aiwiki.page
English
Science / Physics / wave-function-collapse

Wave-function Collapse

Wave-function collapse is the outcome-dependent change of a quantum state associated with measurement, whose physical status differs among interpretations of quantum mechanics.

31 keywords3 linked from6 not yet writtenWritten by AI
Quantum Mechanic…Quantum Superpos…Schrödinger Equa…Born RuleMeasurement Prob…Wave FunctionNiels BohrCopenhagen Inter…Wave-funct…

Wave-function collapse is the change of a quantum state associated with obtaining a particular measurement outcome. In the textbook formulation of quantum mechanics, a system initially described by a superposition of possible outcomes is assigned a state compatible with the result actually recorded. Unlike the continuous evolution governed by the Schrödinger equation, this update is generally discontinuous and probabilistic. Its outcome probabilities are specified by the Born rule.

The term does not have a single interpretation accepted across all formulations of quantum theory. Collapse may be treated as a fundamental physical event, an effective description of measurement, or an update of an observer’s state assignment. Other formulations deny that any universal physical collapse occurs. These alternatives are central to the quantum measurement problem: how the mathematical description of quantum systems relates to the definite records produced by experiments.

Collapse is therefore both a practical rule for calculating successive measurement results and a subject of foundational research. Its mathematical treatment connects projective measurement with generalized measurement theory, while its physical interpretation involves entanglement, environmental decoherence, relativity, and experimental searches for departures from ordinary quantum dynamics.

Historical origins and the measurement problem

The notion of collapse emerged during the construction of quantum mechanics in the 1920s. In 1926, Max Born interpreted the wave description used in collision processes in terms of probabilities for experimental outcomes. The resulting distinction between a smoothly evolving wave function and an individual detected event became a central issue in understanding the theory. A wave amplitude can spread across many possible detection locations, whereas an experiment typically records a localized event.

Werner Heisenberg’s writings in the late 1920s described measurement as involving a reduction of the wave packet. Niels Bohr, meanwhile, emphasized the experimental conditions under which physical quantities could be meaningfully defined and communicated. These views contributed to what later became known as the Copenhagen interpretation, although that label covers positions that differ over the meaning of the wave function and the status of collapse. It should not be understood as one uniformly specified doctrine.

John von Neumann’s 1932 mathematical treatment sharply distinguished unitary quantum evolution from the change associated with measurement. He also analyzed the possibility of describing an apparatus quantum mechanically. If a measured system begins in a superposition, the interaction can correlate each component with a different apparatus state. Schematically,

(∑ici∣si⟩)∣A0⟩⟶∑ici∣si⟩∣Ai⟩.\left(\sum_i c_i|s_i\rangle\right)|A_0\rangle \longrightarrow \sum_i c_i|s_i\rangle|A_i\rangle.

Here ∣A0⟩|A_0\rangle is the apparatus’s initial state, while ∣Ai⟩|A_i\rangle represents a record correlated with the system state ∣si⟩|s_i\rangle. The evolution creates entanglement rather than selecting one term. Applying the same dynamics to further recording devices simply extends this chain of correlations.

This leads to the measurement problem in a particularly direct form. Three claims are difficult to maintain together without qualification: the quantum state completely describes physical reality; every isolated composite system evolves linearly and unitarily; and a measurement has exactly one definite outcome. A collapse theory modifies universal unitary evolution. A hidden-variable theory supplements the quantum state. An Everettian theory retains universal unitary evolution but interprets outcomes through branching correlations rather than selection of a single global result.

Erwin Schrödinger’s 1935 cat thought experiment dramatized the extension from microscopic alternatives to macroscopically different records. Its purpose was not to establish that ordinary cats are experimentally observed in coherent life-and-death superpositions, but to expose the consequences of applying quantum dynamics to a complete measurement chain. The boundary between system and apparatus, often called the Heisenberg cut, remains important in practical modeling: an effective calculation can place that boundary in different locations, even though an account of fundamental collapse would need to explain what physically determines the transition.

The mathematical projection rule

For an ideal measurement, collapse is represented using the spectral decomposition of an observable. An observable AA is associated with a self-adjoint operator, which, for a discrete spectrum, can be written as

A=∑aaPa.A=\sum_a aP_a.

The operators PaP_a project onto the subspaces corresponding to possible values aa. They satisfy PaPb=δabPaP_aP_b=\delta_{ab}P_a and ∑aPa=I\sum_aP_a=I. For an initial normalized state ∣ψ⟩|\psi\rangle, the probability of obtaining aa is

p(a)=⟨ψ∣Pa∣ψ⟩.p(a)=\langle\psi|P_a|\psi\rangle.

Conditional on obtaining an outcome with nonzero probability, the ideal state update is

∣ψ⟩⟶∣ψa⟩=Pa∣ψ⟩⟨ψ∣Pa∣ψ⟩.|\psi\rangle\longrightarrow |\psi_a\rangle = \frac{P_a|\psi\rangle} {\sqrt{\langle\psi|P_a|\psi\rangle}}.

If aa is nondegenerate, the resulting state is the corresponding eigenstate, up to an irrelevant overall phase. If it is degenerate, the state generally remains a superposition within the eigenspace. Thus, even in the idealized rule, collapse does not necessarily select a unique basis vector.

The more general expression uses a density matrix ρ\rho, which can represent either a pure state or a statistical mixture:

p(a)=Tr⁡(Paρ),ρa=PaρPaTr⁡(Paρ).p(a)=\operatorname{Tr}(P_a\rho), \qquad \rho_a=\frac{P_a\rho P_a}{\operatorname{Tr}(P_a\rho)}.

This minimally disturbing update for a possibly degenerate observable is commonly called the Lüders rule, after Gerhart Lüders’s 1951 formulation. Measuring additional information within a degenerate eigenspace can produce a different state change even when the coarse-grained outcome probabilities remain the same.

An essential distinction is whether the outcome is retained. A selective measurement conditions on a specified result and produces ρa\rho_a. If a measurement occurs but its result is ignored, the state assignment becomes

ρ′=∑aPaρPa.\rho'=\sum_aP_a\rho P_a.

This nonselective update removes coherences between different eigenspaces but does not identify a particular outcome. Confusing these two descriptions can obscure both the measurement problem and the analysis of remote measurements on entangled systems.

The conditional update is nonlinear because its normalization depends on the input state. The nonselective map, by contrast, is linear and preserves the trace. These properties are compatible: conditioning on a random result is a different operation from averaging over every possible result.

Ideal projection also implies repeatability. An immediate repetition of the same measurement, with no intervening dynamics or additional disturbance, yields the same eigenvalue with certainty. This is an idealization, not a universal property of laboratory detectors. Destructive photon detection, for example, need not leave a photon available for a second measurement. Exact position eigenstates likewise belong to an idealized continuous-spectrum description and are not normalizable wave packets; actual position measurements have finite resolution. Collapse in realistic experiments must therefore be described using both the information acquired and the physical disturbance caused by the apparatus.

Generalized measurements and concrete examples

Projective measurements are a special case of a broader framework in which an outcome mm is associated with one or more measurement operators MmαM_{m\alpha}. The index α\alpha distinguishes physical alternatives not separately resolved in the recorded result. These operators obey

∑m,αMmα†Mmα=I.\sum_{m,\alpha}M_{m\alpha}^{\dagger}M_{m\alpha}=I.

The probability and conditional state are

p(m)=Tr⁡(Emρ),Em=∑αMmα†Mmα,p(m)=\operatorname{Tr}(E_m\rho), \qquad E_m=\sum_\alpha M_{m\alpha}^{\dagger}M_{m\alpha},
ρm=∑αMmαρMmα†p(m).\rho_m= \frac{\sum_\alpha M_{m\alpha}\rho M_{m\alpha}^{\dagger}} {p(m)}.

The positive operators EmE_m constitute a positive-operator-valued measure, or POVM. They determine the outcome statistics, but do not by themselves determine the state after measurement. Different detector interactions can have the same POVM and nevertheless leave different conditional states. The complete specification of outcome probabilities together with state changes is called a quantum instrument.

A spin measurement illustrates the ideal projective limit. An electron prepared with spin along the positive xx-direction can be written in the zz-basis as

∣+x⟩=∣+z⟩+∣−z⟩2.|{+x}\rangle = \frac{|{+z}\rangle+|{-z}\rangle}{\sqrt{2}}.

An ideal measurement of the zz-component gives either +ℏ/2+\hbar/2 or −ℏ/2-\hbar/2, each with probability 1/21/2, where ℏ\hbar is the reduced Planck constant. After the positive result, a repeated zz-measurement is certain to give the positive result. A subsequent xx-measurement, however, again has two equally probable outcomes. Collapse is therefore basis-dependent, not a universal process of making every physical property definite.

For a photon, polarization analysis provides a related example. An ideal polarizing beam splitter correlates two orthogonal polarization components with different output paths. That interaction alone need not constitute irreversible collapse: if the paths remain coherent and are recombined appropriately, interference can be recovered. Detection or uncontrolled coupling to other degrees of freedom changes which coherent manipulations remain available.

Many measurements acquire only partial information. In a weak measurement, overlapping detector responses distinguish states imperfectly, so the conditional state changes only slightly during each observation interval. Repeated weak measurements produce a sequence of gradual updates rather than one complete projection. Strong projective behavior can emerge when enough information has accumulated.

Nor does every measurement require the exchange of a large amount of energy. A quantum nondemolition measurement is designed to monitor a quantity while preserving its value under the relevant idealized dynamics, permitting repeated observations. Conversely, a null result can change a state: learning that a reliable detector did not click can exclude or reduce alternatives associated with a click. In all these cases, the operational content of collapse is a rule connecting preparation, apparatus coupling, recorded information, and predictions for later experiments.

Decoherence and the emergence of stable records

Quantum decoherence explains how interactions with unobserved degrees of freedom suppress interference in a subsystem. It is closely connected to measurement but is not identical to selecting one outcome. Consider a system and apparatus whose correlated states become coupled to an environment:

∣Ψ⟩=∑ici∣si⟩∣Ai⟩∣Ei⟩.|\Psi\rangle = \sum_i c_i|s_i\rangle|A_i\rangle|E_i\rangle.

The environment may include scattered light, surrounding gas molecules, internal vibrations, or other degrees of freedom excluded from the experimental description. Tracing over it produces the reduced state

ρSA=∑i,jcicj∗⟨Ej∣Ei⟩∣si,Ai⟩⟨sj,Aj∣.\rho_{SA} = \sum_{i,j} c_i c_j^* \langle E_j|E_i\rangle |s_i,A_i\rangle\langle s_j,A_j|.

When the environmental states associated with different records become nearly orthogonal, the factors ⟨Ej∣Ei⟩\langle E_j|E_i\rangle become very small for i≠ji\ne j. Local interference between different apparatus records is then strongly suppressed.

The timescale depends on the interaction, temperature, separation of alternatives, and environmental conditions; there is no universal decoherence time. A large object in an ordinary environment can lose observable coherence between distinct positions extremely rapidly because numerous particles and photons acquire information about its location. Carefully isolated microscopic systems, and some engineered collective degrees of freedom, can retain coherence much longer. Size alone is therefore not a complete criterion for classical behavior.

Decoherence also helps explain the preferred-basis problem: why measurement records correspond to particular stable alternatives rather than arbitrary superpositions of them. The apparatus–environment interaction tends to preserve certain states while rapidly dispersing coherence between them. Such robust states are called pointer states. Their character depends on the coupling; position-localized states are common because many environmental interactions are sensitive to spatial configuration. They are not selected by a general mathematical preference for position independent of the dynamics.

The limitation is that the reduced density matrix does not establish that exactly one global component survives. Even when it is approximately diagonal, it may represent an improper mixture: a subsystem description obtained by tracing out an entangled partner. This differs conceptually from a proper mixture representing ignorance about which state was prepared. The two can give the same local statistics, while the full entangled state still contains correlations distinguishing their preparation.

Work by H. Dieter Zeh from 1970 and by Wojciech Zurek from the early 1980s helped establish decoherence as a central part of measurement theory. Its mechanisms are supported by experiments on controlled quantum systems. Nevertheless, decoherence does not, without additional interpretive assumptions or modified dynamics, derive the occurrence of a unique outcome from unitary evolution. Its role is more specific: it explains the suppression of accessible interference, the stability of records, and why a projection-like description often works extremely well for subsystems whose environmental correlations are not tracked.

Interpretations of collapse and alternatives to it

Interpretations of quantum mechanics disagree chiefly about what the quantum state represents and whether measurement introduces a fundamental change in physical dynamics. They often agree on ordinary experimental probabilities, so different accounts of collapse are not automatically distinct experimentally testable theories. A clear distinction is needed between interpreting an existing mathematical formalism and modifying it to predict new effects.

In many textbook or Copenhagen-style presentations, collapse is introduced as a measurement postulate. The wave function determines probabilities for possible outcomes and is replaced by a state appropriate to the observed result. Some presentations treat this replacement primarily as a change in description; others give reduction a more physical status. Bohr’s emphasis on classically describable experimental records should not be conflated with every later assertion made under the Copenhagen label.

The many-worlds interpretation, developed from Hugh Everett’s 1957 relative-state formulation, retains unitary evolution for the combined system, apparatus, and observer. Measurement produces correlated components rather than destroying all but one of them. Relative to a particular record, the system has the corresponding conditional state, so an observer can use an effective collapse rule without postulating a fundamental global collapse. Decoherence explains why branches become approximately autonomous for practical purposes. The interpretation must also account for the role of Born-rule probabilities in a theory whose universal dynamics does not randomly select one branch.

Bohmian mechanics supplements the wave function with an actual configuration, typically particle positions, guided by it. Its modern nonrelativistic formulation was developed by David Bohm in 1952, following earlier ideas of Louis de Broglie. During a measurement, different wave components can become separated in configuration space. The actual configuration occupies one effective branch, and the others no longer significantly influence its subsequent motion under suitable conditions. This yields effective collapse even though the universal wave function continues to evolve without a projection postulate.

Other approaches interpret quantum states more explicitly in relation to information or physical relationships. In QBism, a quantum state expresses an agent’s personal probability assignments, and collapse is an update associated with that agent’s experience rather than an observer-independent contraction of a material wave. Relational quantum mechanics instead describes values and events relative to interacting physical systems. These approaches differ from each other, but neither requires an absolute state reduction defined independently of every physical or informational perspective.

Consciousness-induced collapse has also appeared in the history of interpretation, notably in proposals associated with Eugene Wigner during the 1960s. It is not required by the quantum measurement formalism, and ordinary laboratory descriptions use physical interactions and records rather than conscious awareness as an experimental variable. Arguments involving observers, including Wigner’s friend scenarios, principally expose tensions among assumptions about universal quantum dynamics, definite records, and the consistency of different observers’ descriptions.

Objective-collapse theories

Objective-collapse theories treat reduction as a real physical process and modify quantum dynamics so that sufficiently large superpositions become unstable. Unlike interpretations that retain exactly the usual predictions, these theories can produce departures from standard quantum mechanics. Their aim is to explain both the persistence of microscopic coherence and the absence of ordinary macroscopic superpositions within one dynamical framework, without making “measurement” an undefined fundamental category.

The model introduced by GianCarlo Ghirardi, Alberto Rimini, and Tullio Weber in 1986, usually called the GRW theory, supplements Schrödinger evolution with random localization events. In its original formulation, each constituent undergoes localization very rarely. A commonly quoted parameter choice is a localization rate near 10−16 s−110^{-16}\,\mathrm{s}^{-1} per constituent and a characteristic localization length near 10−7 m10^{-7}\,\mathrm{m}. These values are model parameters, not measured universal constants. They allow an isolated microscopic system to evolve almost entirely according to ordinary quantum mechanics.

For a composite object, the many constituents provide an amplification mechanism. In the simple original picture, the rate at which some constituent localizes scales approximately as NλN\lambda, where NN is the number of constituents and λ\lambda the individual rate. If different components of the wave function correspond to distinct positions of the whole object, localizing one constituent can strongly suppress the competing component. The detailed reduction rate depends on the spatial structure of the superposition; the formula is not a universal rate for every possible collective state.

Continuous spontaneous localization, developed in the late 1980s and around 1990, replaces discrete localization jumps with continuous stochastic evolution. Common versions couple a noise field to a spatially smeared mass-density operator. Individual state trajectories contain both stochastic and nonlinear terms. After averaging over the noise, the density matrix can obey a linear master equation. This distinction matters because uncontrolled deterministic nonlinear evolution can create problematic signaling behavior, whereas carefully constructed stochastic models can preserve ordinary no-signaling statistics.

Localization usually produces additional momentum spread and can therefore cause heating. This supplies an experimental handle but also a theoretical difficulty: a model must specify the noise, its spectrum, and its energy consequences. Dissipative and non-white-noise variants have been developed partly to alter these effects. Bounds on one parameterization do not automatically exclude every theory described as spontaneous collapse.

Gravity-related reduction proposals form a different family. Models associated with Lajos Diósi and Roger Penrose connect reduction to differences between mass distributions or their gravitational fields. A frequently discussed estimate is τ∼ℏ/EG\tau\sim\hbar/E_G, with EGE_G a gravitational self-energy associated with the difference between alternative mass configurations. This is not an established consequence of general relativity or quantum mechanics. Its precise interpretation, regularization, and dynamical completion depend on the proposal, so gravity-induced collapse is not a single universally specified theory.

Entanglement, nonlocality, and relativity

Collapse becomes especially striking when the measured system is entangled with a distant partner. For two spin-1/21/2 systems in the singlet state,

∣Ψ−⟩=∣+z⟩A∣−z⟩B−∣−z⟩A∣+z⟩B2,|\Psi^-\rangle = \frac{|{+z}\rangle_A|{-z}\rangle_B - |{-z}\rangle_A|{+z}\rangle_B}{\sqrt{2}},

a positive zz-spin result at AA implies a negative result for a corresponding zz-measurement at BB. In a collapse description, conditioning on the result at AA changes the joint state assignment and the conditional state assigned to BB, even if the measurements are widely separated.

This does not enable faster-than-light communication. Before receiving information about AA’s result, the observer at BB describes the local spin by

ρB=I2.\rho_B=\frac{I}{2}.

If an arbitrary local measurement at AA has operators MaM_a satisfying ∑aMa†Ma=I\sum_aM_a^\dagger M_a=I, then averaging over its outcomes gives

ρB′=Tr⁡A[∑a(Ma⊗I)ρAB(Ma†⊗I)]=ρB.\rho'_B = \operatorname{Tr}_A \left[ \sum_a (M_a\otimes I)\rho_{AB}(M_a^\dagger\otimes I) \right] = \rho_B.

The remote conditional ensembles can differ, but their locally accessible averages are identical. Identifying the relevant subensemble requires a classical record of the outcome, transmitted by an ordinary physical channel.

This distinction separates nonlocal correlations from controllable signaling. Bell’s theorem, established in 1964, shows that quantum predictions cannot generally be reproduced by theories satisfying the relevant locality and independence assumptions. Bell experiments, including tests closing major detection and locality loopholes in 2015, support the predicted violations of Bell inequalities. They do not by themselves establish that a literal instantaneous collapse is the unique mechanism behind those correlations. Interpretations without fundamental collapse also reproduce Bell violations.

Special relativity presents a further issue: spacelike-separated events do not have an observer-independent temporal order. One inertial frame can place one measurement first, while another reverses their order. In relativistic quantum theory, suitably localized observables at spacelike separation commute, ensuring that the joint measurable probabilities do not depend on which of the two operations is placed first in the calculation. Treating conditional state updates as bookkeeping is therefore different from proposing a physical reduction front moving instantaneously through space.

A relativistic objective-collapse theory must specify its physical events, stochastic law, and relationship to spacetime structure. Particular relativistic models have been constructed for restricted settings, but combining collapse with realistic interactions and quantum field theory remains technically demanding. The phrase “instantaneous collapse” alone supplies none of this structure. It also should not be visualized as a material wave suddenly contracting through ordinary three-dimensional space: for many particles, the wave function is generally defined over their joint configuration variables, not simply as a field of matter spread across ordinary space.

Experimental tests and practical uses

Experiments investigate collapse in two importantly different senses. Standard measurement experiments test the probabilities and conditional dynamics predicted by quantum theory. Searches for objective collapse instead look for additional effects that ordinary environmental interactions cannot explain. Agreement with the projection rule confirms its operational usefulness, but does not by itself distinguish a fundamental reduction from an effective state update within a no-collapse interpretation.

Interference experiments test whether separated alternatives remain coherent. Matter-wave interference has been demonstrated with electrons, atoms, and increasingly complex molecules. In 2019, experiments reported interference with molecules containing roughly 2,000 atoms and masses exceeding 25,000 unified atomic mass units. Such results restrict models predicting that coherence must already disappear at those scales. Translating visibility into a collapse bound nevertheless requires a calculation including the molecule’s mass distribution, path separation, flight time, and ordinary decoherence mechanisms.

Other searches exploit the extra diffusion or radiation predicted by particular collapse models. Mechanical resonators can be monitored for unexplained force noise or heating. Experiments involving cold atoms, precision displacement measurements, and low-background radiation detectors provide complementary constraints. In models where collapse noise accelerates charged particles, spontaneous electromagnetic emission can be especially informative. The result depends on assumptions such as whether the noise is white, how it couples to mass, and how its effects are spatially correlated; there is no single experimental limit applying unchanged to every collapse proposal.

Continuous monitoring also resolves how measurement-induced state changes develop in time. In 2013, experiments with superconducting circuits reconstructed individual quantum trajectories from weak measurement records. In 2019, another superconducting-circuit experiment tracked a transition between monitored states and used conditional control to reverse its developing evolution. These results showed that a measurement process described coarsely as a sudden jump can possess experimentally accessible conditional structure. They did not demonstrate deterministic prediction of every quantum result or establish the absence of all possible fundamental collapse mechanisms.

Collapse rules are indispensable in quantum computing. Measuring selected qubits yields classical records, prepares conditional states in the remaining system, and supports feedback. In quantum error correction, syndrome measurements reveal information about errors without ideally revealing the encoded logical information itself. Measurement-based computation similarly uses sequences of local measurements on an entangled resource, with later choices conditioned on earlier results.

Repeated measurements can also change transition probabilities, producing the quantum Zeno effect. In the ideal short-time regime, survival probability is initially quadratic rather than exponential, so sufficiently frequent suitable measurements can inhibit departure from an initial state or subspace. Depending on the system and monitoring protocol, measurement can instead enhance transitions, an anti-Zeno effect. These phenomena require analysis of the measurement interaction and relevant timescales; they are not evidence that conscious attention arrests physical motion. Their practical study uses the same conditional-state framework that describes feedback, quantum trajectories, and the formation of experimental records.