Born Rule Problem

The Born-rule problem asks why squared amplitudes should have their stated probabilistic significance in a proposed account of quantum reality. The formal Born rule already gives an unambiguous recipe for probabilities. Foundations asks what those probabilities mean, what assumptions determine their form and how they connect to frequencies or rational expectation.

Helpful background. Born Rule states the operational probability law; Measurement Problem identifies record alternatives.

A numerical rule and its interpretation

For mutually exclusive orthogonal record branches with amplitudes $c_i$, the usual weights are

$$p_i=|c_i|^2,\qquad \sum_i p_i=1.$$

A collapse theory may identify them with probabilities of stochastic events; a pilot-wave theory relates them to a configuration distribution; an Everettian account must explain their role when the unitary state retains all branches. Those are different explanatory tasks even when they recover the same measurement statistics.

Normalization is not an interpretation of probability. Nor does it by itself select the square: many positive functions can be normalized for one chosen decomposition. The stronger demand is consistency across preparations, measurements and physically equivalent descriptions.

What would count as an explanation?

A proposed derivation should identify its input objects, consistency requirements and output claim. If it starts with additive probabilities on projections, its task is to restrict the form of those probabilities. If it starts with preferences over games, it must explain how a theorem about preferences relates to physical expectation. If it starts with an ensemble measure, it must justify why that measure is appropriate to a single observed history.

The postulate page gives a concrete refinement test for alternative amplitude weights; attempts to derive the rule compares the major premise sets. Neither normalization nor a successful algebraic consistency test should be silently promoted into an explanation of every role probability plays.

A useful worked diagnosis is a collapse proposal that specifies the two possible final states but no selection rates. It addresses which physical states are allowed after collapse, but it has no prediction for a run ensemble until it adds probabilities. A unitary proposal that reproduces squared norms has the complementary task of explaining how those norms inform an observer whose global state contains several records. The missing explanatory step is different in the two cases.

Frequencies require a sampling model

For $N$ independent identically prepared trials with success probability $p$, the relative frequency $f_N$ has

$$\mathbb E f_N=p,\qquad \operatorname{Var}(f_N)=\frac{p(1-p)}N.$$

This explains concentration of frequencies once a probability law and independence are supplied. It does not derive the law from the existence of a long sequence. With correlations between trials, the variance includes covariance terms and need not fall as $1/N$.

Likewise, counting mathematical branches is not automatically a frequency measure: branch number can change under harmless refinement of environmental records. Any branch-counting prescription must explain which distinctions count and why its predictions are stable under coarse-graining.

How the proposed routes differ

Gleason’s theorem characterizes event probabilities given Hilbert-space structure and consistency assumptions. Decision-theoretic arguments constrain valuations of quantum games under rationality and equivalence principles. Envariance uses entanglement symmetries and assumptions linking them to local probabilities. Typicality characterizes which histories dominate a chosen measure.

An effective evaluation writes the premises in ordinary language, identifies the mathematical conclusion and then asks whether the probability interpretation has been explained or assumed. Agreement with the Born formula is necessary for an empirically adequate standard account; it does not make all purported derivations logically equivalent.

Exercise: correlated repetitions

A source draws one fair classical bit and repeats it $N$ times. Find the mean and variance of the measured fraction of ones. Why is the ordinary independent-trial variance inapplicable?

Solution

The fraction is either zero or one with equal probability, so its mean is $1/2$ and variance $1/4$ for every $N$. All trials are correlated. The example separates the single-trial probability from the joint sampling law required for frequency concentration.

References

• 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.
• Wallace, David, The Emergent Multiverse: Quantum Theory according to the Everett Interpretation, Oxford University Press (2012), doi:10.1093/acprof:oso/9780199546961.001.0001 — probability in a unitary interpretation.
• Zurek, Wojciech H., “Environment-Assisted Invariance, Entanglement, and Probabilities in Quantum Physics,” Physical Review Letters 90, 120404 (2003), doi:10.1103/PhysRevLett.90.120404 — symmetry-based probability argument.