Objective-collapse theory is a family of proposed modifications of quantum mechanics in which the quantum state undergoes spontaneous, physically real localization. These changes occur independently of observers or measurements and supplement ordinary quantum evolution with stochastic dynamics. Collapse is designed to be negligible for microscopic systems in familiar experiments but rapid for macroscopically distinct alternatives. The aim is to explain definite measurement outcomes through a single physical law rather than a separate measurement postulate. Unlike interpretations that retain standard quantum dynamics, objective-collapse models generally predict experimentally testable departures from it.
Motivation and the measurement problem
In standard quantum mechanics, an isolated system’s wave function evolves according to the linear Schrödinger equation. Linearity allows superpositions: if two states are possible, an appropriate linear combination is also a possible state.
When a measuring apparatus interacts with a superposed system, the same dynamics generally produces an entangled state of system and apparatus. Schematically,
where and represent different apparatus readings. The equation itself does not select one reading as the actual outcome. Reconciling this evolution with definite observations is a central aspect of the measurement problem.
Textbook treatments commonly introduce a separate collapse rule, with outcome probabilities given by the Born rule. Objective-collapse theories instead make reduction part of the dynamics of all physical systems. An apparatus is not exceptional because it is observed; its size, composition, and configuration make spontaneous localization effective.
Environmental decoherence explains why interference between macroscopic alternatives becomes practically inaccessible when systems interact with their surroundings. By itself, however, decoherence does not select a single branch of the total quantum state. Objective collapse adds a physical process intended to make that selection.
General structure
Most established collapse models combine three ingredients:
- Ordinary quantum evolution, which remains an excellent approximation in appropriate regimes.
- Stochastic localization, expressed through random jumps or continuous noise.
- Amplification, through which collapse becomes stronger for larger or more massive composite systems.
The wave-function dynamics is generally nonlinear as well as stochastic. Nonlinearity makes it possible to favor one component of a superposition, while randomness supplies variable outcomes with appropriate probabilities.
A crucial distinction separates the evolution of an individual state from that of a statistical ensemble. An individual wave function may obey a nonlinear stochastic equation, while its noise-averaged density matrix obeys a linear master equation. In standard nonrelativistic models, this structure is important for preventing the randomness from being exploited for controllable faster-than-light communication.
A master equation alone does not specify a complete objective-collapse theory. Different stochastic descriptions can yield the same ensemble evolution, and the physical interpretation of individual histories must also be stated.
Principal models and historical development
Ghirardi–Rimini–Weber model
The model introduced by GianCarlo Ghirardi, Alberto Rimini, and Tullio Weber in 1986, usually called GRW, provided an influential quantitative realization of spontaneous localization.
In its original nonrelativistic formulation, each constituent undergoes random localization events at a constant mean rate . Between events, ordinary Schrödinger evolution applies. A localization acting on constituent has the form
with
Here is the position operator, is the localization length, and the random center has probability density
Historically influential illustrative values are
These are proposed parameters, not measured constants. For an isolated constituent, localization would be extremely rare. For a composite system with independently collapsing constituents, localization events occur at a total rate of approximately . When those constituents are appropriately correlated with macroscopically distinct configurations, a single event can suppress one configuration relative to another. This is the GRW amplification mechanism.
Continuous spontaneous localization
Continuous spontaneous localization, or CSL, developed around the end of the 1980s, replaces discrete jumps with continuous stochastic evolution. Its formulation accommodates identical particles without assigning physically distinguishable labels to them.
In widely studied mass-proportional versions, the noise couples to a spatially smeared mass-density operator. This makes the collapse strength depend on the mass distribution. The resulting amplification is not generally captured by a simple rule: it also depends on the object’s geometry, constituent masses, and the separation between superposed configurations relative to .
CSL is a major framework for experimental tests because it gives quantitative predictions for interference suppression, mechanical noise, heating, and radiation. Variants introduce finite noise correlation times, dissipation, or other changes to the basic dynamics.
Gravity-related proposals
Other proposals connect collapse with gravitation. The Diósi–Penrose family associates instability with differences between the mass distributions of superposed alternatives.
A commonly discussed characteristic timescale is
where is the reduced Planck constant and is the gravitational self-energy associated with the difference between the competing mass distributions.
This relation is a heuristic scale rather than a unique, complete dynamical law. Concrete models require further choices, including how mass density is smoothed at short distances. Diósi’s stochastic models and Penrose’s arguments are related but should not be treated as identical theories. Gravity-related collapse is also not an established consequence of general relativity or a completed theory of quantum gravity.
Physical consequences
Suppression of interference
Localization suppresses coherence between sufficiently separated spatial alternatives. For a single free particle in the GRW model, the collapse contribution in the position representation is
For separations much smaller than , the effect is weak. For separations much larger than , the off-diagonal elements decay at a rate approaching .
Experiments therefore depend not just on an object’s size but on the separation of its superposed configurations and the time for which coherence is maintained.
Momentum diffusion and heating
Spatial localization broadens the momentum distribution. In the simplest nondissipative models, repeated localization produces momentum diffusion and increasing mean kinetic energy.
For one free, nonrelativistic particle of mass , using the GRW convention above,
This formula applies to that specific model and should not be transferred unchanged to mass-proportional CSL.
The energy increase reflects the fact that the ordinary particle Hamiltonian is not conserved under the modified dynamics. Dissipative collapse models alter this behavior and can allow relaxation toward a finite effective temperature.
For charged matter, collapse-induced motion can also produce spontaneous electromagnetic radiation. The predicted spectrum depends on the noise model and on how the constituents respond collectively.
Experimental tests
Objective-collapse theories are testable because their dynamics differs from standard quantum mechanics. Experiments constrain particular models and parameter ranges rather than a single universal “collapse strength.”
Matter-wave experiments. Interferometry with atoms, molecules, and larger objects tests whether spatial coherence survives as predicted. Relevant variables include mass, path separation, and propagation time.
Mechanical-noise measurements. Sensitive resonators, cantilevers, and suspended or levitated objects can constrain the additional force noise predicted by collapse. These are often called non-interferometric tests because they need not create a resolved spatial superposition.
Heating measurements. Precision measurements of energy deposition, including studies of cold atomic systems, limit unexplained heating compatible with particular collapse models.
Radiation searches. Low-background detectors can constrain spontaneous radiation attributed to collapse noise. Translating observations into limits requires assumptions about the noise spectrum and the material’s microscopic response.
Ordinary environmental effects—including residual gas collisions, thermal radiation, vibrations, and electromagnetic disturbances—can also reduce coherence or generate noise. A discrepancy becomes evidence for collapse only if competing explanations are adequately excluded and its dependence on experimental conditions matches a specific prediction.
No objective-collapse model has been established by an experimentally confirmed collapse-specific signal. Bounds on simple white-noise models do not automatically apply to colored-noise, dissipative, or differently coupled variants.
What physically exists in a collapse theory?
A dynamical equation does not by itself settle what the theory says exists in ordinary space. Two prominent proposals are:
- Matter-density ontology: a spatial density calculated from the quantum state represents the distribution of matter. Collapse concentrates that density into approximately definite configurations.
- Flash ontology: discrete localization events are fundamental occurrences in spacetime, and macroscopic objects are represented by patterns of such events.
These alternatives matter because a wave function for many particles is defined on configuration space, not simply on three-dimensional physical space. Different ontologies can share closely related collapse dynamics while assigning different physical meanings to it.
Open issues and limitations
Residual tails
Gaussian localization suppresses distant components but does not make them exactly zero. Consequently, a post-collapse state retains small-amplitude “tails.” Their significance depends on the ontology and on what constitutes a sufficiently definite macroscopic state. This is an interpretive issue even when the tails make no practical difference to predictions.
Relativity and nonlocality
Collapse involving entangled systems is generally nonlocal. Avoiding controllable superluminal signaling is necessary but does not by itself establish compatibility with special relativity.
Relativistic collapse models have been constructed in restricted settings, including flash-based models. Extending them to realistic interacting quantum field theories while controlling divergences, energy production, and causal structure remains an important research problem.
Parameter choice and physical mechanism
Basic models introduce phenomenological constants, especially a collapse rate and localization length, without deriving their values from a more fundamental theory. Experimental bounds narrow the allowed ranges but do not determine a unique model.
The origin of collapse noise is likewise unsettled. In different proposals it may be fundamental, connected with gravity, or associated with additional physical degrees of freedom. These possibilities entail different predictions and different accounts of energy exchange.
Emergence of classical behavior
Localization can help maintain approximately definite positions and recover aspects of classical mechanics, but localization alone does not explain every classical property. Classical behavior also depends on interactions, environmental decoherence, dynamical stability, and the chosen physical ontology. Successful collapse models must therefore reproduce both microscopic quantum phenomena and the observed behavior of macroscopic matter.