Contextuality and No-Go Results

A measurement context is a compatible collection of alternatives that can be implemented together. Contextuality asks whether the representation assigned to one event can be independent of every compatible procedure in which that same event appears.

The traditional Kochen–Specker question is sharp: can all projectors be assigned predetermined $0$ or $1$ values, consistently across overlapping orthogonal decompositions of the identity? In complex Hilbert-space dimension at least three, the answer is no.

Required background. Use Projectors to represent sharp yes-no events and orthogonal resolutions of the identity, and Projection-Valued Measures to interpret each context as a normalized sharp measurement.

Helpful background. Use Gleason’s Theorem to recognize the trace-rule form forced by noncontextual projector probabilities.

Enter the chapter

Read Contextuality first. It defines a context, distinguishes traditional Kochen–Specker noncontextuality from generalized operational noncontextuality, and explains why ordinary disturbance is not the relevant definition.

Then read Kochen–Specker Theorem. It states the dimension-$d\ge3$ valuation obstruction, gives a Gleason-based proof route, and checks a finite Peres–Mermin parity witness.

Use the Kochen–Specker reference card when only the theorem statement, hypotheses, and canonical link are needed.

One event in two contexts

Suppose a rank-one projector $P$ belongs to two sharp measurements,

$$\{P,P_2,\ldots,P_d\}, \qquad \{P,Q_2,\ldots,Q_d\},$$

with each set an orthogonal resolution of the identity. A traditional noncontextual value assignment requires the same number $v(P)$ in both lists. It also requires exactly one projector in each complete list to carry value $1$:

$$v(P)+\sum_{j=2}^{d}v(P_j)=1, \qquad v(P)+\sum_{j=2}^{d}v(Q_j)=1.$$

One context alone is easy to color. The obstruction arises because many contexts overlap, so choices that are locally consistent cannot be extended to one global assignment.

Two meanings of noncontextuality

Framework	Operational equivalence being respected	Typical representation constraint
Traditional Kochen–Specker	The same sharp projector in different compatible PVMs	One context-independent deterministic value $v(P)\in\{0,1\}$
Generalized operational	Indistinguishable preparations, transformations, or measurement events	Operationally equivalent procedures have identical ontological representations

The generalized framework does not identify noncontextuality with outcome determinism. It can treat unsharp measurements, preparations, transformations, and qubit scenarios. Therefore the standard projective qubit exception does not imply that every qubit experiment has a generalized noncontextual model.

Relation to Gleason

Gleason’s theorem classifies noncontextual probability measures on projectors. For dimension at least three, normalized additive assignments take the form

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

Kochen–Specker asks whether such an assignment can be dispersion-free, with every rank-one projector assigned only $0$ or $1$. The trace-rule structure makes that impossible globally. The theorems share a projection-lattice backbone but answer different questions: one classifies probabilities, while the other rules out global sharp values.

Relation to Bell

Bell nonlocality is a contextuality problem with additional spacetime structure, but the canonical assumptions differ.

• Bell considers separated parties and constrains a joint conditional distribution through local factorization and measurement independence.
• Kochen–Specker can be formulated for a single system and constrains values across compatible sharp measurements.
• Bell inequalities are statistical witnesses tested across remote settings.
• The original Kochen–Specker theorem is a logical obstruction to a global value map.

Neither theorem licenses the statement “all hidden-variable theories are impossible.” Bell-nonlocal or contextual models lie outside the excluded classes.

Diagnostic questions

Before calling an experiment contextual, ask:

1. Which laboratory procedures are claimed to be operationally equivalent?
2. Which ontological representation must therefore be identical?
3. Is outcome determinism assumed, derived, or unnecessary?
4. Which compatibility or non-disturbance condition is experimentally checked?
5. Which inequality or logical contradiction separates the noncontextual model from the data?

Without this ledger, apparatus disturbance or uncontrolled drift can be mislabeled as contextuality.

Common mistakes

Treating a context as a vague environment. Here a context is a precisely specified compatible measurement or operational procedure.

Inferring a contradiction from one basis. A single basis always permits one $1$ and the remaining $0$s. Overlap among multiple contexts creates the obstruction.

Applying the standard theorem to $d=2$. The rank-one projectors of a qubit admit traditional colorings. Generalized contextuality is a separate question.

Equating contextuality with disturbance. Contextuality concerns the failure of a representation to respect specified operational equivalences. Ordinary mechanical disturbance does not by itself establish that failure.

Exercises

1. Two PVMs share a projector $P$. What equality does traditional measurement noncontextuality impose?

Solution

It requires the same value $v(P)$ wherever $P$ occurs. The value may not depend on which other orthogonal projectors complete the PVM. The remaining projectors can differ between the contexts, but each complete context must still contain exactly one value $1$.

2. Why does the existence of a traditional qubit coloring not settle generalized contextuality for qubits?

Solution

The coloring concerns deterministic values for rank-one projective measurements. Generalized noncontextuality also constrains operationally equivalent preparations, transformations, and unsharp measurement events, without taking outcome determinism as its definition. Those additional equivalences can yield contextuality tests for qubits.

3. Compare the conclusions of Gleason and Kochen–Specker in one sentence each.

Solution

Gleason classifies normalized noncontextual projector probabilities as trace-rule probabilities $\operatorname{Tr}(\rho P)$. Kochen–Specker shows that these probabilities cannot be replaced globally by context-independent deterministic values $0$ and $1$ in dimension at least three.

References

• A. Cabello, S. Severini, and A. Winter, “Graph-theoretic approach to quantum correlations,” Physical Review Letters 112, 040401 (2014).
• S. Kochen and E. P. Specker, “The Problem of Hidden Variables in Quantum Mechanics,” Indiana University Mathematics Journal 17, 59–87 (1968), doi:10.1512/iumj.1968.17.17004.
• N. D. Mermin, “Simple unified form for the major no-hidden-variables theorems,” Physical Review Letters 65, 3373–3376 (1990).
• R. W. Spekkens, “Contextuality for preparations, transformations, and unsharp measurements,” Physical Review A 71, 052108 (2005).