What Needs Interpreting?

1.1 A detector is a quantum system

1.1.1 Follow the amplitudes into the apparatus

Imagine that you prepare a qubit in \(|+\rangle=(|0\rangle+|1\rangle)/\sqrt2\). You measure Z, write down a result, and repeat. Half the records are approximately zero and half one. Nothing mysterious has happened to the statistics. The difficulty begins when we ask the quantum theory to describe the device making those records.

Represent its ready state by \(|A_r\rangle\), with distinguishable records \(|A_0\rangle,|A_1\rangle\). An ideal interaction satisfies \(U|0,A_r\rangle=|0,A_0\rangle\) and \(U|1,A_r\rangle=|1,A_1\rangle\). The same linear operator must therefore satisfy

We have derived a correlation, not a randomly selected term. The detector is perfectly reliable within each component: there is no amplitude for a zero paired with the wrong record. Reliability and uniqueness are different questions. The state has exactly the correlations an experiment needs, while its relation to one observed record still calls for an account.

A pointer qubit initially in |0⟩ makes this concrete: a CNOT produces α|00⟩+β|11⟩. This is an ideal premeasurement. A stable laboratory record additionally involves amplification and interaction with an environment. Neither step should be silently replaced with a collapse postulate.

1.2 Three claims in tension

1.2.1 State exactly where the problem enters

Let us make the tension precise rather than calling everything “weird.” Consider three commitments: the wavefunction is a complete physical description; every isolated system, including an apparatus, evolves linearly and unitarily; and a completed measurement has one definite outcome in the ordinary sense of a unique macroscopic record. With the additional usual link between a definite pointer value and the corresponding pointer subspace, the state above does not satisfy all three commitments at once for nonzero α and β.

This is a formulation of the measurement problem, not a theorem declaring quantum mechanics contradictory. One can change the account of completeness, the dynamics, the account of outcomes, or the role assigned to the quantum state. Interpretations differ partly because they reject different parts of this package.

The pointer's density operator gives the correct local frequencies. It does not license the further inference that the joint state has secretly become one product state or the other. A partial trace is an operation on a description of a subsystem. It is not a new dynamical law acting on a single experimental run. This distinction will do much of the work in the next chapter.

1.3 Predictions and physical accounts

1.3.1 Separate the formalism from an explanation

There are several questions hiding inside “What does the wavefunction mean?” Is it a physical field, part of a law, a summary of information, or an agent's probability assignment? Which quantities are definite? What makes a record stable? Why should squared amplitudes govern expectations? Asking these separately makes disagreements easier to locate.

Our common operational language uses a preparation ρ, a measurement with positive effects \(E_k\) summing to I, and \(p(k)=\operatorname{Tr}(\rho E_k)\). To predict later experiments we also need an instrument: completely positive maps \(\mathcal I_k\), with \(\sum_k\mathcal I_k\) trace preserving. The conditional state is \(\mathcal I_k(\rho)/p(k)\) when the outcome has nonzero probability. Effects alone do not specify disturbance.

An interpretation must connect these calculations to a clear account of experiments. A modified theory can also change the predictions; objective-collapse models are a central example. We will not rank approaches by how strange their names sound. We will ask what each posits, what equations it uses, which probabilities it reproduces, and which explanatory task remains open.

Keep an assumption ledger as you read. Write “calculation,” “postulate,” “theorem under assumptions,” or “interpretive proposal” next to each important step. This habit is more valuable than memorizing slogans about observers.

Sources and further reading

• Maximilian Schlosshauer, Decoherence, the measurement problem, and interpretations of quantum mechanics. Review of decoherence and its interpretive scope; compare its distinctions with the calculations here.