Canonical Commutation Relation

Equivalent forms

\hat{x}\hat{p} - \hat{p}\hat{x} = i\hbar

[\hat{x}_i, \hat{p}_j] = i\hbar\,\delta_{ij}

Classical mechanics is the special case where this commutator vanishes — ℏ measures exactly how non-classical nature is.

Symbol	Name	SI unit	Dimension	Range
$sigma_x$	Position uncertainty Standard deviation of position in the quantum state	m	L	1e-12 – 1e-8
$sigma_p$	Minimum momentum uncertaintyoutput Smallest momentum spread allowed: hbar/(2*sigma_x)	kg*m/s	MLT^-1	1e-27 – 1e-22
$hbar$	Reduced Planck constant h/(2*pi)	J*s	ML^2T^-1	1.054571817e-34 – 1.054571817e-34

Unit systems

Symbol	SI	CGS	Imperial
$sigma_x$	m	cm	ft
$sigma_p$	kg*m/s	g*cm/s	lb*ft/s
$hbar$	J*s	erg*s	ftlbfs

Where it holds

Exact for canonically conjugate pairs (x and p_x, etc.) in non-relativistic and relativistic quantum theory alike. Modified-commutator proposals (generalized uncertainty principles) arise only in quantum-gravity phenomenology, untested.

Dimensional analysis

[x][p] = L * (M*L*T^-1) = M*L^2*T^-1

[hbar

] = J*s = M*L^2*T^-1

Commutator and hbar share units of action.

\checkmark

Discovery

Max Born and Pascual Jordan · 1925

Decoding Heisenberg's strange arrays of numbers, Born recognized them as matrices and derived pq - qp = h/(2*pi*i) — calling it the fundamental quantum condition. The relation meant so much to Born that it is engraved on his tombstone in Gottingen.

Try this

Why can't you know position and momentum at the same time?

All of quantum weirdness traces back to one algebraic fact: the position and momentum operators do not commute. Their commutator is not zero, not random — it is exactly i times Planck's constant.

Research status: stable

Real-world applications

Squeezed-light interferometry (LIGO sensitivity upgrades)
Quantum-limited amplifiers in superconducting-qubit readout
Canonical quantization recipe for building any quantum field theory
Zero-point motion that keeps liquid helium from freezing at atmospheric pressure

Common misconceptions

It is not about measurement clumsiness — the non-commutativity is intrinsic to the state, present before any measurement
i*hbar is a multiple of the identity operator, not a number you measure directly
Energy and time do NOT satisfy this relation — time is a parameter, not an operator, so the energy-time uncertainty has a different origin

Experimental verification

Every confirmation of the uncertainty principle tests this relation: electron diffraction widths, squeezed light at LIGO (squeezing one quadrature inflates the conjugate one exactly as the commutator demands), and sub-Heisenberg attempts that always fail.

Derivation

Represent p as the operator -i*hbar*d/dx acting on wavefunctions.

Then for any psi:

(x*p - p*x) \psi = -i

*hbar*x*psi' + i*hbar*

d/dx(x*\psi ) = -i

*hbar*x*psi' + i*hbar*psi + i*hbar*

x*\psi ' = i

*hbar*psi.

Hence [x,

p] = i

*hbar as an operator identity.

The Robertson inequality sigma_A*sigma_B >= |<[A,B]>|/2 then yields sigma_x*sigma_p >= hbar/2 — uncertainty is a theorem of this algebra, not a separate postulate.

Limiting cases

hbar -> 0⟶ [x, p] -> 0Observables commute — the classical limit where order of measurement is irrelevant.

Gaussian wavepacket⟶

\sigma _x * \sigma _p =

hbar/2Gaussians saturate the bound — they are minimum-uncertainty states.

Same-direction position pair [x, x] or perpendicular [x, p_y]⟶ commutator

= 0

Compatible observables — both can be known simultaneously with arbitrary precision.

What if…

What if x and p commuted?

Atoms would collapse: nothing would stop electrons from sitting at the nucleus with zero momentum. Matter is stable because the commutator forbids that state.

What if hbar were twice as large?

All uncertainty products would double; atoms would be larger and chemistry would happen at different energy scales.

What if there were a minimal length (quantum gravity)?

Generalized uncertainty proposals add a term $\sim \beta *p^2$ to the commutator, preventing sigma_x from going below the Planck length — experimentally unconstrained so far.

1

Electron confined to an atom-sized region

Given ·

sigma x:: 1e-10

Find · minimum sigma_p and corresponding speed

Steps

sigma_p >= hbar/ $(2*\sigma _x) = 1.054571817e-34 / (2e-10)$
$= 5.27e-25\,\mathrm{kg}*m/s$
$v = \sigma _p / m_e = 5.27e-25 / 9.1093837015e-31 \sim 5.8e_{5} m/s$

Result ·

\sigma _p =

hbar/

(2*\sigma _x) \sim 5.27e-25\,\mathrm{kg}*m/s

, so

v \sim 5.8e_{5} m/s

— confinement alone forces atomic electrons to move fast.

2

From commutator to uncertainty bound

Given ·

commutator:: i*hbar

Find · lower bound on sigma_x * sigma_p

Steps

Robertson inequality: sigma_A*sigma_B >= |<[A,B]>|/2
<[x,p]> $= i$ *hbar for every state
sigma_x*sigma_p >= hbar/ $2 \sim 5.27e-35\,\mathrm{J}*s$

Result · Robertson: sigma_x*sigma_p >= |<[x,p]>|/

2 =

hbar/2.

Heisenberg Uncertainty Principle→Time-Independent Schrödinger Equation→Spin-1/2 Operators→Density Matrix→