Foundations Theorems Overview

Foundational theorems constrain different mathematical objects. Some represent a probability assignment by a density operator; others obstruct a global classical model, a physical transformation, or a proposed law for combining preparations. Their conclusions become clearer when each theorem is attached to its own input and output. This page compares those structures; the chapter contents provides reading routes and links to the detailed treatments.

Four distinct mathematical questions

Representation. Given a normalized additive probability assignment to sharp events, which mathematical state represents it? Gleason’s theorem answers with a unique density operator under its stated Hilbert-space and additivity hypotheses. It constrains an existing assignment. It does not specify an instrument, a Hamiltonian, or an ontological model of individual outcomes.

Global extension. Can context-wise probability data or sharp-value constraints be realized consistently in one classical probability model? Kochen–Specker concerns shared zero-one projector values. Bell concerns a source law and local response factorization. The measurement-cover formulation expresses a related finite marginal-extension problem. These frameworks overlap mathematically but their physical hypotheses are not interchangeable.

Transformation. Can one admissible quantum channel perform the same task for every input in a declared family? No-cloning excludes exact deterministic copying of distinct nonorthogonal pure states. No-broadcasting characterizes finite-dimensional mixed families whose states can be reproduced in both output marginals. The channel class and the joint-versus-marginal target are part of each theorem.

Composition. Can overlapping single-system preparation measures coexist with independent product preparation laws and all relevant quantum predictions? The PBR theorem answers no under its ontological-model assumptions. Its independence condition constrains the composition of preparations, not Bell’s measurement response factorization.

The Born rule plays different logical roles

In Gleason’s theorem, the trace form

$$\mu(P)=\operatorname{Tr}(\rho P)$$

is the conclusion of the representation theorem. In a Bell violation, the Born rule is an input used to calculate the target behavior outside the local set. In no-cloning, the Hilbert-space and channel framework is assumed when comparing possible physical transformations. Treating all three as premise-free derivations of the entire formalism would confuse their logical roles.

A useful check is to ask which objects could still vary after the theorem has been applied. A density operator fixed by its sharp-event probabilities does not fix a measurement instrument. A family of observed Bell probabilities does not uniquely determine an underlying state and devices. A prohibition on one channel task does not forbid every approximate or postselected alternative.

Dimensions, topology, and finiteness remain visible

The projection version of Gleason and the traditional Kochen–Specker theorem have a dimension-two exception. Replacing projectors with a broader effect structure changes the premises rather than erasing the exception. The mathematical Kochen–Specker extension also separates a finite logical certificate from its exact quantum realization.

Bell polytopes use finite setting and outcome sets. Fine’s complete CHSH characterization applies to the binary two-setting scenario; testing CHSH alone is not a universal locality test for larger scenarios. Infinite-dimensional separable states require trace-norm closure, which the mathematical Bell extension handles without assuming a countable product ensemble.

For transformations, an exact impossibility can be strengthened only after choosing an error metric. The cloning extension uses joint-output trace distance for an approximation bound and separately solves a two-state heralded task. The two statements cannot be compared by omitting their success conditions.

Exercise: identify the missing mathematical object

A proposed “derivation of quantum measurement” starts from Gleason’s trace rule and claims that the post-measurement state must be the Lüders state. Identify the missing object and give two different instruments with the same sharp outcome probabilities.

Solution

The missing object is the instrument, a family of completely positive state-update maps. For a finite PVM $(P_i)$, the Lüders branches are $\mathcal I_i(\rho)=P_i\rho P_i$. For arbitrary fixed density operators $\sigma_i$, the measure-and-prepare branches $\mathcal J_i(\rho)=\operatorname{Tr}(\rho P_i)\sigma_i$ have the same branch traces and therefore the same outcome probabilities, but generally different outputs. Summing either family gives a trace-preserving channel. Gleason constrains event probabilities, not the additional physical choice of instrument. See Quantum Instruments.

References

• Bell, John S., Speakable and Unspeakable in Quantum Mechanics, 2nd edition, Cambridge University Press, 2004, doi:10.1017/CBO9780511815676 — model assumptions and foundational implications.
• Gleason, Andrew M., “Measures on the Closed Subspaces of a Hilbert Space,” Journal of Mathematics and Mechanics 6, 885–893 (1957), doi:10.1512/iumj.1957.6.56050 — the projection-probability representation theorem.
• Pusey, Matthew F., Jonathan Barrett, and Terry Rudolph, “On the reality of the quantum state,” Nature Physics 8, 475–478 (2012), doi:10.1038/nphys2309 — the composition-dependent overlap theorem.
• Watrous, John, The Theory of Quantum Information, Cambridge University Press, 2018, doi:10.1017/9781316848142 — channels, instruments, and information-theoretic constraints.