General Relativity as the Foundation of Modern Cosmology

General relativity — the geometric theory of gravitation published by Albert Einstein in 1915 — provides the mathematical backbone for every major framework in modern cosmology, from the expanding universe and the Big Bang theory to gravitational waves and the large-scale structure of the universe. Without its field equations, cosmologists would have no predictive machinery for describing how matter, energy, space, and time interact on scales spanning billions of light-years. This page covers the theory's core mechanics, its causal role in observable cosmological phenomena, the boundaries separating it from competing frameworks, and the genuine tensions that remain unresolved.

• Definition and scope
• Core mechanics or structure
• Causal relationships or drivers
• Classification boundaries
• Tradeoffs and tensions
• Common misconceptions
• Key equations and derivation pathway
• Reference table: GR predictions vs. observational confirmations
• References

Definition and scope

General relativity (GR) replaces Newton's instantaneous gravitational force with a geometric description: mass and energy curve the four-dimensional fabric of spacetime, and objects follow the straightest possible paths — called geodesics — through that curved geometry. The theory is encoded in Einstein's field equations, a set of 10 coupled, nonlinear partial differential equations relating the Einstein tensor $G_{\mu\nu}$ (geometry) to the stress-energy tensor $T_{\mu\nu}$ (matter and energy):

$$G_{\mu\nu} + \Lambda g_{\mu\nu} = \frac{8\pi G}{c^4} T_{\mu\nu}$$

The term $\Lambda$ is the cosmological constant, which Einstein originally introduced in 1917 to permit a static universe and later characterised as his "greatest blunder" — though it has since been reinstated to account for observed accelerated expansion. The scope of GR in cosmology is total: the Friedmann equations, the Lambda-CDM model, gravitational lensing, and the theoretical framework for black holes all derive directly from GR's field equations.

The theory operates at scales where Newtonian gravity either breaks down or produces measurable errors. Mercury's perihelion precesses at 43 arcseconds per century beyond what Newtonian mechanics predicts — a discrepancy precisely accounted for by GR (Einstein, Annalen der Physik, 1915). GR's domain of validity is classical (non-quantum) and applies from stellar interiors out to the observable universe's ~93-billion-light-year diameter.

Core mechanics or structure

Spacetime curvature. The central object in GR is the metric tensor $g_{\mu\nu}$, which encodes distances and time intervals in curved spacetime. The Riemann curvature tensor, derived from the metric, quantifies how geometry deviates from flat Euclidean space. The Einstein tensor $G_{\mu\nu}$ is a contraction of the Riemann tensor that captures the portion of curvature directly sourced by energy and momentum.

Geodesic motion. Freely falling particles — including photons — travel along geodesics. Gravitational attraction in GR is not a force but a consequence of geodesic deviation: two initially parallel geodesics converge in positively curved spacetime. This reframing is what allows GR to predict gravitational lensing, where light bends around massive objects.

The Friedmann–Lemaître–Robertson–Walker (FLRW) metric. Applying GR to a homogeneous, isotropic universe yields the FLRW metric, the foundational solution underpinning all standard Big Bang cosmology. Georges Lemaître derived an expanding-universe solution from GR's field equations in 1927, two years before Edwin Hubble's observational confirmation of recession velocities. The FLRW metric introduces the scale factor $a(t)$, whose time evolution is governed by the Friedmann equations and encodes the entire expansion history of the universe.

Gravitational waves. GR predicts that accelerating masses radiate spacetime perturbations propagating at the speed of light. These gravitational waves were directly detected by LIGO on 14 September 2015 (Abbott et al., Physical Review Letters, 116, 061102, 2016), confirming a prediction made 100 years earlier and opening the field documented at LIGO-Virgo cosmology.

Singularity theorems. Roger Penrose and Stephen Hawking proved in the 1960s–1970s, using GR, that singularities — regions of infinite density — are generic outcomes of gravitational collapse and of tracing cosmic expansion backward in time. These theorems establish the formal existence of a Big Bang singularity within GR's framework, while simultaneously marking the boundary of GR's own predictive power.

Causal relationships or drivers

GR's field equations function as a bidirectional constraint: the distribution of matter and energy determines spacetime geometry, and that geometry determines how matter and energy move and evolve. Four causal chains dominate modern cosmology:

1. Expansion dynamics. The stress-energy content of the universe — radiation density, matter density, and the energy density associated with $\Lambda$ — drives the expansion rate through the Friedmann equations. The Hubble constant $H_0$ is the present-day value of this expansion rate, currently measured at approximately 67–73 km/s/Mpc depending on the method used (Planck Collaboration, 2020, A&A 641, A6; Riess et al., ApJ, 2022).

2. Structure formation. Small density perturbations in the early universe grow under gravitational instability — a process described by cosmological perturbation theory built on linearised GR. Overdense regions collapse to form the galaxies and clusters that constitute the cosmic web.

3. Light propagation. Photons follow null geodesics in curved spacetime. Redshift in an expanding universe is a direct consequence of photons traveling through a spacetime whose scale factor $a(t)$ increases over time — not a Doppler shift in the classical sense.

4. Dark energy coupling. The accelerated expansion detected through Type Ia supernovae in 1998 (Riess et al., AJ, 116, 1009; Perlmutter et al., ApJ, 517, 565) is accommodated within GR by a nonzero $\Lambda$ acting as dark energy, which contributes a negative-pressure term to $T_{\mu\nu}$.

Classification boundaries

GR is the classical field theory of gravity. Its boundaries with adjacent frameworks are functionally important:

GR vs. Newtonian gravity. Newtonian gravity is the weak-field, slow-velocity limit of GR. For gravitational potentials $\Phi \ll c^2$ and velocities $v \ll c$, GR reduces to Newton's inverse-square law. The boundary breaks down near compact objects (neutron stars, black holes) and on cosmological scales.

GR vs. special relativity. Special relativity governs flat (Minkowski) spacetime with no gravity. GR extends special relativity by incorporating curved spacetime; locally (in a small enough region), GR spacetime is always approximately flat, recovering special relativity — the equivalence principle.

GR vs. quantum field theory (QFT). GR and QFT are mutually inconsistent at the Planck scale (~$10^{-35}$ meters, ~$10^{19}$ GeV). GR treats spacetime as a smooth manifold; QFT treats fields as quantised. The resolution of this conflict is the central project of quantum cosmology, loop quantum gravity, and string theory cosmology.

GR vs. modified gravity theories. Scalar-tensor theories (e.g., Brans-Dicke), $f(R)$ gravity, and massive gravity modify GR's action to change its predictions at large scales, often as alternatives to dark matter or dark energy. Observational data — particularly from baryon acoustic oscillations and Planck satellite findings — tightly constrain deviations from standard GR.

Tradeoffs and tensions

The Hubble tension. GR-based early-universe modeling (via the CMB power spectrum measured by Planck) yields $H_0 \approx 67.4$ km/s/Mpc (Planck Collaboration, 2020), while late-universe distance-ladder measurements consistently return $H_0 \approx 73$ km/s/Mpc (SH0ES collaboration; Riess et al., 2022). The 4–5 sigma discrepancy suggests either systematic errors or genuine new physics beyond standard GR-based Lambda-CDM. This tension is one of the active problems catalogued across the cosmology research landscape.

Dark sector unknowns. GR predicts dynamics driven by $T_{\mu\nu}$, but approximately 95% of the universe's energy budget — dark matter (~27%) and dark energy (~68%) per Planck 2018 results — has no independently confirmed particle-physics identity. GR accommodates both as components of $T_{\mu\nu}$ or $\Lambda$, but their nature remains unknown.

Singularity problem. GR's own equations predict its breakdown at the Big Bang singularity and inside black holes. The theory is internally consistent in predicting these boundaries but provides no description of physics at or beyond them.

GR and the cosmic microwave background. GR correctly predicts the angular power spectrum of CMB anisotropies only when cosmic inflation is appended. Inflation is not a prediction of GR but an additional theoretical layer required to solve the horizon and flatness problems.

Common misconceptions

Misconception: Space expands by moving through a medium. GR describes expansion as the metric itself stretching — no background medium exists. Galaxies receding beyond the Hubble radius ($c/H_0 \approx 14.4$ billion light-years) are not moving through space faster than light; the space between them is expanding.

Misconception: GR requires a centre of the universe. The FLRW metric applied in GR is homogeneous and isotropic — every point is equivalent. No centre exists in the model. The big bang occurred everywhere simultaneously, not at a spatial location.

Misconception: The cosmological constant is fine-tuned within GR. The observed value of $\Lambda$ (~$10^{-52}$ m$^{-2}$) is 120 orders of magnitude smaller than quantum field theory's vacuum energy prediction. GR accepts any value of $\Lambda$ — the fine-tuning problem is a problem of QFT, not GR.

Misconception: Gravitational lensing proves GR uniquely. Newtonian gravity also predicts light deflection, but at exactly half the magnitude GR predicts. Sir Arthur Eddington's 1919 solar eclipse measurement of 1.75 arcseconds of deflection (matching GR's prediction) was the first empirical discrimination between the two theories.

Misconception: GR is incompatible with gravitational waves. Einstein himself doubted whether gravitational waves were physical, publishing a retraction draft in 1936 that was corrected before publication. GR fully predicts gravitational radiation; the 2016 LIGO detection confirmed the GR calculation to within measurement precision.

Key equations and derivation pathway

The conceptual and mathematical derivation of GR's cosmological applications follows a fixed logical sequence. These are the stages through which the full framework is constructed, as documented in standard references including Carroll's Spacetime and Geometry (Addison-Wesley, 2004) and Misner, Thorne, and Wheeler's Gravitation (W.H. Freeman, 1973):

1. Special relativity — establish Minkowski spacetime with metric $\eta_{\mu\nu}$ and Lorentz invariance.
2. Equivalence principle — local inertial frames are indistinguishable from flat spacetime; gravity is geometry.
3. General covariance — physical laws must hold in all coordinate systems; tensors are the required mathematical objects.
4. Metric tensor — replace $\eta_{\mu\nu}$ with the general $g_{\mu\nu}$; define covariant derivatives via Christoffel symbols.
5. Curvature tensors — construct Riemann tensor $R^\rho{}{\sigma\mu\nu}$, Ricci tensor $R{\mu\nu}$, and Ricci scalar $R$.
6. Einstein field equations — $G_{\mu\nu} = R_{\mu\nu} - \frac{1}{2}Rg_{\mu\nu} = \frac{8\pi G}{c^4}T_{\mu\nu}$, with optional $\Lambda g_{\mu\nu}$.
7. FLRW ansatz — impose homogeneity and isotropy; reduce field equations to Friedmann equations.
8. Friedmann equations — $H^2 = \frac{8\pi G}{3}\rho - \frac{kc^2}{a^2} + \frac{\Lambda c^2}{3}$; govern expansion rate as a function of density and curvature $k$.
9. Equation of state — specify $w = p/\rho$ for each component (radiation: $w=1/3$; matter: $w=0$; $\Lambda$: $w=-1$) to integrate Friedmann equations forward/backward in time.
10. Observational outputs — luminosity distance, angular diameter distance, comoving distance, and lookback time are all derived from the scale factor $a(t)$ produced by steps 7–9.

The full derivation is explored further across the cosmologyauthority.com reference network and in dedicated pages on the Friedmann equations and cosmological perturbation theory.

Reference table: GR predictions vs. observational confirmations

The law belongs to the people. Georgia v. Public.Resource.Org, 590 U.S. (2020)