What Is a Foundation

A foundational question asks which structures, assumptions and empirical commitments make a theory the theory it is. In quantum mechanics this includes the state space, probability rule, composition of systems, physical meaning of measurement records and possible alternatives. It is not limited to interpretation, and it does not suspend the requirement for testable predictions and clear mathematics.

Three kinds of foundational work

Structural work asks which principles constrain the formalism: for example, whether a probability assignment on quantum events must have the Born form. Empirical work asks what observations exclude: for example, whether data fit a specified local correlation model. Interpretive work asks what the successful formalism describes and which explanatory commitments accompany it.

These tasks interact. A reconstruction theorem can expose the assumptions behind a probability rule. A Bell experiment can constrain a class of proposed explanations. An interpretation can suggest a modified dynamics that creates a new empirical question. The conclusion must retain the type of evidence that supports it.

A worked classification: the qubit

A qubit state can be represented by a density operator $\rho$, and a two-outcome measurement by effects $E,I-E$, giving

$$p(0)=\operatorname{Tr}(\rho E).$$

Computing this probability for specified matrices is an ordinary quantum calculation. Asking why physical states form a Bloch ball rather than another convex set is structural foundations. Asking whether a detector’s statistics agree with the rule is empirical testing, including its calibration model. Asking whether $\rho$ describes an observer-independent property or an agent’s commitments is interpretation.

A single equation can thus support several questions without answering all of them. Confusion begins when a correct calculation is treated as proof of an ontology, or a verbal interpretation is treated as a replacement for an unprovided calculation.

Constraints and alternatives

A useful foundational claim identifies a model class and something that would fail within it. “Quantum theory is nonclassical” is too vague until “classical” means, for example, a simplex state space, Bell factorization, noncontextual valuations or noninvasive temporal measurements. Different meanings lead to different theorems.

An alternative theory also needs more than a list of quantum features it dislikes. It must specify states, dynamics, probabilities and composition sufficiently to calculate a declared experiment. The operational framework provides one way to keep that interface explicit.

Foundations and ordinary practice

Most reliable laboratory calculations can proceed without settling every ontology question. This does not make foundational questions meaningless; it shows that operational success and explanatory completeness are distinct standards. Conversely, unresolved interpretation does not imply that standard outcome predictions are arbitrary or untested.

Use Layers of a Foundations Claim to classify statements, and the volume overview to choose a route through the subject.

Exercise: classify a claim

“Every Bell-local model obeys this bound; our calibrated quantum model predicts a violation; the observed data reject the local null; therefore only one interpretation is possible.” Classify each clause and identify the unsupported step.

Solution

The clauses are a theorem, a model prediction, an experimental inference and an interpretive uniqueness claim. The first three can be supported by their respective assumptions and evidence. They do not identify a unique interpretation among all accounts that reproduce the violating correlations.

References

• Bell, John S., Speakable and Unspeakable in Quantum Mechanics, 2nd edition, Cambridge University Press (2004) — physical precision in foundational reasoning.
• Hardy, Lucien, “Quantum Theory From Five Reasonable Axioms,” arXiv (2001), quant-ph/0101012 — a structural reconstruction program.