Locality and Counterfactual Definiteness in Bell’s Theorem
At a glance
- 1. Separate the ideas
Locality concerns the influence of distant events; counterfactual definiteness concerns outcomes of measurements that were not performed.
- 2. Specify the model
In a Bell experiment, identify the settings, outcomes, and any shared hidden information represented by lambda.
- 3. Apply Bell’s condition
Bell-local causality requires joint outcome probabilities to factor into local response probabilities, given lambda.
- 4. Read the conclusion
A Bell-inequality violation rules out a conjunction of assumptions, not locality or counterfactual definiteness in isolation.
Locality and counterfactual definiteness are often grouped together under the phrase “local realism,” but they are not the same claim. The distinction matters when interpreting Bell’s theorem: experiments that violate a Bell inequality rule out a class of models, rather than showing by themselves that one particular philosophical assumption is false.
What locality means in a Bell experiment
In a Bell experiment, two distant observers—often called Alice and Bob—each choose a measurement setting and record an outcome. Let and denote their settings, and and their outcomes. A hidden variable represents information the model says is relevant to the pair, such as shared information from their common source.
In the usual Bell framework, Bell-local causality means that once is specified, the joint probabilities for the outcomes factor into probabilities for each observer’s local result:
The equation says that Alice’s response, given the relevant past information, does not depend on Bob’s setting or outcome, and vice versa. It is a precise condition used in Bell arguments. “Locality” has other meanings in physics, so it is best to state which one is meant.
With measurement independence—the assumption that is not correlated with the observers’ setting choices—Bell locality is stronger than no signalling. No signalling requires that Alice cannot use her choice of setting to send a controllable message to Bob. Quantum mechanics respects no signalling, but its entangled-state correlations can still violate Bell inequalities. Thus “no faster-than-light communication” and “Bell-local causality” are not interchangeable.
What counterfactual definiteness means
Counterfactual definiteness (CFD) is the idea that outcomes are well-defined even for measurements that were possible but not performed. For example, if Alice measures along direction rather than , CFD in a familiar deterministic hidden-variable picture says there was nevertheless a definite answer that would have been obtained at for the same underlying state.
That idea is distinct from locality. Whether an unperformed measurement has a definite outcome is a claim about the status of possible outcomes. Whether one observer’s result depends on a distant setting is a claim about causal structure or conditional probabilities. You can discuss either claim without automatically asserting the other.
The term CFD is not always used in exactly the same way. Sometimes it means the intuitive claim that all possible measurements have pre-existing values; sometimes it is used as shorthand for deterministic hidden-variable models. Because of that variation, it is clearer to specify the mathematical assumptions of a model rather than rely on the label “realism.”
How the assumptions enter Bell’s theorem
For a deterministic local hidden-variable model, the outcomes can be written as functions such as and . Such a model assigns results for each possible setting, including settings not chosen in a particular run. It therefore has the familiar counterfactual-outcome picture as well as locality.
But Bell inequalities do not require that outcomes be deterministic or that every unperformed measurement be described as having a pre-existing value in that intuitive sense. They also follow for stochastic Bell-local models, in which and can be probabilistic. The factorization condition above is enough for the standard Bell argument, together with assumptions such as measurement independence: the distribution of is not correlated with the observers’ setting choices in the relevant way.
For example, the Clauser–Horne–Shimony–Holt (CHSH) inequality says that a Bell-local model with measurement independence must satisfy . Quantum mechanics predicts values up to for suitable entangled states and measurement settings, and experiments observe violations of the local bound. The result conflicts with the combined assumptions used to derive the inequality.
A concrete way to see the difference
Imagine Alice can choose between settings and , while Bob can choose between and . In an individual run, each chooses only one setting. A deterministic table might assign four possible answers to Alice—one for each of her settings—and four to Bob, all before the choices are made. That table expresses definite counterfactual outcomes. If Alice’s entries depend only on her own setting and , and Bob’s only on his and , it also has the relevant locality structure.
Now imagine instead that the model gives probabilities for results at the chosen settings, without claiming that every unchosen result had a definite value in advance. That rejects the simple deterministic-table picture, but it does not automatically violate Bell locality: the probabilities could still factor as above. Conversely, a theory could assign definite outcomes yet let Alice’s outcome depend on Bob’s distant setting. That would have definite outcomes but fail Bell locality.
These examples show why the two concepts should not be collapsed into “local realism.” The phrase can hide which assumption a particular argument actually uses.
What Bell-test violations do—and do not—show
A Bell violation rules out models satisfying the assumptions behind the violated inequality. In the standard analysis, these include Bell-local causality and measurement independence, along with the experiment’s relevant statistical and detection assumptions. It does not, by itself, prove that counterfactual definiteness is false, nor does it establish that usable faster-than-light signalling occurs.
Different interpretations of quantum mechanics respond to Bell’s result in different ways. Some abandon Bell locality; some question the independence of measurement settings from the variables describing the source; others use different accounts of outcomes or measurement. The data constrain the structure of any adequate account, but they do not select one interpretation simply by naming “realism” as the alternative.
Common mistakes
- Treating locality and CFD as synonyms: locality concerns dependence between distant events; CFD concerns unperformed measurements.
- Equating Bell locality with no signalling: no signalling is weaker and is compatible with Bell-inequality violations.
- Saying Bell’s theorem disproves realism: “realism” is too broad unless the specific mathematical assumptions are stated.
- Assuming Bell inequalities require deterministic pre-existing values: stochastic local models also obey Bell inequalities under the standard assumptions.
- Concluding that entanglement sends messages faster than light: Bell correlations are nonclassical, but they cannot be controlled to transmit information faster than light.
References
Spotted an error? Send a correction. Read how we write and check articles in our editorial policy.