Einstein field equations

Intuition [Beginner]

Gravity is not a force. It is the shape of space and time. A heavy object like the Sun bends spacetime around itself, and everything nearby -- planets, light rays, clocks -- responds to that bend. The Einstein field equations are the rule that tells you exactly how much bending a given amount of matter produces.

Think of a stretched rubber sheet. Place a bowling ball on it. The sheet sags. Roll a marble nearby and it curves toward the bowling ball, not because the ball pulls the marble, but because the marble is following the curve of the sheet. In general relativity the "sheet" is spacetime itself, and the "bowling ball" is any concentration of mass and energy.

The equation comes in two halves. On the left: the Einstein tensor $G_{\mu \nu }$, which describes how spacetime is curved. On the right: the stress-energy tensor $T_{\mu \nu }$, which describes how matter and energy are distributed. The proportionality constant $8\pi G/c^4$ sets the scale -- it says that you need an enormous amount of energy to produce even a tiny amount of curvature, which is why gravity is so weak compared to electromagnetism.

In words: matter tells spacetime how to curve; curved spacetime tells matter how to move. The first half is the field equation itself. The second half is the geodesic equation, which follows from the curvature the field equation produces.

Visual [Beginner]

The picture captures the core idea: the amount of grid distortion (left side) equals the amount of matter (right side), scaled by the coupling constant. Where there is no matter, the grid returns to flatness. Where matter is dense, the grid bends sharply.

Worked example [Beginner]

A vacuum is a region with no matter or energy -- so $T_{\mu \nu }=0$ everywhere. What do the Einstein equations say about spacetime in a vacuum?

The Einstein field equations are $G_{\mu \nu }=\frac{8\pi G}{c^4}{T}_{\mu \nu }$. If $T_{\mu \nu }=0$, then $G_{\mu \nu }=0$. The Einstein tensor is defined as $G_{\mu \nu }=R_{\mu \nu }-\frac{1}{2}R{g}_{\mu \nu }$, where $R_{\mu \nu }$ is the Ricci tensor and $R$ is the scalar curvature.

Taking the trace of both sides by contracting with $g^{\mu \nu }$ gives $G=R-\frac{1}{2}(4)R=-R$. So $G_{\mu \nu }=0$ means $R_{\mu \nu }-\frac{1}{2}R{g}_{\mu \nu }=0$. Taking the trace: $R-2R=-R=0$, so $R=0$. Substituting back: $R_{\mu \nu }=0$.

In a vacuum the Einstein equations reduce to $R_{\mu \nu }=0$. This does not mean spacetime is flat -- the full Riemann curvature tensor can still be nonzero. It means only the contracted part vanishes. The Schwarzschild solution (describing spacetime outside a spherical mass) satisfies $R_{\mu \nu }=0$ while having genuine curvature, including the gravitational field that bends light and slows clocks.

Minkowski space (flat spacetime with metric ${\eta }_{\mu \nu }$) is also a vacuum solution: all Christoffel symbols vanish, so $R_{\mu \nu }=0$ and $G_{\mu \nu }=0$.

Check your understanding [Beginner]

Formal definition [Intermediate+]

The Einstein field equations (EFE) are

$$\boxed{G_{\mu \nu }+\Lambda g_{\mu \nu }=\frac{8\pi G}{c^4}{T}_{\mu \nu },}$$

where:

• $G_{\mu \nu }=R_{\mu \nu }-\frac{1}{2}R{g}_{\mu \nu }$ is the Einstein tensor 13.03.01 pending,
• $R_{\mu \nu }$ is the Ricci tensor of the Levi-Civita connection of the metric $g_{\mu \nu }$,
• $R=g^{\mu \nu }{R}_{\mu \nu }$ is the Ricci scalar (scalar curvature),
• $\Lambda$ is the cosmological constant (discussed below),
• $T_{\mu \nu }$ is the stress-energy tensor 11.01.01 pending,
• $G$ is Newton's gravitational constant,
• $c$ is the speed of light.

The Einstein tensor is symmetric ($G_{\mu \nu }=G_{\nu \mu }$) because $R_{\mu \nu }$ is symmetric and $g_{\mu \nu }$ is symmetric. In four dimensions the equation represents 10 independent component equations (a symmetric $4\times 4$ tensor has 10 independent components after accounting for symmetry).

The trace-reversed form

Contracting the EFE with $g^{\mu \nu }$ and using $g^{\mu \nu }{G}_{\mu \nu }=R-2R=-R$ in four dimensions gives $-R+4\Lambda =8\pi GT/c^4$, where $T=g^{\mu \nu }{T}_{\mu \nu }$ is the trace of the stress-energy tensor. Substituting back:

$$R_{\mu \nu }=\frac{8\pi G}{c^4}(T_{\mu \nu }-\frac{1}{2}T{g}_{\mu \nu })+\Lambda g_{\mu \nu }.$$

In vacuum ($T_{\mu \nu }=0$, $\Lambda =0$) this reduces to $R_{\mu \nu }=0$, as derived in the worked example.

The cosmological constant

The term $\Lambda g_{\mu \nu }$ can be moved to the right side and absorbed into the stress-energy tensor as a contribution $T_{\mu \nu }^{(\Lambda )}=-\frac{c^4\Lambda }{8\pi G}{g}_{\mu \nu }$. This is the stress-energy of the vacuum: a uniform energy density ${\rho }_{\Lambda }=\frac{c^2\Lambda }{8\pi G}$ with pressure $p_{\Lambda }=-{\rho }_{\Lambda }{c}^2$. The cosmological constant is not a fudge or an afterthought -- it is the vacuum energy of spacetime, and observational cosmology (supernova data, CMB, baryon acoustic oscillations) has established $\Lambda >0$ with ${\rho }_{\Lambda }\approx 5.96\times 10^{-27}{\text{kg/m}}^3$, driving the accelerated expansion of the universe.

The observed value of $\Lambda$ corresponds to an energy density of about $5.4\times 10^{-10}{\text{J/m}}^3$. The discrepancy between this value and the naive quantum-field-theory estimate of $\sim 10^{113}{\text{J/m}}^3$ is the cosmological constant problem -- one of the deepest unsolved problems in theoretical physics.

The Bianchi identity and conservation of energy

The contracted second Bianchi identity 13.03.01 pending states

$${\nabla }^{\mu }{G}_{\mu \nu }=0.$$

This is a geometric identity, not a physical postulate -- it follows from the definition of $G_{\mu \nu }$ in terms of the Riemann tensor. Taking the covariant divergence of the EFE then gives

$${\nabla }^{\mu }{T}_{\mu \nu }=0,$$

which is the local conservation of energy and momentum. This is not an independent physical law in GR -- it is a consequence of the field equations, enforced by the geometry. Any matter model that fails to satisfy ${\nabla }^{\mu }{T}_{\mu \nu }=0$ is inconsistent with the Einstein equations.

The Newtonian limit

In the weak-field, slow-motion limit, the metric is approximately $g_{\mu \nu }={\eta }_{\mu \nu }+h_{\mu \nu }$ with $|h_{\mu \nu }|\ll 1$. For a static, pressureless source ($T_{tt}=\rho c^2$, all other components negligible), the $tt$ component of the EFE reduces to

$${\nabla }^2\Phi =4\pi G\rho ,$$

where $\Phi$ is the Newtonian gravitational potential related to $h_{tt}$ by $h_{tt}=-2\Phi /c^2$. This is Poisson's equation for Newtonian gravity. The factor $8\pi G/c^4$ in the EFE was chosen by Einstein to reproduce exactly this limit, and this fixes the coupling constant with no remaining freedom.

Counterexamples and common slips

• The EFE is a set of 10 coupled nonlinear partial differential equations for the 10 components of $g_{\mu \nu }$. Because it is nonlinear, you cannot superpose solutions. Two Schwarzschild spacetimes placed side by side do not produce a valid two-body solution -- the field of each one changes the spacetime in which the other lives.
• The vacuum equations $R_{\mu \nu }=0$ do not mean "no gravity." They mean "no matter." Gravitational fields persist in vacuum -- this is the content of the Schwarzschild solution and of gravitational waves.
• The cosmological constant $\Lambda$ is not a correction to the field equations. It can be moved to either side of the equation. On the left, it is a property of the geometry. On the right, it is the vacuum energy. The two interpretations are mathematically equivalent; the difference is interpretive.
• The EFE determines the metric $g_{\mu \nu }$ given the matter distribution $T_{\mu \nu }$, but $T_{\mu \nu }$ itself depends on $g_{\mu \nu }$ (for instance, the fluid stress-energy involves $g_{\mu \nu }$ in the normalisation of the four-velocity). The equations are therefore coupled: geometry determines the motion of matter, and matter determines the geometry.

Key theorem with proof [Intermediate+]

Theorem (Uniqueness of the Einstein equations). The Einstein tensor equation $G_{\mu \nu }=\kappa T_{\mu \nu }$ (with $G_{\mu \nu }=R_{\mu \nu }-\frac{1}{2}R{g}_{\mu \nu }$) is the unique symmetric $(0,2)$-tensor equation for the metric that satisfies all three of the following:

(a) Newtonian limit. In the weak-field, slow-motion limit, the equation reproduces Poisson's equation ${\nabla }^2\Phi =4\pi G\rho$.

(b) Divergence-free. The left side satisfies ${\nabla }^{\mu }{G}_{\mu \nu }=0$ identically (the contracted Bianchi identity), ensuring energy-momentum conservation ${\nabla }^{\mu }{T}_{\mu \nu }=0$.

(c) Linear in second derivatives. The equation involves at most second derivatives of the metric, and these appear linearly (the Riemann tensor is linear in second derivatives of $g_{\mu \nu }$).

Proof. We establish uniqueness by exhausting the possibilities.

Step 1: Available tensors. Any tensor constructed from the metric and its derivatives that is symmetric and of type $(0,2)$ is, to lowest meaningful order, a linear combination of $g_{\mu \nu }$, $R_{\mu \nu }$, and $R{g}_{\mu \nu }$. Higher-order terms (products of curvature tensors, derivatives of curvature) involve more than second derivatives of $g_{\mu \nu }$ or are nonlinear in second derivatives.

Step 2: Linearity in second derivatives. Condition (c) restricts the left side to the most general form linear in second derivatives of $g_{\mu \nu }$:

$$A_{\mu \nu }=\alpha R_{\mu \nu }+\beta R{g}_{\mu \nu }+\Lambda g_{\mu \nu },$$

where $\alpha$, $\beta$, and $\Lambda$ are constants. The term $\Lambda g_{\mu \nu }$ involves zero derivatives of the metric.

Step 3: Divergence-free condition. Computing ${\nabla }^{\mu }{A}_{\mu \nu }=0$ and using the contracted Bianchi identity ${\nabla }^{\mu }(R_{\mu \nu }-\frac{1}{2}R{g}_{\mu \nu })=0$:

$${\nabla }^{\mu }{A}_{\mu \nu }=\alpha {\nabla }^{\mu }{R}_{\mu \nu }+\beta {\nabla }_{\nu }R.$$

From the contracted Bianchi identity, ${\nabla }^{\mu }{R}_{\mu \nu }=\frac{1}{2}{\nabla }_{\nu }R$. Substituting:

$${\nabla }^{\mu }{A}_{\mu \nu }=(\frac{\alpha }{2}+\beta ){\nabla }_{\nu }R.$$

For this to vanish for all metrics, we need $\beta =-\alpha /2$. Setting $\alpha =1$ by convention gives $\beta =-1/2$ and

$$A_{\mu \nu }=R_{\mu \nu }-\frac{1}{2}R{g}_{\mu \nu }+\Lambda g_{\mu \nu }=G_{\mu \nu }+\Lambda g_{\mu \nu }.$$

Step 4: Newtonian limit. For $\Lambda =0$, matching to Poisson's equation in the weak-field limit fixes the proportionality constant to $\kappa =8\pi G/c^4$. A nonzero $\Lambda$ contributes at order $\Lambda r^2$ to the potential and is negligible at solar-system scales for the observed value of $\Lambda$.

Therefore the Einstein field equations $G_{\mu \nu }+\Lambda g_{\mu \nu }=(8\pi G/c^4)T_{\mu \nu }$ are the unique tensor equation satisfying (a), (b), and (c). $■$

Remark. Dropping condition (c) permits higher-derivative theories (e.g., $f(R)$ gravity, Lovelock gravity in dimensions greater than four). In four dimensions, Lovelock's theorem states that the Einstein tensor is the unique symmetric, divergence-free $(0,2)$-tensor that is linear in second derivatives and involves at most the metric and its first two derivatives. This is a stronger version of the uniqueness result.

Bridge. The uniqueness theorem builds toward 13.05.01 pending, where the Schwarzschild solution is the simplest non-flat vacuum solution of the equations whose uniqueness is established here, and appears again in 13.08.01, where the Friedmann equations follow from feeding the FLRW ansatz into the Einstein tensor computed for a spatially homogeneous and isotropic metric. The foundational reason the field equations take exactly this form is the contracted Bianchi identity, which enforces ${\nabla }^{\mu }{G}_{\mu \nu }=0$ and thereby guarantees energy-momentum conservation as a consequence of the geometry. This is exactly the bridge between differential geometry and physics: the Einstein tensor is the unique geometric object whose divergence vanishes identically and which is linear in second derivatives of the metric. Putting these together with the Newtonian limit, the coupling constant $8\pi G/c^4$ is fixed with no remaining freedom, and the central insight is that no other symmetric, divergence-free, second-order equation for the metric is possible in four dimensions.

Exercises [Intermediate+]

Lean formalization [Intermediate+]

The Einstein field equations sit atop a chain of formalised prerequisites that Mathlib does not yet provide. The chain is:

1. Pseudo-Riemannian (Lorentzian) metrics. Mathlib has positive-definite Riemannian metrics but not indefinite-signature pseudo-Riemannian metrics as a distinct structure. The signature constraint $(-,+,+,+)$ is essential for GR.

2. Levi-Civita connection. The unique torsion-free metric-compatible connection on a pseudo-Riemannian manifold. Mathlib has affine connections on vector bundles but not the Levi-Civita connection of a metric.

3. Curvature tensors. The Riemann tensor $R^{\rho }{}_{\sigma \mu \nu }$, its contractions (Ricci tensor $R_{\mu \nu }$, scalar curvature $R$), and the Einstein tensor $G_{\mu \nu }=R_{\mu \nu }-\frac{1}{2}R{g}_{\mu \nu }$. None of these exist in Mathlib.

4. Contracted Bianchi identity. The theorem ${\nabla }^{\mu }{G}_{\mu \nu }=0$ is a formal statement about the curvature tensors that would need to be proved from the definitions.

5. Einstein-Hilbert action. The variational principle $\delta \int R\sqrt{-g}{d}^{4}x=\int G_{\mu \nu }\sqrt{-g}\delta g^{\mu \nu }{d}^{4}x$ connects the Lagrangian formulation to the tensor equation.

A natural formalisation target is: given a pseudo-Riemannian manifold $(M,g)$, define the Einstein tensor and prove it is divergence-free. The Einstein field equations themselves would then be a definition (a structure parametrised by $T_{\mu \nu }$ and $\Lambda$) rather than a theorem.

lean_status: none reflects the absence of items 1--4. This unit ships without a lean_module and is reviewer-attested.

The Einstein-Hilbert action and variational principles [Master]

The Einstein field equations can be derived from a variational principle. The Einstein-Hilbert action is

$$S_{\mathrm{EH}}=\frac{c^4}{16\pi G}{\int }_{M}R\sqrt{-g}{d}^{4}x,$$

where $R$ is the scalar curvature and $\sqrt{-g}$ is the square root of the negative of the metric determinant (the natural volume element on a Lorentzian manifold). Varying $S_{\mathrm{EH}}$ with respect to $g^{\mu \nu }$ and discarding boundary terms (or adding the Gibbons-Hawking-York boundary term) yields $G_{\mu \nu }=0$.

Theorem (Einstein-Hilbert variational principle). The Euler-Lagrange equations of the action $S_{\mathrm{EH}}$ with respect to variations $\delta g^{\mu \nu }$ vanishing on $\partial M$ are the vacuum Einstein equations $G_{\mu \nu }=0$.

Proof. The variation of the scalar curvature decomposes as

$$\delta R=R_{\mu \nu }\delta g^{\mu \nu }+g_{\mu \nu }\square (\delta g^{\mu \nu })-{\nabla }_{\mu }{\nabla }_{\nu }(\delta g^{\mu \nu }),$$

where $\square =g^{\alpha \beta }{\nabla }_{\alpha }{\nabla }_{\beta }$. The variation of the volume element is $\delta \sqrt{-g}=-\frac{1}{2}\sqrt{-g}{g}_{\mu \nu }\delta g^{\mu \nu }$. Combining:

$$\delta \int R\sqrt{-g}{d}^{4}x=\int (R_{\mu \nu }-\frac{1}{2}R{g}_{\mu \nu })\delta g^{\mu \nu }\sqrt{-g}{d}^{4}x+\int {\nabla }_{\alpha }(\cdots {)}^{\alpha }\sqrt{-g}{d}^{4}x.$$

The second integral is a total divergence and vanishes for variations with compact support (or is cancelled by the Gibbons-Hawking-York boundary term $S_{\mathrm{GHY}}=\frac{c^4}{8\pi G}{\int }_{\partial M}K{d}^{3}x$, where $K$ is the trace of the extrinsic curvature of $\partial M$). Setting the coefficient of $\delta g^{\mu \nu }$ to zero yields $G_{\mu \nu }=0$. $\square$

Adding matter via $S_m=\int {\mathcal{L}}_m\sqrt{-g}{d}^{4}x$ defines the stress-energy tensor as $T_{\mu \nu }=-\frac{2}{\sqrt{-g}}\frac{\delta S_m}{\delta g^{\mu \nu }}$. The full action $S=S_{\mathrm{EH}}+S_m$ then yields $G_{\mu \nu }=(8\pi G/c^4)T_{\mu \nu }$. The cosmological constant enters as a volume term $S_{\Lambda }=\frac{c^4\Lambda }{8\pi G}\int \sqrt{-g}{d}^{4}x$, whose variation contributes $\Lambda g_{\mu \nu }$ to the left side.

The Palatini formalism treats the connection ${\Gamma }_{\mu \nu }^{\rho }$ and the metric $g_{\mu \nu }$ as independent variables. Varying the action $\int g^{\mu \nu }{R}_{\mu \nu }(\Gamma )\sqrt{-g}{d}^{4}x$ with respect to $\Gamma$ yields the metric-compatibility condition ${\nabla }_{\rho }{g}_{\mu \nu }=0$, identifying $\Gamma$ as the Levi-Civita connection. Varying with respect to $g^{\mu \nu }$ then yields the Einstein equations. The Palatini approach is the starting point for $f(R)$ theories of gravity and for the first-order formulation of GR used in loop quantum gravity, where the connection and the frame field (vierbein) are treated as independent variables on a gauge group.

The Gibbons-Hawking-York boundary term $S_{\mathrm{GHY}}=\frac{c^4}{8\pi G}{\int }_{\partial M}K{d}^{3}x$, where $K$ is the trace of the extrinsic curvature of $\partial M$, is not an additional physical postulate. It is required to make the variational principle well-posed when the induced metric on $\partial M$ (rather than its normal derivative) is held fixed. Without it, the boundary contribution from the total divergence in the Einstein-Hilbert variation does not vanish, and the action is not stationary for solutions of the field equations.

Lovelock gravity and the dimensionality constraint [Master]

Lovelock's 1971 theorem generalises the uniqueness of the Einstein tensor to arbitrary dimension ^{[Lovelock 1971]}. The $k$-th Lovelock tensor is

$$L_{\mu \nu }^{(k)}=\frac{1}{2^k}{\delta }_{\mu \nu {\rho }_1{\sigma }_1\cdots {\rho }_{k-1}{\sigma }_{k-1}}^{{\alpha }_1{\beta }_1\cdots {\alpha }_k{\beta }_k}\prod _{i=1}^k{R}^{{\rho }_i{\sigma }_i}{}_{{\alpha }_i{\beta }_i},$$

where ${\delta }_{\cdots }^{\cdots }$ is the generalised Kronecker delta (the fully antisymmetrised product of Kronecker deltas). For $k=0$, $L_{\mu \nu }^{(0)}=g_{\mu \nu }$. For $k=1$, $L_{\mu \nu }^{(1)}=G_{\mu \nu }$. For $k\ge 2$, the Lovelock tensor vanishes identically in $n<2k+2$ dimensions because the antisymmetrisation over more indices than the dimension has no independent components.

Theorem (Lovelock 1971). In a four-dimensional spacetime, the only tensor $A_{\mu \nu }$ satisfying (i) symmetry, (ii) dependence on $g_{\mu \nu }$ and its first two derivatives only, (iii) linearity in second derivatives, and (iv) ${\nabla }^{\mu }{A}_{\mu \nu }=0$, is $A_{\mu \nu }=\alpha G_{\mu \nu }+\beta g_{\mu \nu }$ for constants $\alpha$, $\beta$.

In four dimensions, only $k=0$ and $k=1$ survive, giving $A_{\mu \nu }=\alpha G_{\mu \nu }+\beta g_{\mu \nu }$. The constant $\beta$ plays the role of the cosmological constant $\Lambda$. In five dimensions, $k=2$ (the Gauss-Bonnet tensor) is nonvanishing and contributes additional terms to the field equations. The Gauss-Bonnet Lagrangian $R^2-4R_{\mu \nu }{R}^{\mu \nu }+R_{\mu \nu \rho \sigma }{R}^{\mu \nu \rho \sigma }$ is a topological invariant in four dimensions (its integral equals the Euler characteristic by the Chern-Gauss-Bonnet theorem, so it contributes only a total divergence to the action), but is dynamical in five and higher dimensions.

The physical consequence is sharp: in four dimensions, no modification of GR that preserves the basic assumptions (symmetric, divergence-free, second-order) is possible without either introducing additional fields (scalar-tensor theories, $f(R)$ gravity) or admitting higher-derivative terms (which typically introduce ghost instabilities via the Ostrogradsky theorem). Lovelock's theorem is the rigidity result behind Einstein's equations.

Energy conditions and the singularity theorems [Master]

The EFE alone does not determine $T_{\mu \nu }$ -- that depends on the matter model. Energy conditions are inequalities constraining $T_{\mu \nu }$ that codify physically reasonable matter:

• Weak energy condition (WEC): $T_{\mu \nu }{t}^{\mu }{t}^{\nu }\ge 0$ for all timelike $t^{\mu }$. Every observer measures non-negative energy density.
• Null energy condition (NEC): $T_{\mu \nu }{k}^{\mu }{k}^{\nu }\ge 0$ for all null $k^{\mu }$. The weakest standard condition; required for the area theorem of black-hole mechanics.
• Strong energy condition (SEC): $(T_{\mu \nu }-\frac{1}{2}T{g}_{\mu \nu })t^{\mu }{t}^{\nu }\ge 0$ for all timelike $t^{\mu }$. Through the EFE, this requires $R_{\mu \nu }{t}^{\mu }{t}^{\nu }\ge 0$ -- gravity is always attractive.
• Dominant energy condition (DEC): $T_{\mu \nu }{t}^{\mu }$ is a non-spacelike, future-directed vector for every future-directed timelike $t^{\mu }$. Energy does not flow faster than light.

The energy conditions enter the singularity theorems of Hawking and Penrose as hypotheses. The prototype is:

Theorem (Penrose 1965). If $(M,g)$ is a globally hyperbolic spacetime containing a non-compact Cauchy surface $\Sigma$, a closed trapped surface $\mathcal{T}$, and the null energy condition holds, then $(M,g)$ contains an incomplete null geodesic.

The proof constructs a contradiction from the assumption that every null geodesic is complete, using the Raychaudhuri equation for null congruences and the existence of conjugate points. The NEC ensures that the expansion of null geodesics emanating from $\mathcal{T}$ decreases (focusing), while the trapped-surface condition provides initial negative expansion. The key geometric input is the Raychaudhuri equation

$$\frac{d\theta }{d\lambda }=-\frac{1}{2}{\theta }^2-{\hat{\sigma }}_{ab}{\hat{\sigma }}^{ab}+{\hat{\omega }}_{ab}{\hat{\omega }}^{ab}-R_{ab}{k}^a{k}^b,$$

where $\theta$ is the expansion, ${\hat{\sigma }}_{ab}$ the shear, and ${\hat{\omega }}_{ab}$ the rotation. For hypersurface-orthogonal congruences the rotation vanishes, and the NEC ensures the last term is non-positive, giving $d\theta /d\lambda \le -{\theta }^2/2$. This forces $\theta \to -\infty$ in finite affine parameter, producing a conjugate point and violating the assumptions of geodesic completeness ^{[Penrose 1965]}.

The observed acceleration of the cosmic expansion (requiring $\ddot{a}>0$ in the Friedmann equations) violates the SEC, indicating that either the cosmological constant or dark energy provides stress-energy that does not satisfy this condition. The NEC is the one energy condition not observationally violated. The interplay between energy conditions, quantum inequalities, and quantum energy violations (Casimir effect) is an active area of research in quantum field theory in curved spacetime.

The initial-value formulation and well-posedness [Master]

The EFE constitute 10 coupled nonlinear PDEs for $g_{\mu \nu }$. The initial-value formulation, due to Lichnerowicz (1944) and Choquet-Bruhat (1952), casts them as a Cauchy problem. On a spacelike hypersurface $\Sigma$ with unit normal $n^{\mu }$, specify the induced metric ${\gamma }_{ij}$ (the "first fundamental form") and the extrinsic curvature $K_{ij}$ (the "second fundamental form", encoding how $\Sigma$ is embedded in spacetime).

The Einstein equations split into two groups:

• Constraint equations (four equations on $\Sigma$, not evolution equations):

  • Hamiltonian constraint: $R^{(3)}+K^2-K_{ij}{K}^{ij}=\frac{16\pi G}{c^4}\rho$
  • Momentum constraints: $D_j(K^{ij}-{\gamma }^{ij}K)=\frac{8\pi G}{c^4}{j}^i$

  Here $R^{(3)}$ is the scalar curvature of $(\Sigma ,{\gamma }_{ij})$, $D$ is the Levi-Civita connection of ${\gamma }_{ij}$, and $\rho =T_{\mu \nu }{n}^{\mu }{n}^{\nu }$, $j^i=-{\gamma }^{ij}{T}_{j\mu }{n}^{\mu }$ are the energy density and momentum density measured by observers with four-velocity $n^{\mu }$.

• Evolution equations (six second-order PDEs determining how ${\gamma }_{ij}$ and $K_{ij}$ evolve off $\Sigma$):

  • ${\partial }_t{\gamma }_{ij}=-2\alpha K_{ij}+D_i{\beta }_j+D_j{\beta }_i$
  • ${\partial }_t{K}_{ij}=\alpha (R_{ij}^{(3)}+K{K}_{ij}-2K_{ik}{K}^k{}_j)-D_i{D}_j\alpha +\alpha (8\pi G{S}_{ij}/c^4-\frac{1}{2}S{\gamma }_{ij}-\frac{1}{2}\rho {\gamma }_{ij})+{\mathcal{L}}_{\beta }{K}_{ij}$

  where $\alpha$ (lapse) and ${\beta }^i$ (shift) encode the coordinate freedom, and $S_{ij}={\gamma }_i{}^{\mu }{\gamma }_j{}^{\nu }{T}_{\mu \nu }$.

Theorem (Choquet-Bruhat 1952; Choquet-Bruhat and Geroch 1969). Given initial data $(\Sigma ,{\gamma }_{ij},K_{ij})$ satisfying the constraint equations with ${\gamma }_{ij}$ and $K_{ij}$ sufficiently smooth, there exists a unique maximal globally hyperbolic development $(M,g)$ satisfying the Einstein equations and containing $(\Sigma ,{\gamma }_{ij},K_{ij})$ as a Cauchy surface. This development is unique up to diffeomorphism.

The remaining four degrees of freedom (lapse and shift) correspond to coordinate choices (gauge freedom). Local existence was established by Choquet-Bruhat in 1952 using energy estimates for reduced wave equations obtained by imposing harmonic coordinates ${\square }_g{x}^{\mu }=0$ ^{[Choquet-Bruhat 1952]}. Global uniqueness of the maximal development was proved by Choquet-Bruhat and Geroch in 1969 using Zorn's lemma and a patched-gluing argument ^{[Choquet-Bruhat Geroch 1969]}. This result is the foundation of numerical relativity: the constraint equations define the admissible initial-data sets, and the evolution equations propagate them.

Linearised gravity and gravitational waves [Master]

In the weak-field regime, write $g_{\mu \nu }={\eta }_{\mu \nu }+h_{\mu \nu }$ with $|h_{\mu \nu }|\ll 1$ and work to first order in $h$. The linearised Ricci tensor is

$$R_{\mu \nu }^{(1)}=\frac{1}{2}({\partial }_{\mu }{\partial }^{\alpha }{h}_{\alpha \nu }+{\partial }_{\nu }{\partial }^{\alpha }{h}_{\alpha \mu }-\square h_{\mu \nu }-{\partial }_{\mu }{\partial }_{\nu }h),$$

where $h={\eta }^{\mu \nu }{h}_{\mu \nu }$ and $\square ={\eta }^{\alpha \beta }{\partial }_{\alpha }{\partial }_{\beta }$. Defining the trace-reversed perturbation ${\bar{h}}_{\mu \nu }=h_{\mu \nu }-\frac{1}{2}h{\eta }_{\mu \nu }$, the linearised EFE become

$$\square {\bar{h}}_{\mu \nu }=-\frac{16\pi G}{c^4}{T}_{\mu \nu }.$$

This is a wave equation for each component of ${\bar{h}}_{\mu \nu }$, sourced by the stress-energy tensor. In vacuum ($T_{\mu \nu }=0$), the solutions are gravitational waves: transverse perturbations of the metric propagating at the speed of light.

Gauge freedom allows the imposition of the transverse-traceless (TT) gauge: ${\partial }^{\mu }{\bar{h}}_{\mu \nu }=0$ and $\bar{h}=0$, with ${\bar{h}}_{0\mu }=0$ in addition. In this gauge the perturbation has only two independent components, corresponding to the two polarisations $h_{+}$ and $h_{\times }$. For a plane wave propagating in the $z$-direction:

$$h_{\mu \nu }^{\mathrm{TT}}=\left(\begin{array}{cccc}0& 0& 0& 0\\ 0& h_{+}& h_{\times }& 0\\ 0& h_{\times }& -h_{+}& 0\\ 0& 0& 0& 0\end{array}\right).$$

The $h_{+}$ polarisation stretches along one axis while compressing along the perpendicular axis; $h_{\times }$ does the same rotated by $45^{\circ }$.

Theorem (Quadrupole formula). For a slowly moving, weakly self-gravitating source with mass quadrupole moment $Q_{ij}(t)$, the radiated gravitational-wave amplitude at distance $r$ is

$$h_{ij}^{\mathrm{TT}}(t,r)=\frac{2G}{c^{4}r}{\ddot{Q}}_{ij}^{\mathrm{TT}}(t-r/c),$$

where the double dot denotes the second time derivative and the TT projection extracts the transverse-traceless part. The total power radiated is $P=\frac{G}{5c^5}⟨{\stackrel{...}{Q}}_{ij}{\stackrel{...}{Q}}^{ij}⟩$, where the angle brackets denote a time average and the triple dot denotes the third time derivative.

The factor $G/c^5\approx 2.6\times 10^{-53}{\text{W}}^{-1}$ makes gravitational radiation extraordinarily weak: even the Hulse-Taylor binary pulsar radiates only $\sim 7\times 10^{24}$ W (about 1% of the Sun's luminosity), despite involving two neutron stars in a tight orbit. The first direct detection by LIGO in 2015 (GW150914) measured a strain $h\sim 10^{-21}$ from the merger of two black holes at a distance of $\sim 400$ Mpc, producing a peak gravitational luminosity of $\sim 3.6\times 10^{49}$ W -- briefly outshining the electromagnetic luminosity of the entire observable universe. The linearised theory builds toward 13.07.01 pending, where gravitational waves are treated in full nonlinear generality.

Full proof set [Master]

Proposition 1 (Contracted Bianchi identity implies divergence-free Einstein tensor). On any pseudo-Riemannian manifold $(M,g)$ with Levi-Civita connection, the Einstein tensor $G_{\mu \nu }=R_{\mu \nu }-\frac{1}{2}R{g}_{\mu \nu }$ satisfies ${\nabla }^{\mu }{G}_{\mu \nu }=0$ identically.

Proof. The second Bianchi identity is

$${\nabla }_{\lambda }{R}_{\rho \sigma \mu \nu }+{\nabla }_{\rho }{R}_{\sigma \lambda \mu \nu }+{\nabla }_{\sigma }{R}_{\lambda \rho \mu \nu }=0.$$

Contract on indices $\lambda$ and $\mu$ using $g^{\lambda \mu }$:

$${\nabla }^{\mu }{R}_{\rho \sigma \mu \nu }+{\nabla }_{\rho }{R}_{\sigma \mu }{}^{\mu }{}_{\nu }-{\nabla }_{\sigma }{R}_{\lambda \rho }{}^{\lambda }{}_{\nu }=0.$$

Since $R_{\sigma \mu }{}^{\mu }{}_{\nu }=R_{\sigma \nu }$ and $R_{\lambda \rho }{}^{\lambda }{}_{\nu }=R_{\rho \nu }$:

$${\nabla }^{\mu }{R}_{\rho \sigma \mu \nu }+{\nabla }_{\rho }{R}_{\sigma \nu }-{\nabla }_{\sigma }{R}_{\rho \nu }=0.$$

Contract on $\rho$ and $\nu$ using $g^{\rho \nu }$:

$$g^{\rho \nu }{\nabla }^{\mu }{R}_{\rho \sigma \mu \nu }+{\nabla }^{\nu }{R}_{\sigma \nu }-{\nabla }_{\sigma }R=0.$$

The first term: $g^{\rho \nu }{\nabla }^{\mu }{R}_{\rho \sigma \mu \nu }={\nabla }^{\mu }(g^{\rho \nu }{R}_{\rho \sigma \mu \nu })={\nabla }^{\mu }{R}^{\nu }{}_{\sigma \mu \nu }=-{\nabla }^{\mu }{R}_{\sigma \mu }=-{\nabla }^{\mu }{R}_{\mu \sigma }$, where the last equality uses symmetry $R_{\mu \nu }=R_{\nu \mu }$. The first two terms therefore cancel:

$$-{\nabla }^{\mu }{R}_{\mu \sigma }+{\nabla }^{\nu }{R}_{\sigma \nu }-{\nabla }_{\sigma }R=-{\nabla }_{\sigma }R=0?$$

Wait -- the first two terms are identical (rename dummy index $\mu \to \nu$), so they cancel pairwise:

$$\underset{\text{from first term}}{\underbrace{-{\nabla }^{\mu }{R}_{\mu \sigma }}}+\underset{={\nabla }^{\mu }{R}_{\mu \sigma }}{\underbrace{{\nabla }^{\nu }{R}_{\sigma \nu }}}-{\nabla }_{\sigma }R=-{\nabla }_{\sigma }R.$$

This gives ${\nabla }_{\sigma }R=2{\nabla }^{\mu }{R}_{\mu \sigma }$, i.e., ${\nabla }^{\mu }{R}_{\mu \nu }=\frac{1}{2}{\nabla }_{\nu }R$. Now compute:

$${\nabla }^{\mu }{G}_{\mu \nu }={\nabla }^{\mu }{R}_{\mu \nu }-\frac{1}{2}({\nabla }_{\nu }R)g_{\mu \nu }{g}^{\mu \nu }\cdot \frac{1}{4}\cdot 4.$$

More carefully: ${\nabla }^{\mu }(R{g}_{\mu \nu })=({\nabla }^{\mu }R)g_{\mu \nu }={\nabla }_{\nu }R$. So ${\nabla }^{\mu }{G}_{\mu \nu }={\nabla }^{\mu }{R}_{\mu \nu }-\frac{1}{2}{\nabla }_{\nu }R$. Substituting ${\nabla }^{\mu }{R}_{\mu \nu }=\frac{1}{2}{\nabla }_{\nu }R$:

$${\nabla }^{\mu }{G}_{\mu \nu }=\frac{1}{2}{\nabla }_{\nu }R-\frac{1}{2}{\nabla }_{\nu }R=0.\square$$

Proposition 2 (Newtonian limit of the Einstein equations). For a static, pressureless source with $T_{tt}=\rho c^2$ and all other components negligible, the tt-component of the linearised Einstein equations in the weak-field limit $g_{\mu \nu }={\eta }_{\mu \nu }+h_{\mu \nu }$ yields Poisson's equation ${\nabla }^2\Phi =4\pi G\rho$ where $h_{tt}=-2\Phi /c^2$.

Proof. In the Newtonian gauge (static, diagonal perturbation with $h_{ij}=0$ for spatial indices), the linearised Ricci tensor gives

$$R_{tt}^{(1)}=-\frac{1}{2}{\nabla }^2{h}_{tt},$$

since all time derivatives vanish and the spatial Laplacian ${\nabla }^2={\delta }^{ij}{\partial }_i{\partial }_j$. The trace-reversed field equation for the $tt$ component is, for dust ($p=0$, $T=-\rho c^2$ in the mostly-plus convention):

$$T_{tt}-\frac{1}{2}T{\eta }_{tt}=\rho c^2-\frac{1}{2}(-\rho c^2)(-1)=\frac{1}{2}\rho c^2.$$

The linearised trace-reversed EFE give $R_{\mu \nu }^{(1)}=(8\pi G/c^4)(T_{\mu \nu }-\frac{1}{2}T{\eta }_{\mu \nu })$:

$$-\frac{1}{2}{\nabla }^2{h}_{tt}=\frac{8\pi G}{c^4}\cdot \frac{1}{2}\rho c^2=\frac{4\pi G}{c^2}\rho .$$

Substituting $h_{tt}=-2\Phi /c^2$:

$$-\frac{1}{2}{\nabla }^2(-2\Phi /c^2)=\frac{4\pi G}{c^2}\rho ⟹\frac{1}{c^2}{\nabla }^2\Phi =\frac{4\pi G}{c^2}\rho ⟹{\nabla }^2\Phi =4\pi G\rho .\square$$

Synthesis. The Einstein field equations are the foundational reason that gravity is geometry: the Einstein tensor $G_{\mu \nu }$ is the unique symmetric, divergence-free $(0,2)$-tensor linear in second derivatives of the metric, and the coupling constant $8\pi G/c^4$ is fixed by the Newtonian limit. The central insight of the variational formulation is that the Einstein-Hilbert action identifies the Einstein tensor as the Euler-Lagrange expression of the simplest possible scalar Lagrangian ($R$ itself), and the pattern generalises through the Lovelock tower in higher dimensions. Putting these together with the energy conditions, the singularity theorems of Hawking and Penrose show that geodesic incompleteness is a generic consequence of the Einstein equations with physically reasonable matter -- the bridge is between the local differential geometry of curvature and the global causal structure of spacetime. This is exactly the structure that appears again in 13.05.01 pending (the Schwarzschild solution as the prototype singular spacetime), in 13.07.01 pending (gravitational waves as dynamical curvature propagating on a background), and in 13.08.01 (cosmological solutions where the Friedmann equations govern the expansion of the universe).

Connections [Master]

• Riemann curvature tensor 13.03.01 pending. The Einstein tensor $G_{\mu \nu }=R_{\mu \nu }-\frac{1}{2}R{g}_{\mu \nu }$ is constructed from contractions of the Riemann tensor. The field equations are a statement about curvature, and every specific solution (Schwarzschild, Kerr, FLRW) is found by computing the Riemann tensor for a metric ansatz and imposing $G_{\mu \nu }=(8\pi G/c^4)T_{\mu \nu }$.

• Tensors on manifolds 13.02.01. The field equation $G_{\mu \nu }=(8\pi G/c^4)T_{\mu \nu }$ is a tensor equation between two $(0,2)$-tensor fields. Its coordinate independence follows from the tensoriality of both sides.

• Thermodynamics and the stress-energy tensor 11.01.01 pending. $T_{\mu \nu }$ encodes the density, flux, and pressure of matter and energy. Its form depends on the matter model (perfect fluid, electromagnetic field, scalar field, vacuum energy). The divergence-free condition ${\nabla }^{\mu }{T}_{\mu \nu }=0$ is the general-relativistic statement of energy-momentum conservation.

• Schwarzschild solution 13.05.01 pending. The simplest non-flat solution of the vacuum Einstein equations $R_{\mu \nu }=0$. The Schwarzschild metric describes spacetime outside any static, spherically symmetric mass distribution.

• Cosmology 13.08.01. Feeding the FLRW metric ansatz (spatially homogeneous and isotropic) into the Einstein equations yields the Friedmann equations governing the scale factor $a(t)$ of the universe.

• Newton's laws 09.01.02 pending. Newton's law of gravitation $F=GMm/r^2$ and Poisson's equation ${\nabla }^2\Phi =4\pi G\rho$ are the non-relativistic limits of the Einstein field equations. The EFE reduce to Poisson's equation in the weak-field, slow-motion regime.

• Special relativity 10.05.01 pending. Minkowski spacetime (${\eta }_{\mu \nu }$) is a vacuum solution of the Einstein equations with $\Lambda =0$. SR is the local physics in each tangent space of a GR spacetime.

• Gravitational waves 13.07.01 pending. The linearised Einstein equations yield a wave equation for metric perturbations, predicting gravitational radiation. The full nonlinear theory is treated in the gravitational-waves unit.

Historical & philosophical context [Master]

Einstein presented the field equations to the Prussian Academy of Sciences in Berlin on 25 November 1915, in the paper "Die Feldgleichungen der Gravitation" ^{[Einstein 1915]}. He had been pursuing a theory of gravitation based on the curvature of spacetime since 1907, when the equivalence principle first suggested that gravity is a feature of geometry. The path from the equivalence principle to the field equations took eight years.

The period from 1912 to 1915 was one of intense struggle. Einstein and Grossmann published the "Entwurf" theory in 1913, which had the correct geometric picture (gravity = curvature of a four-dimensional manifold) but the wrong field equations. Einstein spent two years trying to fix the Entwurf equations before realising, in the autumn of 1915, that the correct equations were obtained from the Ricci tensor and the Ricci scalar. He published several versions in November 1915 before arriving at the final form on 25 November.

David Hilbert, working independently in Gottingen, derived the field equations from a variational principle (the Einstein-Hilbert action) and submitted his paper five days before Einstein's presentation ^{[Hilbert 1915]}. The priority question has been debated extensively; the consensus, established by Corry, Renn, and Stachel (1997), is that Hilbert's first submission contained a different (incorrect) form of the equations and was revised after seeing Einstein's final paper. Einstein and Hilbert corresponded during November 1915 and each acknowledged the other's contribution. There is no real priority dispute: Einstein had the physical insight and the correct equations first; Hilbert had the elegant variational derivation.

The Newtonian limit was the decisive check. On 18 November 1915, Einstein used his (nearly final) field equations to compute the perihelion precession of Mercury and obtained $43$ arcseconds per century -- exactly the observed anomaly. He later recounted that this calculation gave him heart palpitations. The correct prediction of Mercury's orbit was the first empirical triumph of the theory, decades before the solar-eclipse light-bending test and the gravitational-redshift measurement.

The cosmological constant $\Lambda$ was introduced by Einstein in 1917 to allow a static universe ^{[Einstein 1917]}. After Hubble's 1929 discovery of cosmic expansion, Einstein abandoned $\Lambda$, reportedly calling it his "greatest blunder." The discovery of cosmic acceleration in 1998 (Perlmutter, Riess, Schmidt) revived $\Lambda$ as the simplest explanation, and $\Lambda >0$ is now a central feature of the standard cosmological model.

The initial-value formulation was developed by Lichnerowicz (1944) and Choquet-Bruhat (1952), establishing that the Einstein equations constitute a well-posed hyperbolic system ^{[Choquet-Bruhat 1952]}. The global existence theory (Choquet-Bruhat and Geroch 1969) confirms that admissible initial data determine a unique maximal development -- a remarkable result for a system of nonlinear equations with gauge freedom.

Bibliography [Master]

Primary literature:

• Einstein, A., "Die Feldgleichungen der Gravitation," Sitzungsberichte Preuss. Akad. Wiss. (1915), 844-847. [The field equations in their final form.]
• Einstein, A., "Erklarung der Perihelionbewegung der Merkur aus der allgemeinen Relativitatstheorie," Sitzungsberichte Preuss. Akad. Wiss. (1915), 831-839. [Mercury perihelion prediction, the first triumph.]
• Hilbert, D., "Die Grundlagen der Physik. (Erste Mitteilung)," Nachr. Ges. Wiss. Gottingen, Math.-Phys. Kl. (1915), 395-407. [Variational derivation of the field equations.]
• Einstein, A., "Kosmologische Betrachtungen zur allgemeinen Relativitatstheorie," Sitzungsberichte Preuss. Akad. Wiss. (1917), 142-152. [Introduction of the cosmological constant.]
• Lovelock, D., "The Einstein tensor and its generalizations," J. Math. Phys. 12 (1971), 498-501. [Lovelock's theorem.]
• Penrose, R., "Gravitational collapse and space-time singularities," Phys. Rev. Lett. 14 (1965), 57-59. [First singularity theorem.]
• Choquet-Bruhat, Y., "Theoreme d'existence pour certains systemes d'equations aux derivees partielles non lineaires," Acta Math. 88 (1952), 141-225. [Local existence for the Einstein equations.]
• Choquet-Bruhat, Y. & Geroch, R., "Global aspects of the Cauchy problem in general relativity," Comm. Math. Phys. 14 (1969), 329-335. [Global uniqueness of maximal developments.]

Textbooks and monographs:

• Schutz, B. F., A First Course in General Relativity, 2nd ed. (Cambridge, 2009). [Standard intermediate treatment of the field equations.]
• Wald, R. M., General Relativity (University of Chicago Press, 1984). [Master-tier treatment; Ch. 4 on the Einstein equation is definitive.]
• Carroll, S. M., Spacetime and Geometry: An Introduction to General Relativity (Addison-Wesley, 2004). [Accessible treatment of curvature, the Einstein equation, and the Newtonian limit.]
• Hartle, J. B., Gravity: An Introduction to Einstein's General Relativity (Addison-Wesley, 2003). [Physics-first approach; Ch. 8 derives the field equations from physical arguments.]
• Weinberg, S., Gravitation and Cosmology (Wiley, 1972). [Ch. 7 covers the field equations and their Newtonian limit in detail.]
• Misner, C. W., Thorne, K. S. & Wheeler, J. A., Gravitation (Freeman, 1973). [The comprehensive reference; see in particular the variational-principle derivation in Part IV.]

Historical and philosophical:

• Corry, L., Renn, J. & Stachel, J., "Belated Decision in the Hilbert-Einstein Priority Dispute," Science 278 (1997), 1270-1273. [Resolved the Einstein-Hilbert priority question.]
• Norton, J. D., "How Einstein Found His Field Equations, 1912-1915," Historical Studies in the Physical Sciences 14 (1984), 253-316. [The definitive historical account.]
• Pais, A., Subtle is the Lord: The Science and the Life of Albert Einstein (Oxford, 1982). [Biography with detailed treatment of the path to the field equations.]

---

Wave 3 physics unit, deepened 2026-05-21. hooks_out targets 13.05.01 and 11.01.01 are confirmed. Target 13.08.01 is proposed. Status remains draft pending Tyler's review and the GR chapter retro per PHYSICS_PLAN.