Unitarity, No-Cloning, Complementarity, and Observers

Unitarity, no-cloning, and observer complementarity constrain different objects. Unitarity is a property of a complete closed-system map; a reduced observer sees a quantum channel that need not be invertible. No-cloning forbids a state-independent physical device from copying arbitrary nonorthogonal states. Complementarity proposes that interior and exterior descriptions need not be simultaneously available to one observer. Keeping these statements distinct is essential before asking whether black-hole information has been duplicated.

Helpful background. Semiclassical Gravity supplies evaporation and backreaction scales; Quantum Information supplies channels, dilation, recovery, and no-cloning; Quantum Fields in Curved Spacetime supplies near-horizon modes and infalling states; Causality and Global Structure supplies the horizon-crossing causal geometry.

Maps, copies, and comparisons

Let $X$ be a complete input and $Y$ a complete output. Closed evolution is unitary when

$$\rho_Y=U\rho_XU^\dagger, \qquad U^\dagger U=UU^\dagger=I.$$

An observer retaining only $R\subset Y$ instead sees

$$\mathcal N_{X\to R}(\rho_X) =\operatorname{Tr}_{Y\setminus R}\!\left(U\rho_XU^\dagger\right),$$

a completely positive trace-preserving channel. It can be noisy even though the complete evolution is unitary. The exact distinction between unitary, isometric, and reduced channels is standard Watrous 2018, Chapters 2–3 and becomes the foundation for every recovery claim in this chapter.

A cloning claim is stronger than saying the same classical label appears in two descriptions. A perfect universal cloner would map every unknown state $\lvert\psi\rangle$ and fixed blank to two independent copies. Inner-product preservation makes that impossible for any pair of nonorthogonal inputs. Orthogonal records can be copied, and one logical operator can have multiple representations in a constrained code, so the theorem must be applied to the actual state space and map.

An operational duplication experiment adds another layer: one observer must collect and decode exterior radiation, cross the horizon, receive an interior signal, and compare the records before hitting the singularity. Causal time, signal energy, decoding complexity, and backreaction can prevent that sequence even when both descriptions appear in a semiclassical narrative.

Six steps through the observer problem

The pages proceed from exact channel theory to the disputed meaning of observer-dependent descriptions.

1. Unitarity and Quantum Channels distinguishes a unitary, an isometry, and the CPTP channel obtained after tracing a physical environment.
2. The No-Cloning Theorem proves the overlap contradiction and shows why copying orthogonal basis states is not excluded.
3. Black-Hole Complementarity states which exterior and interior observables are proposed to be complementary and which consistency burdens remain.
4. Duplication Experiments and Causal Bounds derives the exponentially shrinking interior communication window and its signal-energy cost.
5. Old Black Holes and Decoding explains Hayden–Preskill recovery, separating scrambling from the complexity of running a decoder.
6. Observer Dependence without Contradiction asks when restricting an observer’s algebra is a consistent change of description and when it conflicts with linear quantum mechanics.

The order guards against a common mistake: a causal obstruction to comparing records does not repeal no-cloning, and no-cloning does not by itself identify which semiclassical operator assignment is wrong.

Complementarity as a conditional proposal

Black-hole complementarity proposes that an exterior observer can describe information as thermalized and later radiated, while an infalling observer experiences an approximately smooth horizon. The two descriptions are not to be combined into one observer-accessible experiment. Susskind, Thorlacius, and Uglum formulated this together with a stretched-horizon account and duplication bounds Susskind, Thorlacius, and Uglum 1993, §§ 2–4.

This proposal is not just coordinate dependence. It needs a rule for the domains of interior and exterior operators, preservation of ordinary correlators in each domain, and consistency on any overlap. If both copies are represented as independent factors in one global state, an observer’s inability to compare them may not remove an entropy or linearity contradiction. Later firewall arguments sharpen that distinction.

For an old black hole already entangled with early radiation, Hayden and Preskill show in a random-unitary model that a newly deposited message can become recoverable from only slightly more late radiation than the message size after scrambling Hayden and Preskill 2007, § 3, Eqs. (3)–(7). The theorem is about existence and average fidelity of recovery. It does not automatically provide an efficient decoder or authorize a low-backreaction comparison behind the horizon.

Results and open questions as of 2026-08-19

Channel and no-cloning theorems are settled under their hypotheses. It remains unresolved whether complementarity supplies a globally consistent interior for realistic evaporation. Competing positions differ over factorization, operator domains, code subspaces, and operational criteria.

Evidence or framework	Result	Scope and limitation
Channel dilation Watrous 2018, Chapters 2–3	Every finite-dimensional reduced channel has an isometric environment representation	Does not identify the physical gravitational environment or prove its accessibility
No-cloning Wootters and Zurek 1982, pp. 802–803	No fixed linear device clones arbitrary nonorthogonal states	Orthogonal records and redundant logical representations are not excluded
Stretched-horizon complementarity Susskind, Thorlacius, and Uglum 1993, §§ 2–4	Exterior recovery and smooth infall are assigned to complementary descriptions	Global operator consistency and old-hole entanglement require additional conditions
Hayden–Preskill model Hayden and Preskill 2007, § 3, Eqs. (3)–(7)	An old randomizing black hole can mirror a message after scrambling	Random-unitary typicality and formal recovery do not guarantee efficient or low-backreaction decoding

The conservative position treats complementarity as an operational restriction compatible with semiclassical observations but not yet a microscopic construction. A stronger encoding position represents interior information nonlocally in exterior degrees of freedom on a restricted code subspace. A firewall position argues that the old-hole entanglement assumptions prevent smooth state-independent interior operators. Conclusions therefore depend on the Hilbert space and operator algebra being claimed.

The conclusions could change after a microscopic model constructs compatible interior and exterior operator algebras, or after an efficient recovery protocol survives causal and backreaction bounds, or after a theorem proves that the required observer restriction is consistent across arbitrary superpositions and overlapping code subspaces.

Boundaries of the chapter

Generic channel theorems remain with Quantum Information, fixed-background mode construction with Quantum Fields in Curved Spacetime, and evaporation scales with Semiclassical Gravity. This chapter applies those results to observer comparisons; it does not assume a specific string or holographic realization and does not treat operational impossibility as proof of microscopic consistency.

Check the map

An isometry sends $\lvert0\rangle_X$ to $\lvert00\rangle_{RE}$ and $\lvert1\rangle_X$ to $\lvert11\rangle_{RE}$. Is the reduced map from $X$ to $R$ unitary, and has the complete evolution cloned an arbitrary input?

Solution

The reduced map dephases $X$ in the computational basis, so it is CPTP but not unitary. The complete isometry preserves a superposition as an entangled state $\alpha\lvert00\rangle+\beta\lvert11\rangle$, not as $(\alpha\lvert0\rangle+\beta\lvert1\rangle)^{\otimes2}$. It copies the orthogonal basis label but does not clone an arbitrary input state. This one example separates all three concepts.

References

• Hayden, Patrick, and John Preskill. 2007. “Black Holes as Mirrors: Quantum Information in Random Subsystems.” Journal of High Energy Physics 2007 (09): 120. Open text.
• Susskind, Leonard, Lárus Thorlacius, and John Uglum. 1993. “The Stretched Horizon and Black Hole Complementarity.” Physical Review D 48: 3743–3761. Open text.
• Watrous, John. 2018. The Theory of Quantum Information. Cambridge: Cambridge University Press. DOI.
• Wootters, William K., and Wojciech H. Zurek. 1982. “A Single Quantum Cannot Be Cloned.” Nature 299: 802–803. DOI.