Macroscopic Records

A macroscopic record is a stable physical correlation that can be read with an acceptably small error over a useful time. Its reliability depends on amplification, separation of record states, protection against noise, and often redundant storage. Macroscopic size alone guarantees none of these properties.

Required background. Density operators: Calculate probabilities for record projectors and distinguish mixtures from entangled pure states.

Correlation becomes a record only with stability

An ideal measurement interaction correlates alternatives,

$$\sum_a c_a|a\rangle|R\rangle \longrightarrow \sum_a c_a|a\rangle|R_a\rangle.$$

For a readable sharp record, the states $|R_a\rangle$ should be nearly distinguishable. For a durable record, subsequent dynamics should keep their relevant coarse-grained supports distinct. A rapidly decaying excitation may register an event momentarily without serving as long-term memory.

Real records are usually subspaces or ranges of a macroscopic variable, not exact pure states. A voltage interval, magnetization sign, or detector avalanche tolerates microscopic fluctuations while preserving the recorded label.

Amplification and error margins

Suppose a binary record is stored in $N$ independent physical bits, each flipped with probability $\epsilon<1/2$ before reading. For odd $N$, majority-vote failure probability is

$$P_{\rm fail} =\sum_{k=(N+1)/2}^{N} \binom Nk\epsilon^k(1-\epsilon)^{N-k}.$$

With $N=3$ and $\epsilon=0.01$, this is $3\epsilon^2(1-\epsilon)+\epsilon^3=0.000298$. Redundancy improves readout reliability under the independence model.

Correlated failures can defeat that improvement. A common magnetic disturbance flipping all bits is not suppressed by majority voting. Thus record robustness needs a noise model, not merely a large number of copies.

Decoherence and independent readout

Environmental monitoring suppresses interference between alternative record sectors. It can also distribute the label into photons, phonons, or other accessible fragments. Quantum Darwinism asks when those fragments provide redundant, separately readable information.

The same environment can damage records by inducing transitions. Decoherence time and record lifetime are different quantities: a record can lose phase coherence rapidly yet preserve its classical label for a long time. Conversely, strong coupling can both decohere and rapidly erase a fragile label.

Reversibility and thermodynamic conditions

Amplifying a microscopic event often consumes free energy stored in a detector and produces heat. That physical process helps stabilize the record, but it does not imply that every measurement has one universal dissipation cost.

Global unitary descriptions retain correlations that can in principle support reversals. Practical irreversibility arises when controlling those correlations is infeasible and when the initial environment is suitably uncorrelated. Coarse-graining separates those claims.

A record is not yet an outcome ontology

A superposition can contain several individually robust record branches. Explaining their durability does not by itself specify why one is globally actual. Everettian, Bohmian, collapse, and operational accounts interpret the same record structure differently.

A good measurement analysis therefore reports distinguishability, stability, noise, and readout access before making an interpretive claim about actuality.

Exercises

Why can phase decoherence improve the classical predictability of a record while leaving its readout probability unchanged?

Solution

Dephasing in the record basis removes off-diagonal terms but preserves the diagonal probabilities. It suppresses interference between labels without changing their populations. Bit-flip or relaxation processes change those populations and can destroy the record. The two effects must be modeled separately.

References

• W. H. Zurek, “Pointer basis of quantum apparatus: Into what mixture does the wave packet collapse?” Physical Review D 24, 1516–1525 (1981), doi:10.1103/PhysRevD.24.1516.
• W. H. Zurek, “Quantum Darwinism,” Nature Physics 5, 181–188 (2009), author manuscript.