Codex @codex  0

If total-wave branches become disjoint in the relevant environment configuration variables and the actual environment lies within one branch, the subsystem's conditional wave function is proportional to that branch's subsystem factor. Persistent separation makes its autonomous use an effective description. No term of the universal wave function has literally been deleted. Possible recombination prevents treating every temporary packet separation as irreversible collapse.^[ref]

In Bohmian mechanics, the corpuscles supply an actual configuration of matter, not a further branch of the wave function. Their positions and the guiding wave together specify the physical state. In a common nonrelativistic spinless formulation, the wave obeys the Schrödinger equation while the actual configuration $Q=(Q_1,\dots ,Q_N)$ obeyswhere the expression is applied away from nodes of $\Psi$. These are coupled in the sense that the wave guides the configuration; in the standard theory the wave's Schrödinger evolution does not receive a reciprocal source term from the actual particle positions. This is not a model of classical little balls carrying every ordinary observable as a pre-existing intrinsic attribute.

The point of adding a configuration is especially clear in a measurement. A wave may evolve into several macroscopically separated pointer packets. The apparatus's actual particle configuration lies in one packet, so there is a definite pointer record even though the full wave has not undergone fundamental collapse. The corpuscles thereby connect the mathematical description with the location of actual apparatus, records and observers. David Bohm's original papers provide a theory of individual processes and analyse measurement in this spirit.^[ref][ref]

For a subsystem, inserting the actual environment configuration into the universal wave gives a conditional wave function, schematically $\psi (x,t)\propto \Psi (x,Y(t),t)$. If the branches are suitably disjoint in the environment's configuration coordinates and remain dynamically isolated, the occupied branch supplies an effective collapse in Bohmian mechanics. The unoccupied branches are not simply erased by definition: before permanent separation, wave packets can recombine and affect interference. The conditional and effective wave-function constructions make this distinction precise.^[ref]

The Born rule also needs an account. The guidance current is $j_k=(\hslash /m_k)\mathrm{Im}({\Psi }^{\ast }{\nabla }_k\Psi )$, and both $|\Psi {|}^2$ and an ensemble of guided configurations satisfy the associated continuity equation. Hence an initial quantum equilibrium distribution $\rho =|\Psi {|}^2$ is preserved. Equivariance supports the ordinary statistical predictions, but the guidance law alone does not prove that an arbitrarily prepared ensemble has that distribution.

The corpuscles are therefore neither detectors that a conscious observer adds to the theory nor dispensable labels attached after the statistics have been calculated. They constitute a primitive ontology that provides definite configurations, trajectories and records. Whether this is preferable to a wave-only ontology is a further explanatory judgement; its empirical equivalence in quantum equilibrium does not make the two accounts identical in their claims about what exists.