Skip to content

Generalized Contextuality (Spekkens)

This page introduces Robert Spekkens' framework of generalized noncontextuality, which extends the notion of contextuality beyond projective measurements to arbitrary operational scenarios.

Beyond Kochen–Specker

The traditional Kochen–Specker approach to contextuality focuses on projective measurements and the assignment of definite values to rays in Hilbert space. While powerful, this framework has limitations:

  • It applies only to projective measurements (not POVMs or other generalized measurements)
  • It deals with state-independent scenarios
  • It does not naturally accommodate preparations and transformations

Spekkens' generalized noncontextuality framework addresses all these limitations by recasting contextuality in purely operational terms.

The Operational Framework

Operational Theories

An operational theory describes experiments in terms of:

  • Preparations (\(P\)): Procedures that prepare a system in some state
  • Transformations (\(T\)): Operations applied to the system
  • Measurements (\(M\)): Procedures that yield outcomes

The theory specifies probabilities \(p(k|P, T, M)\) for outcome \(k\) given preparation \(P\), transformation \(T\), and measurement \(M\).

Ontological Models

An ontological model posits that:

  1. The system has an underlying ontic state \(\lambda\) from some space \(\Lambda\)
  2. Preparations determine a probability distribution \(\mu(\lambda|P)\) over ontic states
  3. Measurements determine response functions \(\xi(k|\lambda, M)\) giving outcome probabilities
  4. The operational probabilities arise from averaging over \(\lambda\):
\[p(k|P, M) = \int_\Lambda \xi(k|\lambda, M) \, \mu(\lambda|P) \, d\lambda\]

Operational Equivalences

The key insight is that operationally indistinguishable procedures might be realized differently in the laboratory.

Preparation Equivalence

Two preparations \(P_1\) and \(P_2\) are operationally equivalent if they yield the same outcome probabilities for all measurements:

\[p(k|P_1, M) = p(k|P_2, M) \quad \forall k, M\]

Even though \(P_1\) and \(P_2\) may differ in their physical implementation, they are indistinguishable from the operational perspective.

Measurement Equivalence

Two measurements \(M_1\) and \(M_2\) (or specific effects within them) are operationally equivalent if they yield the same statistics for all preparations:

\[p(k|P, M_1) = p(k|P, M_2) \quad \forall k, P\]

Transformation Equivalence

Similarly, transformations \(T_1\) and \(T_2\) are operationally equivalent if they produce indistinguishable input-output behavior.

Generalized Noncontextuality

Spekkens' principle of generalized noncontextuality demands that operational equivalences be reflected in the ontological model:

Generalized Noncontextuality

An ontological model is noncontextual if:

  • Preparation noncontextuality: Operationally equivalent preparations are represented by the same distribution over ontic states.
  • Measurement noncontextuality: Operationally equivalent measurements are represented by the same response functions.
  • Transformation noncontextuality: Operationally equivalent transformations are represented by the same stochastic maps on ontic states.

In symbols:

  • If \(P_1 \sim P_2\) operationally, then \(\mu(\lambda|P_1) = \mu(\lambda|P_2)\)
  • If \(M_1 \sim M_2\) operationally, then \(\xi(k|\lambda, M_1) = \xi(k|\lambda, M_2)\)

Types of Contextuality

The generalized framework identifies distinct types of contextuality:

Preparation Contextuality

A theory exhibits preparation contextuality if operationally equivalent preparations must be represented by different distributions over ontic states.

Example: In quantum mechanics, the completely mixed state \(\rho = I/d\) can be prepared by mixing any orthonormal basis with equal weights. These different decompositions are operationally equivalent but cannot all be represented by the same ontic distribution in a noncontextual model.

Measurement Contextuality

A theory exhibits measurement contextuality if operationally equivalent measurements cannot be represented by identical response functions.

This generalizes the traditional KS notion: different contexts for the same projector are operationally equivalent ways to measure that projector.

Transformation Contextuality

A theory exhibits transformation contextuality if operationally equivalent transformations cannot be represented by the same ontic evolution.

The Leibniz Principle

Spekkens' noncontextuality can be understood as an ontological Leibniz principle:

If two things are operationally indistinguishable, they should be ontologically identical.

This is a principle of parsimony: the hidden-variable model should not contain structure that is invisible at the operational level.

Quantum mechanics violates this principle—it is contextual precisely because operational equivalences hide ontological distinctions.

Quasiprobability Representations

A remarkable connection exists between contextuality and negativity in quasiprobability representations.

Wigner Function

The Wigner function \(W(q, p)\) is a quasiprobability distribution that can represent quantum states in phase space. For many states, \(W\) takes negative values—a signature of non-classicality.

Spekkens' Result

Spekkens showed that:

Negativity-Contextuality Equivalence

A theory admits a noncontextual ontological model if and only if it admits a non-negative quasiprobability representation.

Equivalently:

  • Noncontextual ⟺ Non-negative quasiprobabilities possible
  • Contextual ⟺ Negativity in any quasiprobability representation

This connects an abstract property (contextuality) to a concrete, quantifiable feature (Wigner negativity).

Contextuality Inequalities

The generalized framework enables the derivation of noncontextuality inequalities—constraints on operational statistics that must be satisfied by any noncontextual model.

Structure of Inequalities

A noncontextuality inequality has the form:

\[\sum_i c_i \, p(k_i | P_i, M_i) \leq B_{NC}\]

where \(B_{NC}\) is the noncontextual bound. Quantum mechanics can violate these inequalities, demonstrating contextuality.

Advantages Over KS

  • Apply to any operational scenario (not just projective measurements)
  • Can witness state-dependent contextuality
  • Enable experimental tests with realistic measurements

Experimental Implications

The generalized framework has made contextuality experimentally accessible:

Prepare-and-Measure Scenarios

Simple prepare-and-measure experiments (no entanglement required) can demonstrate contextuality:

  1. Prepare states using operationally equivalent procedures
  2. Perform measurements with operationally equivalent effects
  3. Observe violation of noncontextuality inequalities

Robustness to Noise

Unlike KS proofs (which require exact orthogonality), generalized contextuality inequalities:

  • Are robust to experimental imperfections
  • Have statistical significance testable with finite data
  • Apply to noisy, realistic quantum systems

Connection to KS Contextuality

The generalized framework subsumes traditional KS contextuality:

  • KS contextuality arises from measurement noncontextuality applied to projective measurements
  • The projector \(|v\rangle\langle v|\) in different bases represents operationally equivalent measurements
  • Assigning the same response function to all contexts would require consistent 0-1 assignments—impossible for KS sets

But generalized contextuality is broader: it captures scenarios beyond KS, including preparation contextuality, which has no traditional KS analog.

Applications

Generalized contextuality has found applications in:

Quantum Information

  • Parity-oblivious multiplexing: Quantum advantage linked to contextuality
  • State discrimination: Contextuality advantages for certain tasks
  • Cryptography: Security proofs based on contextuality

Foundations

  • Classifying quantum phenomena: Contextuality as a unifying concept
  • Toy models: Spekkens' toy model as a noncontextual reference
  • Quantum reconstruction: Contextuality as a core feature

Computation

  • Magic states: Preparation contextuality relates to magic state resource theory
  • MBQC: Measurement contextuality underlies computational power

References

  • R. W. Spekkens, "Contextuality for preparations, transformations, and unsharp measurements," Phys. Rev. A 71, 052108 (2005)
  • R. W. Spekkens, "Negativity and contextuality are equivalent notions of nonclassicality," Phys. Rev. Lett. 101, 020401 (2008)
  • M. D. Mazurek, M. F. Pusey, R. Kunjwal, K. J. Resch, and R. W. Spekkens, "An experimental test of noncontextuality without unphysical idealizations," Nat. Commun. 7, 11780 (2016)
  • R. Kunjwal and R. W. Spekkens, "From the Kochen-Specker theorem to noncontextuality inequalities without assuming determinism," Phys. Rev. Lett. 115, 110403 (2015)