Poisson brackets — definition, properties, connection to commutators: Question bank

Reminders you will lean on (all built in the parent note):

The bracket ${f, g} = \sum_{i} (\frac{\partial f}{\partial q _{i}} \frac{\partial g}{\partial p _{i}} - \frac{\partial f}{\partial p _{i}} \frac{\partial g}{\partial q _{i}})$ — " $q$ -then- $p$ minus $p$ -then- $q$ ".
Equation of motion: $\frac{df}{d t} = {f, H} + \frac{\partial f}{\partial t}$ .
Fundamental brackets: ${q_{i}, p_{j}} = δ_{ij}$ , ${q_{i}, q_{j}} = {p_{i}, p_{j}} = 0$ .
Quantization map: ${f, g} \to \frac{1}{i ℏ} [\hat{f}, \overset{g}{^}]$ .

True or false — justify

True or false:

{f, f} = 0

for every phase-space function

f

True — antisymmetry forces

{f, f} = - {f, f}

, and the only number equal to its own negative is zero. No calculation needed.

True or false:

{p, q} = 1

, just like

{q, p} = 1

False. The bracket is antisymmetric, so

{p, q} = - {q, p} = - 1

. Unlike a dot product, order flips the sign.

True or false: if

{f, H} = 0

then

f

is automatically conserved.

Only if

f

has no explicit time dependence. The full law is

\frac{df}{d t} = {f, H} + \frac{\partial f}{\partial t}

, so a time-dependent

f

can have

{f, H} = 0

yet still change.

True or false: the Poisson bracket needs Hamilton's equations built into its definition.

False. The bracket is pure kinematics — just derivatives. Dynamics enters only when you set the second slot to

H

and plug in Hamilton's equations.

True or false:

{f, c} = 0

for any constant

c

True — a constant has zero derivative with respect to every

q_{i}

and

p_{i}

, so every term in the sum vanishes.

True or false: since

{q, p} = 1

and quantization sends

{f, g} \to \frac{1}{i ℏ} [\hat{f}, \overset{g}{^}]

, we get

[\overset{q}{^}, \overset{p}{^}] = 1

False. Multiply back:

{q, p} = \frac{1}{i ℏ} [\overset{q}{^}, \overset{p}{^}]

gives

[\overset{q}{^}, \overset{p}{^}] = i ℏ

. Dropping the

i ℏ

breaks the units and the

ℏ \to 0

limit.

True or false: the Poisson bracket of two conserved quantities is again conserved.

True — this is Poisson's theorem, a consequence of the Jacobi identity. If

{A, H} = {B, H} = 0

, Jacobi forces

{H, {A, B}} = 0

True or false:

{f, g}

is a number.

False in general — it is another function on phase space, itself dependent on

(q, p, t)

. It reduces to a number only when the derivatives happen to collapse (as for

{q_{i}, p_{j}} = δ_{ij}

True or false: the bracket is symmetric under swapping the coordinate and momentum roles, so

{q, q}

and

{p, p}

are both

1

False.

{q_{i}, q_{j}} = 0

and

{p_{i}, p_{j}} = 0

; only mixed

q

–

p

pairs give the nonzero

δ_{ij}

. Two of the same kind never pair up in the definition.

Spot the error

A student writes

{f g, h} = f {g, h} \cdot g {f, h}

. What's wrong?

The Leibniz rule is a sum, not a product:

{f g, h} = f {g, h} + {f, h} g

. The bracket is built from first derivatives, which obey the additive product rule.

A student concludes "

H

is always conserved because

{H, H} = 0

." What's the flaw?

{H, H} = 0

only gives

\frac{d H}{d t} = \frac{\partial H}{\partial t}

. If

H

depends explicitly on time (e.g. a driven oscillator), energy is not conserved despite the zero bracket.

A student claims

{L_{x}, L_{z}} = + L_{y}

. Where's the sign slip?

The correct result is

{L_{x}, L_{z}} = - L_{y}

. The general rule

{L_{i}, L_{j}} = ϵ_{ij k} L_{k}

has

ϵ_{132} = - 1

, so the ordering

x, z

produces a minus sign.

A student writes

\frac{df}{d t} = {H, f} + \frac{\partial f}{\partial t}

. What's wrong?

The order is reversed. It must be

{f, H}

; writing

{H, f}

flips the sign of the dynamical term, giving the wrong direction of time evolution.

A student says "brackets and commutators are literally the same object." Correct that.

They share the same algebra (antisymmetry, bilinearity, Leibniz, Jacobi) but live in different worlds: brackets act on classical functions, commutators on operators. The map between them carries a factor

\frac{1}{i ℏ}

A student expands

{a f + b g, h}

ab {f, h} + {g, h}

. Fix it.

Bilinearity is

{a f + b g, h} = a {f, h} + b {g, h}

— the constants stay attached to their own terms, they do not multiply together.

Why questions

Why does the bracket use

q

-then-

p

minus

p

-then-

q

rather than a plus?

The minus sign is what makes the bracket antisymmetric, encoding that

q

and

p

are conjugate partners with opposite roles in Hamilton's equations (

\overset{q}{˙} = + \partial H / \partial p

\overset{p}{˙} = - \partial H / \partial q

Why does the Jacobi identity matter physically?

It guarantees the bracket of two conserved quantities is conserved, so you can generate new conservation laws — e.g. two components of angular momentum yielding the third.

Why must the quantization map carry a factor of

\frac{1}{i ℏ}

It fixes the units (a classical bracket and a commutator have different dimensions) and ensures

ℏ \to 0

correctly recovers the classical equation of motion.

Why can Hamilton's equations be seen as a special case of

\frac{df}{d t} = {f, H}

Set

f = q_{i}

to get

\overset{q}{˙}_{i} = \partial H / \partial p_{i}

, and

f = p_{i}

to get

\overset{p}{˙}_{i} = - \partial H / \partial q_{i}

. The bracket formalism contains mechanics as the coordinate-choice of

f

Why is

{f, H} = 0

the natural test for a conserved quantity?

Because the bracket with

H

is the rate of change of

f

along the flow (barring explicit time dependence). Zero flow means the quantity is carried unchanged by the dynamics.

Edge cases

For a free particle

H = \frac{p ^{2}}{2 m}

, what is

{p, H}

and why?

It is

0

H

has no

q

-dependence, so

\partial H / \partial q = 0

, and the

\frac{\partial p}{\partial q}

term also vanishes. Momentum is conserved.

Under a constant force,

f = p - F t

. Is

{f, H}

enough to describe

\frac{df}{d t}

No —

f

depends explicitly on

t

, so

\frac{df}{d t} = {f, H} + \frac{\partial f}{\partial t}

, and the

\frac{\partial f}{\partial t} = - F

term is essential.

f

is a constant function on phase space, what is

{f, g}

for any

g

Zero — all derivatives of a constant vanish, so every term in the sum is zero regardless of

g

What is

{q_{1}, p_{2}}

in a multi-coordinate system?

Zero. The fundamental bracket

{q_{i}, p_{j}} = δ_{ij}

is nonzero only when the indices match; different coordinates (

1 \neq = 2

) do not pair.

In the classical limit

ℏ \to 0

, what happens to the Heisenberg equation

\frac{d f ^}{d t} = \frac{1}{i ℏ} [\hat{f}, \hat{H}]

It reduces to the classical

\frac{df}{d t} = {f, H}

, since

\frac{1}{i ℏ} [\hat{f}, \hat{H}] \to {f, H}

— the quantum and classical laws share one structure.

Connections

Hamiltonian Mechanics — the equations that turn the bracket into dynamics.
Noether's Theorem & Conservation Laws — the ${f, H} = 0$ conservation tests above.
Angular Momentum Algebra — source of the ${L_{i}, L_{j}}$ sign traps.
Commutators in Quantum Mechanics — the $\frac{1}{i ℏ}$ quantization traps.
Canonical Transformations — why bracket structure is the thing preserved.
Liouville's Theorem — same antisymmetric structure, phase-space volume.