Compare · Black Holes

Black Hole Complementarity vs Final-State Projection

Black Hole Information Paradox· within family

	Black Hole Complementarity Frontier	Final-State Projection Historical
Proposed	1993	2004
Key figures	Leonard Susskind, Larus Thorlacius, John Uglum	Gary Horowitz, Juan Maldacena
In one sentence	Susskind, Thorlacius, and Uglum proposed in 1993 that the apparent contradiction at a black hole horizon is allowed: an outside observer sees information get encoded on the horizon, an infalling observer passes through smoothly, and no single observer ever gets to compare the two pictures. Both descriptions are real; their contradiction is observationally inaccessible.	Horowitz and Maldacena proposed in 2003 that information escapes a black hole not by leaking out gradually but by a postselected boundary condition at the singularity: the quantum state there is required to be maximally entangled with the early Hawking radiation. Under this condition the apparent loss of information from the outside is matched by an exact recovery, and unitarity is preserved by construction. The proposal was historically influential, but Gottesman and Preskill showed within months that postselection enables superluminal signaling unless additional restrictions are imposed. The framework has since been largely superseded by the Island Formula and Replica Wormhole approach.
Predictions	From outside, information about any system that falls into a black hole eventually becomes encoded in the outgoing [[Hawking radiation]]; no detector outside the horizon ever finds that information lost From inside, an infalling observer crosses the horizon smoothly and experiences no high-energy quanta or other dramatic local physics there; the smooth horizon is required by the equivalence principle The two pictures are observationally incompatible: no measurement protocol allows a single observer to verify both the smooth interior and the holographic encoding of the same information The number of degrees of freedom needed to store the information on the horizon scales as the horizon area, not its volume, in units of the Planck area; this is the original holographic-principle claim	The final state at the singularity is a specific maximally-entangled state with the Hawking radiation, fixed by the postselection prescription rather than emerging from dynamical evolution The Hawking radiation, when computed under the postselection, is no longer thermal: it carries quantum correlations with the matter that fell in, encoded by the boundary condition If the framework is valid, any test for causality violations (signals traveling backward in time or faster than light) should return null results; but the original proposal does not say how that is enforced, so an extra signaling restriction has to be bolted on (the Gottesman-Preskill 2003 constraint) Information leaks out in a burst when the final state is reached, not slowly as the Page curve predicts; that would show up as a sudden change in the correlation structure of the radiation at a specific time, a signature distinguishable in principle from the gradual leakage of the Island Formula sibling variants, though measuring the radiation in that detail is far beyond any practical experiment
Where it breaks	AMPS 2012 showed that complementarity, as originally stated, is inconsistent for old black holes that have evaporated more than half their mass: the three assumptions (information preserved, late radiation correlated with early, smooth horizon) cannot all hold simultaneously The 'stretched horizon' is a useful bookkeeping construct but is not derived from first principles; it sits at the Planck scale where the original 1993 analysis is not under controlled approximation Complementarity does not specify how information actually transfers from the infalling matter to the outgoing radiation; the proposal frames the dual-picture consistency requirement without giving a mechanism The exact correspondence between the two observer perspectives requires non-local effects across the horizon that are difficult to formulate explicitly within standard quantum field theory in curved spacetime	Gottesman and Preskill 2003 showed that postselection of the form Horowitz-Maldacena required enables faster-than-light signaling and acausal influence on the past, unless additional restrictions are imposed that the original proposal does not provide The framework treats unitarity as a boundary condition rather than deriving it from a physical mechanism; many physicists find this unsatisfying as an explanation, even if mathematically consistent The 2019 Island Formula and Replica Wormhole results derive the unitary Page curve from the gravitational path integral directly, without requiring any postselection prescription; this is widely viewed as a more physically grounded resolution Postselection requires picking out a specific microstate at the singularity from a vast space; the original prescription does not say which microstate or why it has the specific entanglement structure required to preserve unitarity
Key unresolved problem	The missing-mechanism problem: no one can write down the actual physical process that copies information from matter falling in onto the outgoing Hawking radiation, so the rescue stays an assertion rather than a derivation.	The faster-than-light problem: the kind of postselection this proposal needs, forcing the final state into a chosen outcome, would let signals travel faster than light and so break cause and effect, a fatal flaw the original work left unresolved.
Reader vote	No votes yet	No votes yet

Black Hole Complementarity

1993 · Frontier

Final-State Projection

2004 · Historical

Proposed

1993

2004

Key figures

Leonard Susskind, Larus Thorlacius, John Uglum

Gary Horowitz, Juan Maldacena

In one sentence

Susskind, Thorlacius, and Uglum proposed in 1993 that the apparent contradiction at a black hole horizon is allowed: an outside observer sees information get encoded on the horizon, an infalling observer passes through smoothly, and no single observer ever gets to compare the two pictures. Both descriptions are real; their contradiction is observationally inaccessible.

Horowitz and Maldacena proposed in 2003 that information escapes a black hole not by leaking out gradually but by a postselected boundary condition at the singularity: the quantum state there is required to be maximally entangled with the early Hawking radiation. Under this condition the apparent loss of information from the outside is matched by an exact recovery, and unitarity is preserved by construction. The proposal was historically influential, but Gottesman and Preskill showed within months that postselection enables superluminal signaling unless additional restrictions are imposed. The framework has since been largely superseded by the Island Formula and Replica Wormhole approach.

Predictions

From outside, information about any system that falls into a black hole eventually becomes encoded in the outgoing [[Hawking radiation]]; no detector outside the horizon ever finds that information lost
From inside, an infalling observer crosses the horizon smoothly and experiences no high-energy quanta or other dramatic local physics there; the smooth horizon is required by the equivalence principle
The two pictures are observationally incompatible: no measurement protocol allows a single observer to verify both the smooth interior and the holographic encoding of the same information
The number of degrees of freedom needed to store the information on the horizon scales as the horizon area, not its volume, in units of the Planck area; this is the original holographic-principle claim

The final state at the singularity is a specific maximally-entangled state with the Hawking radiation, fixed by the postselection prescription rather than emerging from dynamical evolution
The Hawking radiation, when computed under the postselection, is no longer thermal: it carries quantum correlations with the matter that fell in, encoded by the boundary condition
If the framework is valid, any test for causality violations (signals traveling backward in time or faster than light) should return null results; but the original proposal does not say how that is enforced, so an extra signaling restriction has to be bolted on (the Gottesman-Preskill 2003 constraint)
Information leaks out in a burst when the final state is reached, not slowly as the Page curve predicts; that would show up as a sudden change in the correlation structure of the radiation at a specific time, a signature distinguishable in principle from the gradual leakage of the Island Formula sibling variants, though measuring the radiation in that detail is far beyond any practical experiment

Where it breaks

AMPS 2012 showed that complementarity, as originally stated, is inconsistent for old black holes that have evaporated more than half their mass: the three assumptions (information preserved, late radiation correlated with early, smooth horizon) cannot all hold simultaneously
The 'stretched horizon' is a useful bookkeeping construct but is not derived from first principles; it sits at the Planck scale where the original 1993 analysis is not under controlled approximation
Complementarity does not specify how information actually transfers from the infalling matter to the outgoing radiation; the proposal frames the dual-picture consistency requirement without giving a mechanism
The exact correspondence between the two observer perspectives requires non-local effects across the horizon that are difficult to formulate explicitly within standard quantum field theory in curved spacetime

Gottesman and Preskill 2003 showed that postselection of the form Horowitz-Maldacena required enables faster-than-light signaling and acausal influence on the past, unless additional restrictions are imposed that the original proposal does not provide
The framework treats unitarity as a boundary condition rather than deriving it from a physical mechanism; many physicists find this unsatisfying as an explanation, even if mathematically consistent
The 2019 Island Formula and Replica Wormhole results derive the unitary Page curve from the gravitational path integral directly, without requiring any postselection prescription; this is widely viewed as a more physically grounded resolution
Postselection requires picking out a specific microstate at the singularity from a vast space; the original prescription does not say which microstate or why it has the specific entanglement structure required to preserve unitarity

Key unresolved problem

The missing-mechanism problem: no one can write down the actual physical process that copies information from matter falling in onto the outgoing Hawking radiation, so the rescue stays an assertion rather than a derivation.

The faster-than-light problem: the kind of postselection this proposal needs, forcing the final state into a chosen outcome, would let signals travel faster than light and so break cause and effect, a fatal flaw the original work left unresolved.

Reader vote

No votes yet