No exact finite-dimensional CCR

In plain terms

Position and momentum obey the most famous relation in quantum mechanics, the canonical commutation relation, which says that applying the two operations in opposite orders gives results that differ by a fixed universal amount. Everything characteristic of the theory, the uncertainty principle above all, flows from that relation. CSD holds that the world has only finitely many distinguishable states, so the question is whether the relation can survive in such a world.

It can't, and the proof fits in a few sentences. In quantum mechanics, position and momentum are represented by square tables of numbers called matrices, and a finite world means finite tables. Adding up the entries along the main diagonal of a matrix gives a number called its trace, and the trace has a useful property, which is that the trace of a product of two matrices doesn't depend on the order in which they're multiplied. The left side of the relation is the difference between two products in opposite orders, so its trace is always zero. The right side is a fixed nonzero constant placed on every diagonal entry, so its trace is that constant times the number of states, which isn't zero. Zero can't equal something nonzero, so no finite tables can satisfy the relation.

So any finite theory owes an account of where the relation fails, and the question isn't whether it fails but where.

In CSD

The corpus pays this debt exactly rather than apologising for it. The model is the truncated oscillator, the harmonic oscillator, the standard quantum model of vibration, with its infinite ladder of energy levels cut off at a finite top rung. On that ladder the relation holds perfectly on every energy level except the very top one, where it fails by a single localised defect of precisely the size the trace argument demands. The defect isn't noise to be minimised, as it's the exact compensation the counting requires, pushed into the one place a finite ladder must end.

For any state a laboratory can actually prepare, with negligible population at the top rung, the departure is invisible. The result is deliberately stated without reference to CSD's own assumptions, because the obligation binds every finite- dimensional reconstruction of quantum mechanics, so a rival programme claiming finiteness inherits the same debt whether or not it has noticed.

Mathematically

No finite matrices satisfy [Q,P] = c times identity with c nonzero, as the trace of a commutator vanishes while the trace of c times identity is c times the dimension. As a corollary, the constant can't be i h-bar in finite dimension, and the corpus theorem is no_exact_finite_ccr.

The companion construction gives the sharp positive statement, a conjugate pair on the N-level oscillator whose commutator is exactly i everywhere except the top Fock level, where a rank-one defect of magnitude N sits, which is exactly what makes the trace vanish.

Related entries

• Field arenaTake finitely many field modes, cap the number of quanta each can hold, and the whole toolbox of quantum field theory, propagators, light cones, Wick's theorem and thermal states, reappears as exact theorems on one finite arena.
• Wick's theoremThe bookkeeping rule at the heart of quantum field theory, which reduces any correlation of field operators to a sum over pairings, survives a finite cutoff exactly, and the threshold at which it fails is part of the theorem.
• Hermitian operatorA single symmetry condition on an operator guarantees that every number it can produce in a measurement is real, and it delivers the whole machinery of outcomes and eigenstates at no extra cost.
• Robertson uncertainty relationRobertson's inequality is the precise form of the uncertainty principle, and in CSD the spreads it bounds are the spreads of quantities genuinely distributed over the underlying state space and not merely abstract expectation values.
• Lie-Trotter product formulaAlternate many tiny steps of two operations that don't commute and the result converges to the exponential of their sum, with an error that shrinks like one over the number of steps, and CSD proves this in textbook generality, in a form ready for the Mathlib library, because its open-systems tier needed it and nothing on the shelf existed.