mass quadrupole, and the field equation delivers $\bar h^{\mathrm{TT}}_{ij}=\dfrac{2G}{r}\,\ddot I^{\mathrm{TT}}_{ij}(t-r)$ with $I_{ij}=\int \rho\,(x_ix_j-\tfrac13\delta_{ij}r^2)\,\mathrm{d}^3x$ . Why: this is the $\ell=2$ moment and nothing lower, as both arguments demand — the full derivation is owned by aq-gr-quadrupole-formula.

Step 4. Consistency check: the outgoing waveform organizes as $\psi_4=\sum_{\ell\ge 2,m}\psi_4^{\ell m}\,{}_{-2}Y_{\ell m}$ , dominated by the $(2,2)$ mode for a circular binary. Why: every LIGO waveform begins at $\ell=2$ for the same spin-2 reason — the helicity count is directly visible in the data. *As a consistency check, the rule $\ell_{\min}=s$ reproduces $\ell_{\min}=1$ for electromagnetic dipole radiation.*

Limiting cases

The helicity content is sharpest when the same null-momentum little group is applied to fields of different spin — the graviton's " $2$ " then reads as one entry in a ladder.

Case	Result	Meaning
Scalar (spin 0)	$\phi \to \phi$	Invariant under transverse rotations; helicity $0$ , one state. A hypothetical scalar gravity component — absent in pure GR.
Photon (spin 1)	$\epsilon_\pm \to e^{\mp i\psi}\,\epsilon_\pm$	Transverse polarization vector winds once per turn; helicity $\pm 1$ , two states. Longitudinal and timelike modes are pure gauge.
Graviton (spin 2)	$(h_+ \pm i h_\times) \to e^{\mp 2i\psi}\,(h_+ \pm i h_\times)$	Symmetric transverse-traceless tensor winds twice per turn; helicity $\pm 2$ , two states — the case of this article.
Massive spin-2 (contrast)	$2s+1 = 5$ states	With a rest frame the full $SO(3)$ acts; five polarizations. Massless gravity keeps only two — the three would-be longitudinal modes are gauged away.

Worked examples

1 · The 45° rotation and the 180° periodicity (confidence builder)

Take a pure plus wave, $(h_+,h_\times)=(1,0)$ , and rotate the frame about the propagation axis by $\psi=45°$ , then by $\psi=180°$ , using $h'_+=\cos2\psi\,h_+ +\sin2\psi\,h_\times$ , $h'_\times=-\sin2\psi\,h_+ +\cos2\psi\,h_\times$ . What happens, and how does it show helicity 2?

Step 1. At $\psi=45°$ the relevant angle is $2\psi=90°$ , so $\cos2\psi=0$ and $\sin2\psi=1$ . Then $h'_+=0\cdot1+1\cdot0=0$ and $h'_\times=-1\cdot1+0\cdot0=-1$ . A pure plus wave has become a pure cross wave (with a sign): a 45° frame rotation maps + into ×.

Step 2. At $\psi=180°$ the angle is $2\psi=360°$ , so $\cos2\psi=1$ , $\sin2\psi=0$ , and $(h'_+,h'_\times)=(1,0)$ — the pattern returns to itself after only a half-turn of the frame, not a full turn.

Step 3. Cross-check with the phase form (equation 4): $h_++ih_\times=1$ picks up $e^{-2i\psi}$ . At $\psi=45°$ that is $e^{-i\pi/2}=-i$ , giving $(h'_+,h'_\times)=(0,-1)$ — identical to Step 1; at $\psi=180°$ it is $e^{-2i\pi}=1$ , back to the start. The exponent $2$ is exactly the helicity, and its consequence is the $\pi$ -periodicity. Contrast a photon, where $e^{-i\psi}$ needs a full $360°$ to return.

2 · How stringent is the massless-graviton prediction?

The GWTC-3 dispersion analysis bounds a graviton mass at $m_g \le 1.27\times10^{-23}\ \mathrm{eV}/c^2$ . Translate this into a reduced Compton wavelength and estimate the fractional slowdown $1-v_g/c$ it would allow for a $100\ \mathrm{Hz}$ wave. Use $\hbar c = 1.973\times10^{-7}\ \mathrm{eV\,m}$ and $h = 4.136\times10^{-15}\ \mathrm{eV\,s}$ .

Step 1. A nonzero mass gives a reduced Compton wavelength $\lambda_g=\hbar/(m_g c)=\hbar c/(m_g c^2)$ . Inserting the bound: $\lambda_g = (1.973\times10^{-7})/(1.27\times10^{-23}) = \mathbf{1.55\times10^{16}\ m}$ , i.e. about $\mathbf{1.6\times10^{13}\ km}$ — larger than any binary source, so gravity is indistinguishable from massless.

Step 2. The dispersion relation $v_g/c=\sqrt{1-(m_g c^2/E)^2}$ gives $1-v_g/c\approx\tfrac12(m_g c^2/E)^2$ for a tiny mass. A $100\ \mathrm{Hz}$ wave has quantum energy $E=hf=4.14\times10^{-13}\ \mathrm{eV}$ .

Step 3. Then $m_g c^2/E = (1.27\times10^{-23})/(4.14\times10^{-13}) = 3.07\times10^{-11}$ , so the allowed fractional slowdown is $1-v_g/c \approx \tfrac12(3.07\times10^{-11})^2 = \mathbf{4.7\times10^{-22}}$ — an arrival-time smearing of at most milliseconds across the band for a source hundreds of Mpc away. The non-detection of any such smearing turns "the graviton is massless spin-2" from an assumption into a measured bound: the two helicity states arrive together, with no third mode and no dispersion.

Physics of the portrait

The spin-2 identification is a fact about the symmetry group of spacetime, not about quantum gravity. The person who built the language is Eugene Wigner, who asked what an elementary particle even is and answered: an irreducible representation of the Poincaré group, labelled by mass and, for a massless particle, one number — the helicity.

Origins · 1 minEugene Wigner and how this was discovered

Eugene Wigner

1902–1995 · classification of relativistic particles by the little group of the Poincaré group

Eugene Paul Wigner, a Hungarian-American architect of twentieth-century symmetry, gave "what is a particle?" its modern answer. In a landmark 1939 paper, "On unitary representations of the inhomogeneous Lorentz group," he classified every relativistic particle by how it transforms under the Poincaré group. His key device was the little group: the subgroup that leaves a particle's momentum fixed. For a massive particle at rest the little group is the full rotation group $SO(3)$ , whose representations carry spin $s$ with $2s+1$ states. For a massless particle, with no rest frame, the momentum is null and the little group is instead $\mathrm{ISO}(2)$ ; its finite representations are labelled by a single number — the helicity. This is why a photon has two states rather than three, and a graviton two rather than five: masslessness replaces the rotation group by $\mathrm{ISO}(2)$ , and only the extreme helicities survive. When Fierz and Pauli that same year wrote the consistent spin-2 field theory, and Weinberg later showed a massless spin-2 particle must couple to conserved energy-momentum — recovering general relativity — they were adding dynamics to the skeleton Wigner had drawn. He shared the 1963 Nobel Prize for his symmetry principles; the graviton's helicity $\pm 2$ is, quite literally, an entry in Wigner's table.

History

1918

— Noether ties conservation laws to symmetry, the seed of the momentum arguments forbidding low-multipole radiation

1939

— Wigner classifies relativistic particles by the little group of their momentum, defining helicity for the massless case

1939

— Fierz and Pauli write the consistent Lagrangian for a massive spin-2 field, exposing its five states

1965

— Weinberg shows a soft massless spin-2 particle must couple universally to conserved energy-momentum, forcing general relativity

1970

— van Dam, Veltman and Zakharov find the discontinuity of the massless limit

2017

— the LIGO-Virgo observation of GW170814 first constrains gravitational-wave polarizations, favouring pure tensor modes

Going further

Texts. Misner, Thorne & Wheeler, Gravitation (1973), §35.6 and Box 35.1 · Wald, General Relativity (1984), §4.4 · Carroll, Spacetime and Geometry (2004), §7.1–7.4 · Maggiore, Gravitational Waves Vol. 1 (2008), §2.2 and §3.5 · Weinberg, Gravitation and Cosmology (1972), §10.2 · Schutz, A First Course in General Relativity (2009), Ch. 9. The rank-2 rotation $h'_{ij}=R_{ik}R_{jl}h_{kl}$ yields the double-angle structure and eigenvalue $e^{\mp 2i\psi}$ ; the degree-of-freedom count is $10-4-4=2$ . The little-group classification is due to Wigner (1939). The bound $m_g\lesssim1.27\times10^{-23}\ \mathrm{eV}/c^2$ with $\lambda_g\gtrsim1.6\times10^{13}\ \mathrm{km}$ are LIGO–Virgo–KAGRA GWTC-3 values; the fractional slowdown $\sim4.7\times10^{-22}$ at $100\ \mathrm{Hz}$ follows from $1-v_g/c\approx\tfrac12(m_gc^2/E)^2$ .

Prerequisites. Transverse-Traceless Gauge · Plus and Cross Polarizations · Gauge Freedom in Linearized Gravity · Linearized Einstein Equations. Leads to. Quadrupole Formula · Multipole Expansion of Gravitational Radiation · Isaacson Stress-Energy of Gravitational Waves · Geodesic Deviation and Gravitational Waves.

Ask the tutor: Ask me to rotate the polarization tensors by hand, contrast the photon's helicity ±1 with the graviton's ±2, or trace exactly how spin-2 forces radiation to start at the quadrupole.

Is "spin-2" a quantum statement, or can I derive it classically?

It is entirely classical here. Nothing in the derivation quantizes a field or introduces a creation operator — we ask how the classical amplitudes $h_+,h_\times$ transform under a rotation about the propagation axis and find the phase $e^{\mp 2i\psi}$ . Helicity is the eigenvalue of the rotation generator, so reading off the exponent $\pm 2$ is the spin-2 identification. Quantization later reinterprets these two states as the graviton's polarizations, but the group-theoretic content — a rank-2 traceless tensor carries helicity $\pm 2$ — is a fact about Poincaré representations, true whether or not gravitons are ever detected.

Where exactly does the factor of two in e^∓2iψ come from?

From the rank of the tensor, and nothing else. The metric perturbation $h_{ij}$ has two indices, and a rotation acts on each: $h'_{ij}=R_{ik}R_{jl}h_{kl}$ . Two factors of a single-angle rotation combine through $\cos^2\psi-\sin^2\psi=\cos2\psi$ and $2\sin\psi\cos\psi=\sin2\psi$ , producing double-angle functions. A vector (rank 1, like the photon's potential) is hit by one factor of $R$ and winds once, $e^{\mp i\psi}$ ; a scalar (rank 0) is untouched. So the "2" counts indices, not any physical doubling of rotation rate — the pattern merely repeats after a half-turn.

Why does masslessness reduce five states to two?

A massive spin-2 particle can be brought to rest, where the full rotation group $SO(3)$ acts and all $2s+1=5$ orientations are physically distinct. A massless particle travels at $c$ with no rest frame, so $SO(3)$ never acts; the subgroup fixing its null momentum — Wigner's little group $\mathrm{ISO}(2)$ — acts instead, and its representations carry a single helicity. Gauge invariance removes the would-be helicity $0$ and $\pm 1$ states, leaving only $\pm 2$ — the same reason a photon has two polarizations, not three. Masslessness, not spin, sets the count.

If a detector saw a "breathing" mode, what would it mean?

It would mean a helicity-0 (scalar) polarization, in which a ring of test masses expands and contracts isotropically rather than being stretched one way and squeezed the perpendicular way. Pure general relativity forbids this: its two allowed modes, plus and cross, are traceless and therefore area-preserving, carrying helicity $\pm 2$ only. A confirmed breathing mode would require a scalar-tensor or more general metric theory and would falsify GR's strict spin-2 prediction. Current detector-network data, notably the three-detector event GW170814, favour pure tensor polarization with no evidence for scalar or vector modes.

How is the spin-2 structure actually used in real gravitational-wave analysis?

Operationally, through the Weyl scalar $\psi_4$ , which encodes the outgoing radiation as a spin-weight $-2$ quantity. Numerical-relativity codes extract $\psi_4$ and expand it in ${}_{-2}Y_{\ell m}$ , a basis that exists only for $\ell\ge 2$ precisely because the field has helicity $\pm 2$ . For a circular binary the $(2,2)$ mode dominates, and waveform models (effective-one-body, phenomenological, surrogate) are organized mode-by-mode in this basis. So every template matched against LIGO data is built on the helicity-2 structure derived here — not an abstraction but the daily working language of the field.

Does the graviton mass bound prove the graviton is exactly massless?

No experiment can prove an exactly zero mass; it can only push the upper bound down. The current LIGO–Virgo–KAGRA bound $m_g\lesssim1.27\times10^{-23}\ \mathrm{eV}/c^2$ corresponds to a reduced Compton wavelength above about $1.6\times10^{13}\ \mathrm{km}$ — larger than any binary source — so gravity behaves as massless over everything we observe. A nonzero mass is also theoretically awkward: it would restore the helicity-0 mode and, through the van Dam–Veltman–Zakharov discontinuity, fail to smoothly reproduce GR in the massless limit. Tight empirical bound plus these consistency concerns is why GR's spin-2 graviton is taken to be exactly massless.