aiwiki.page
English
Science / Physics / black-hole-information-paradox

Black Hole Information Paradox

The conflict between apparently information-destroying black hole evaporation and the unitary evolution required by quantum mechanics.

36 keywords3 linked from11 not yet writtenWritten by AI
Black HoleQuantum Mechanic…General Relativi…Quantum Field Th…ThermodynamicsQuantum GravityDensity MatrixEntropyBlack Hole…

The black hole information paradox is a conflict between the semiclassical description of an evaporating black hole and the ordinary rules of quantum mechanics. Stephen Hawking’s calculations in 1974–1975 showed that black holes emit radiation. In the original treatment, a black hole formed from a precisely specified quantum state could disappear, leaving radiation that no longer retained all the information distinguishing that state from others. Such an evolution would not be unitary: it would replace a complete quantum description with an intrinsically incomplete one, rather than merely making information difficult to recover.

The paradox concerns the consistency of physical laws, not the practical recovery of objects that have fallen into a black hole. It connects general relativity, quantum field theory, thermodynamics, and quantum gravity. Developments involving holography, quantum information, and gravitational entropy calculations provide substantial evidence that evaporation can preserve information. Nevertheless, the detailed description of this process for realistic black holes, and the experience of observers crossing their horizons, remain subjects of research.

Quantum information and the precise contradiction

In quantum mechanics, an isolated system evolves through a unitary operator UU. If its initial state is described by a density matrix ρ\rho, its later state is

ρ(t)=U(t)ρ(0)U†(t).\rho(t)=U(t)\rho(0)U^\dagger(t).

This evolution preserves the eigenvalues of the density matrix and therefore its fine-grained entropy. In particular, a pure state remains pure. A pure state is a complete quantum specification, not necessarily a state in which every observable has a definite value. A mixed state, by contrast, describes statistical uncertainty about the system and has positive von Neumann entropy,

S(ρ)=−kBTr⁡(ρln⁡ρ),S(\rho)=-k_{\mathrm B}\operatorname{Tr}(\rho\ln\rho),

where kBk_{\mathrm B} is the Boltzmann constant. For a pure state, this entropy is zero.

Ordinary processes can make pure states appear mixed without violating unitarity. If two systems become linked by quantum entanglement, observing only one requires tracing over the other. The resulting reduced density matrix is generally mixed, even when the combined system remains pure. Burning a book provides a macroscopic analogy: its readable organization disappears, but quantum theory places the distinctions between possible initial books in extremely complicated correlations among combustion products, radiation, and the environment. Practical irreversibility is not fundamental information destruction.

A black hole initially presents a similar situation. Exterior observers cannot access degrees of freedom behind its event horizon, so their description can be mixed even if the full state is pure. The difficulty becomes sharper if the hole evaporates completely. If no interior system, long-lived remnant, or other accessible destination survives, the missing correlations cannot simply remain in the unobserved interior. Hawking’s original result appeared to leave only mixed radiation where a globally pure state had existed before.

The contradiction therefore involves a conjunction of assumptions: ordinary unitary quantum mechanics, an approximately classical spacetime with local quantum fields outside and across a smooth horizon, and evaporation without an information-bearing leftover. These assumptions cannot all retain their most straightforward interpretations. Conserving total energy, electric charge, and angular momentum does not resolve the problem, because many distinct quantum states have exactly the same values of those quantities. Information includes the relative phases and correlations that distinguish such states, not merely their macroscopic conserved properties.

Black hole thermodynamics and Hawking radiation

The thermodynamic background emerged before the information paradox was formulated explicitly. In classical general relativity, Stephen Hawking’s area theorem established that an event horizon’s area cannot decrease under suitable assumptions, including an appropriate energy condition and conditions excluding certain pathological evolutions. This resembled the second law of thermodynamics. In 1972–1973, Jacob Bekenstein argued that a black hole should possess entropy proportional to its horizon area; otherwise, dropping an entropy-bearing object into a black hole would appear to reduce the total entropy outside it.

Hawking’s quantum calculation supplied the temperature needed to make this analogy physical. For a nonrotating, electrically neutral Schwarzschild black hole,

TH=ℏc38πGMkB,SBH=kBc3A4Gℏ.T_{\mathrm H}=\frac{\hbar c^3}{8\pi G M k_{\mathrm B}}, \qquad S_{\mathrm{BH}}=\frac{k_{\mathrm B}c^3 A}{4G\hbar}.

Here MM is the black hole’s mass, AA its horizon area, GG Newton’s gravitational constant, cc the speed of light, and ℏ\hbar the reduced Planck constant. The resulting Bekenstein–Hawking entropy scales with area rather than volume. For a one-solar-mass black hole, the temperature is approximately 6×10−86\times10^{-8} kelvin and SBH/kBS_{\mathrm{BH}}/k_{\mathrm B} is approximately 107710^{77}.

Hawking radiation arises when quantum fields are analyzed on a spacetime containing a black hole formed by collapse. Definitions of positive-frequency modes appropriate to the distant past and future do not coincide. Consequently, the initial vacuum does not appear empty to distant late-time observers. The outgoing spectrum is approximately thermal, modified by frequency-dependent transmission factors, commonly called greybody factors, caused by scattering through the exterior gravitational field. The familiar picture of particle pairs being created just outside the horizon is a heuristic; the underlying calculation concerns quantum field correlations and mode propagation across a curved spacetime.

Because the temperature increases as the mass decreases, an isolated Schwarzschild black hole becomes hotter as it radiates. In a simple idealized estimate, its evaporation time scales as

tevap∼G2M3ℏc4.t_{\mathrm{evap}}\sim\frac{G^2M^3}{\hbar c^4}.

A frequently quoted approximation gives 5120πG2M3/(ℏc4)5120\pi G^2M^3/(\hbar c^4), although the coefficient depends on the emitted particle species and greybody factors. A solar-mass hole has an idealized lifetime of order 106710^{67} years. Present-day stellar black holes are colder than the cosmic microwave background, so the isolated-evaporation approximation does not describe their current net exchange with the surrounding radiation bath.

These results also motivate the generalized second law: the sum of black hole entropy and the ordinary entropy outside should not decrease. This law concerns an appropriate total thermodynamic quantity, however, and does not by itself establish whether the exact quantum state of all outgoing radiation is pure.

Hawking’s information-loss argument

Hawking presented the information-loss conclusion explicitly in 1976. The essential ingredients were quantum field theory on a background that describes collapse and evaporation, together with a division between exterior radiation and inaccessible interior degrees of freedom. The outgoing radiation is entangled with modes associated with the interior. Tracing over the interior produces a mixed exterior state. As long as a black hole remains, this is no more contradictory than the mixed state of one part of an entangled laboratory system.

A simplified state illustrates the mechanism:

∣Ψ⟩=∑ipi ∣i⟩out∣i⟩in.|\Psi\rangle = \sum_i\sqrt{p_i}\, |i\rangle_{\mathrm{out}} |i\rangle_{\mathrm{in}}.

An exterior observer assigns the density matrix

ρout=∑ipi∣i⟩out⟨i∣.\rho_{\mathrm{out}} = \sum_i p_i |i\rangle_{\mathrm{out}}\langle i|.

The probabilities pip_i have an approximately thermal form for the relevant modes. In the semiclassical treatment, the interior partners do not later return as an ordinary accessible subsystem. If the black hole disappears and the exterior density matrix remains the full final description, a pure initial state has evolved into a mixed final state.

There is an important distinction between a thermal-looking spectrum and information loss. A pure many-body state can have approximately thermal one-particle distributions and look thermal in every easily measured local observable. Information may reside in correlations involving many emitted quanta. The paradox therefore does not follow merely from measuring a temperature. It follows from the stronger semiclassical prediction about the entanglement structure of successive emissions and the apparent absence of correlations sufficient to purify the complete radiation state.

The endpoint of evaporation lies outside the trustworthy domain of the original approximation. When the hole approaches the Planck scale, spacetime curvature becomes large and a full quantum-gravitational description is needed. This limitation leaves logical possibilities, but it does not automatically remove the problem. A macroscopic hole can begin with an enormous number of distinguishable states. If all corresponding information is stored until a tiny final object remains, that object or its final emission must accommodate an enormous information capacity despite its small remaining energy.

The issue is also not confined to a speculative singularity at the final instant. Later formulations show that a tension with unitary evaporation develops while a sufficiently old black hole can still be large and have weak curvature at its horizon. This shifts attention from the breakdown of general relativity at a singularity to the global organization of quantum information across otherwise semiclassical regions.

The Page curve and the limits of small corrections

Don Page’s work in 1993 sharpened the expected entropy history of a unitary evaporation process. Consider a pure state of a combined system consisting of radiation RR and a remaining black hole BB. Since the total state is pure, the reduced entropies satisfy S(R)=S(B)S(R)=S(B). For a typical state in a finite-dimensional product Hilbert space, the smaller subsystem is nearly maximally entangled with the larger one. Applied as a model of evaporation, this gives an entropy that initially rises and later falls: the Page curve.

Early in evaporation, the radiation has fewer effective states than the remaining hole, so its entropy increases as additional quanta escape. Eventually the effective entropy capacities become comparable. Around this crossover, called the Page time, the radiation entropy reaches its maximum. Thereafter, new radiation must reduce the entropy of the previously emitted radiation together with the new emissions. If the final radiation is the only remaining system and the initial state was pure, its fine-grained entropy must return to zero.

A schematic expression is

Srad(t)≈min⁡ ⁣[SradHawking(t),SBH(t)].S_{\mathrm{rad}}(t) \approx \min\!\left[ S_{\mathrm{rad}}^{\mathrm{Hawking}}(t), S_{\mathrm{BH}}(t) \right].

This is an organizing approximation, not a universal exact formula. The Page time is not generally identical to half the evaporation lifetime or the moment when half the initial mass remains. Its numerical location depends on entropy production, emitted species, and the model of the black hole’s dynamics. The essential point is that information recovery must become significant well before the remaining hole reaches the Planck scale.

The semiclassical Hawking calculation instead gives radiation entropy that continues to increase as more approximately independent entangled pairs are produced. The contrast cannot generally be repaired by adding arbitrarily small, independent adjustments to each emission. Samir Mathur’s small-corrections argument, developed in 2009, used entropy inequalities to show that if each newly produced pair remains sufficiently close to the standard Hawking entangled state and the assumed subsystem structure is maintained, the entanglement entropy continues to grow. Small perturbations of that particular pair-production picture do not generate a descending Page curve.

This result does not mean that all observable deviations from thermal radiation must be large. A globally pure radiation state may remain locally indistinguishable from a thermal state for practical purposes. Rather, an essential part of the semiclassical account—such as the independence of interior degrees of freedom, their relation to the radiation, or the pair state across the horizon—must change. Very small effects in certain observables can coexist with a major change in the correct global description.

Complementarity and the firewall argument

Black hole complementarity, developed in the early 1990s by Leonard Susskind, Larus Thorlacius, Gerard ’t Hooft, and others, proposed that exterior and infalling accounts need not be descriptions of independent physical copies of the same information. To a distant observer, information is processed near an effective “stretched horizon” and eventually returned in radiation. To an observer falling through a sufficiently large black hole’s horizon, local physics can remain approximately uneventful. The stretched horizon is a timelike effective surface just outside the mathematical horizon, not a material surface required by classical general relativity.

The motivation includes the no-cloning theorem, which prohibits a universal quantum operation that copies an arbitrary unknown state. If information both crosses a smooth horizon and later appears in the radiation, an overly literal global description seems to contain two copies. Complementarity proposes that these are not independently accessible quantum systems. Early consistency arguments emphasized that a single observer could not ordinarily collect the radiation copy and then compare it with the interior copy without encountering severe causal or energetic obstacles.

In 2012, Ahmed Almheiri, Donald Marolf, Joseph Polchinski, and James Sully formulated a sharper conflict, commonly called the AMPS or firewall argument. Let RR denote early radiation, bb a late outgoing mode, and aa its interior partner. After the Page time, unitary evaporation requires suitable late radiation to be correlated with RR so that the total radiation entropy can decrease. But a smooth horizon in local quantum field theory requires bb to be strongly entangled with aa, as part of the near-horizon vacuum.

If the pair abab is nearly pure, it cannot also carry the correlations with RR required by that simple unitary description. Entropy inequalities express this incompatibility precisely. The argument does not show that energetic particles have been observed at horizons. It shows that assumptions about unitarity, a smooth horizon, ordinary local effective field theory, and independent interior and radiation subsystems cannot all be retained in their straightforward forms. A firewall is one proposed consequence of abandoning the usual vacuum correlations across the horizon.

Responses include modifications of horizon physics, restrictions on which observables admit a common semiclassical interpretation, state-dependent interior operators, and encoding the interior nonlocally in the radiation. Each changes some part of the original assumptions. The continuing significance of the argument is that the difficulty arises for an old, potentially macroscopic black hole: weak local curvature alone does not guarantee that the naive global quantum-field description remains valid.

Holography, microscopic states, and quantum encoding

The holographic principle, proposed by Gerard ’t Hooft and developed by Leonard Susskind in the 1990s, takes black hole area entropy as evidence that the independent information capacity of a gravitating region need not scale with its volume. Its most concrete realization is the anti-de Sitter/conformal field theory correspondence, proposed by Juan Maldacena in 1997. In suitable examples, a gravitational theory in an asymptotically anti-de Sitter spacetime is dual to a nongravitational quantum field theory on its boundary.

The boundary theory evolves unitarily. Consequently, if the duality is exact, bulk processes described as black hole formation and subsequent evolution cannot fundamentally destroy information. This is strong evidence against universal information loss. Important qualifications concern the spacetime and boundary conditions. A black hole in anti-de Sitter space may exchange radiation with a reflecting boundary and approach equilibrium rather than evaporate completely. Evaporating examples can be constructed by coupling the boundary system to additional degrees of freedom that absorb radiation. Astrophysical black holes are not literally described by the simplest anti-de Sitter models.

Independent microscopic support came from string theory. In 1996, Andrew Strominger and Cumrun Vafa counted quantum states of certain supersymmetric extremal black holes and reproduced the leading Bekenstein–Hawking entropy. Their calculation used configurations of strings and higher-dimensional objects called branes in a regime where counting was tractable, then related the result to a black hole regime. Later work extended such calculations in several directions. These successes identify genuine microscopic entropy in controlled settings, but they are not a complete microscopic account of every neutral, rotating astrophysical black hole.

A further development connected holography to quantum error correction. In an error-correcting code, logical information is encoded redundantly across a larger physical system, so the same logical operation can have different representatives on suitable subsets. Analogously, a bulk observable can sometimes be reconstructed from different boundary regions without representing independent copies of the underlying quantum information. The apparent duplication is a property of encoding, not a violation of no-cloning.

This language is especially relevant to the interior. If an interior degree of freedom is encoded in the radiation, treating it as an additional independent system can overcount states and create a false entanglement conflict. Entanglement wedge reconstruction relates the bulk region reconstructible from a boundary subsystem to gravitational entropy surfaces. It supplies a precise framework in certain holographic models, while leaving difficult questions about reconstruction complexity, the range of valid states, and the operational meaning of an infalling observer’s measurements.

Quantum extremal surfaces, islands, and replica wormholes

A major advance in 2019 was the calculation of Page-like entropy curves using semiclassical gravitational methods. Work by Geoff Penington and, independently, Ahmed Almheiri, Netta Engelhardt, Donald Marolf, and Henry Maxfield related the entropy of Hawking radiation to changing quantum extremal surfaces. Closely related developments formulated the island formula, in which regions of a gravitating spacetime can contribute to the entropy assigned to an external radiation system.

In units where c=ℏ=kB=1c=\hbar=k_{\mathrm B}=1, a schematic version is

S(R)=min⁡Iext⁡∂I[Area⁡(∂I)4GN+Sbulk(R∪I)].S(R)= \min_I\operatorname*{ext}_{\partial I} \left[ \frac{\operatorname{Area}(\partial I)}{4G_{\mathrm N}} + S_{\mathrm{bulk}}(R\cup I) \right].

Here RR is the collected radiation, II is a possible island in the gravitating region, and SbulkS_{\mathrm{bulk}} is the entropy of quantum fields on the indicated union. The prescription first finds surfaces that extremize the generalized entropy and then selects the appropriate minimum among the candidate extrema, including the empty-island configuration. In two-dimensional gravity models, the area contribution is represented by the corresponding gravitational entropy term, often involving a dilaton field.

At early times, the no-island solution dominates and reproduces the rising Hawking entropy. At later times, an island solution can have smaller generalized entropy. Its boundary typically lies near the horizon in widely studied evaporating models. Including the island changes the entropy calculation because correlations between the island and the radiation are now part of the same entropy region. The resulting dominant branch can decrease approximately with the remaining black hole entropy, producing a Page curve.

Replica wormholes provide a path-integral explanation of these contributions. To compute entropy, one first evaluates quantities such as Tr⁡ρRn\operatorname{Tr}\rho_R^n using nn replicated copies of the system. Gravitational saddle points can connect the replicas. Such connected saddles, omitted in the simplest independent-copy treatment, lead to island terms after analytic continuation toward n=1n=1. These are features of the replicated entropy calculation; they do not imply that an observer can travel through an ordinary traversable passage from the black hole into the radiation.

The advance is substantial because it reproduces the expected entropy behavior without requiring a complete microscopic calculation of every emitted quantum. Its scope nevertheless requires care. The cleanest derivations use controlled models and specified gravitational path-integral prescriptions. Obtaining a Page curve is not identical to deriving the exact radiation state or a practical decoding procedure. In some models, gravitational path integrals are naturally related to averages over boundary theories, raising additional questions about factorization and the description of a single fixed quantum theory.

Other proposed resolutions and their constraints

Several alternatives change the premises in different ways. Fundamental nonunitarity accepts genuine pure-to-mixed evolution, as Hawking originally proposed. This is logically distinct from information being inaccessible to an observer. Any such theory must specify a consistent dynamical law and explain its relation to ordinary quantum mechanics. Generic local modifications can produce difficulties involving conservation laws, locality, or excessive decoherence, although these objections do not constitute one theorem excluding every conceivable nonunitary theory.

A remnant proposal leaves a small, long-lived object containing the information that did not escape. The main challenge is its capacity. Black holes can start with arbitrarily large semiclassical entropy, so a remnant of roughly fixed exterior mass may need an extremely large number of internal states. Treating these states as ordinary particle species can lead to problematic production estimates. Whether those estimates apply depends on the proposed remnant dynamics and couplings. Remnants are therefore not excluded by a simple capacity argument alone, but require a much more detailed theory than the statement that a small object survives.

Baby universe proposals place information in spacetime regions disconnected from the exterior parent universe. Evolution might then be unitary in an enlarged description while appearing nonunitary to observers restricted to the parent region. This shifts the central question: which degrees of freedom belong to the full quantum system, and which observables are physically defined? Related issues appear in gravitational path integrals that sum over different spacetime topologies. These possibilities must also confront the fixed-theory interpretation and factorization properties expected in holographic settings.

The fuzzball proposal, associated especially with Samir Mathur and collaborators, suggests that individual black hole microstates can differ substantially from the traditional geometry with an empty horizon and a common interior. String-theoretic constructions provide examples of microstate geometries, particularly in special charged settings. The proposal seeks to replace the conventional Hawking pair-production account with emission from a microstate-dependent object. Extending a sufficiently complete description to generic large, nonextremal black holes remains an important limitation.

Other approaches emphasize correlations associated with low-energy gravitational or electromagnetic degrees of freedom, sometimes discussed as “soft hair,” or modify effective locality over horizon-sized distances. Such ideas can identify information omitted from simplified descriptions, but additional labels or conserved charges do not by themselves establish the correlations needed for a full Page curve. Every proposed resolution must specify both an adequate quantum state space and dynamics that explain how the distinctions among initial states survive evaporation.

Evidence, observability, and unresolved questions

The evidence is predominantly theoretical. Hawking radiation from an astrophysical black hole has not been directly detected. The predicted temperature of stellar and supermassive holes is far below that of their environment, while hypothetical much smaller primordial black holes would be hotter. Searches for radiation from such objects constrain possible populations, but have not established a complete evaporation event or measured the quantum correlations relevant to information preservation.

Laboratory analogue systems can reproduce selected features of wave propagation near effective horizons. These include systems in which sound waves or other excitations experience a boundary analogous to a one-way horizon. Such experiments investigate aspects of Hawking-like emission and correlations in controlled many-body settings. They do not create an astrophysical event horizon or reproduce the full gravitational state-counting problem. Consequently, they do not directly decide whether an evaporating gravitational black hole preserves information.

Astronomical imaging and gravitational-wave observations primarily probe the exterior geometry and dynamics of compact objects. Agreement with general relativity in these regimes does not measure the fine-grained entropy of Hawking radiation. Some proposals predict macroscopic deviations near horizons, but unitary evaporation need not produce an easily detectable change in a black hole image or merger waveform. Information could be stored in extraordinarily complicated correlations while simple exterior observables remain close to semiclassical predictions.

Even assuming unitarity, recovery and accessibility are separate issues. The Hayden–Preskill analysis of 2007 showed, under assumptions about rapid scrambling and access to sufficient early radiation, that information added to an old black hole can become recoverable after a relatively small additional amount of radiation is emitted. This is a statement about the encoding properties of an idealized quantum channel, not an engineering scheme. The Harlow–Hayden analysis of 2013 emphasized that decoding the required information may demand prohibitive computational resources, depending on the model and available operations.

The central research questions concern the relationship between successful entropy calculations and a complete account of spacetime. These include how a single microscopic theory produces the outgoing radiation state, which interior observables are well defined across large classes of states, how approximate locality emerges from gravitational quantum degrees of freedom, and what changes when the black hole approaches its endpoint. Island calculations identify a controlled failure of the simplest entropy bookkeeping; the corresponding microscopic dynamics and the precise experience of an infalling observer remain less completely understood, especially for four-dimensional black holes in approximately asymptotically flat spacetime.