Skip to content
Back to Quantum Field Theory

Conceptual Notes

Active and Passive Transformations

August 3, 2026 4 min read

Objects Involved in a Transformation

When quantum field theory textbooks discuss transformations, they are usually describing the relation between fields and coordinates. Tong’s lecture notes, for example, say:

we are dealing with an active transformation in which the field is truly shifted.

If we were instead dealing with a passive transformation in which we relabel our choice of coordinates.

Descriptions of this kind conceal an object that is rarely named explicitly: spacetime.

If fields and coordinates were the only objects involved, then shift field and relabel coordinates would not differ in substance; they would merely be two views of the same relative motion.

Once spacetime is included explicitly, we can discuss active and passive transformations more carefully, as well as why spacetime is so often omitted from the discussion.

Active Transformations

An active transformation changes the distribution of a field over spacetime. For a real scalar field 𝜙, let spacetime be a manifold 𝑀. An active transformation may then be represented as

𝐴:FF𝜙𝜙

Here 𝜙(𝑝) =𝜙(𝑝): the value of the transformed field 𝜙 at 𝑝 equals the value of the original field 𝜙 at 𝑝.

It is important that neither the spacetime manifold nor its coordinate system changes under an active transformation. What changes is the field configuration.

For example, a Lorentz transformation sends 𝜙 𝜙, where 𝜙(𝑝) =𝜙(𝑝) and

𝛬𝜇𝜈𝑥𝜈(𝑝)=𝑥𝜇(𝑝).

A common source of confusion is that this expression looks exactly like a Lorentz coordinate transformation. In the active picture, however, it is not a coordinate transformation. It relates the coordinates of 𝑝 to those of 𝑝, with the field taking the same value at these two points before and after the transformation.

In other words, the object acted on by the “Lorentz transformation” here is the field configuration, not the coordinate system.

Passive Transformations

A passive transformation changes the coordinate system on spacetime. It changes the coordinates assigned to the same spacetime point from 𝑥𝜇 to 𝑥𝜇:

𝑃:FF𝑥𝜇𝑥𝜇

For a Lorentz transformation,

𝑝𝑀,𝑥𝜇(𝑝)=𝛬𝜇𝜈𝑥𝜈(𝑝).

Neither spacetime nor the field configuration changes under a passive transformation. This is the transformation commonly meant when we speak of a “Lorentz transformation.”

Comparing Active and Passive Transformations

To compare the two descriptions more carefully, introduce the coordinate map 𝑥:

𝑥:𝑀4𝑝(𝑥0(𝑝),𝑥1(𝑝),𝑥2(𝑝),𝑥3(𝑝)).

Then 𝑥𝜇 =𝑟𝜇 𝑥. Introduce ˜𝜙 such that

˜𝜙(𝑥(𝑝))=𝜙(𝑝),

or equivalently 𝜙 =˜𝜙 𝑥.

The active transformation can therefore be written as

˜𝜙(𝑥(𝑝))=˜𝜙(𝑥(𝑝)).

Now introduce

𝛬:44𝛼𝜇𝛬𝜇𝜈𝛼𝜈.

The component expression

𝛬𝜇𝜈𝑥𝜈(𝑝)=𝑥𝜇(𝑝)

may then be written as

𝛬𝑥(𝑝)=𝑥(𝑝).

Define ˜𝜙 𝑥 =𝜙. It follows that

˜𝜙(𝛬𝑥(𝑝))=˜𝜙(𝑥(𝑝)).

Suppressing 𝑝 and the tilde, and writing composition with the map simply as 𝛬𝑥, gives the familiar expression

𝜙(𝛬𝑥)=𝜙(𝑥).

If 𝛬1 denotes the inverse of 𝛬—without proving its existence here—then

𝑥(𝑝)=𝛬1𝑥(𝑝).

Therefore,

˜𝜙(𝑥(𝑝))=˜𝜙(𝛬1𝑥(𝑝)).

Again suppressing 𝑝 and the tilde, and writing composition as 𝛬1𝑥, yields

𝜙(𝑥)=𝜙(𝛬1𝑥).

For a passive transformation, spacetime and the field configuration remain unchanged:

𝜙(𝑝)=𝜙(𝑝).

The coordinate system changes, meaning that the map 𝑥 is replaced by 𝑥 such that

𝑥(𝑝)=𝛬𝑥(𝑝).

Define

¯𝜙𝑥=𝜙.

Then

𝜙(𝑝)=𝜙(𝑝)¯𝜙(𝑥(𝑝))=˜𝜙(𝑥(𝑝))¯𝜙(𝛬𝑥(𝑝))=˜𝜙(𝑥(𝑝)).

Suppressing 𝑝 and the composition symbol gives

¯𝜙(𝛬𝑥)=˜𝜙(𝑥).

From 𝑥(𝑝) =𝛬 𝑥(𝑝), we obtain

𝑥(𝑝)=𝛬1𝑥(𝑝).

Therefore,

𝜙(𝑝)=𝜙(𝑝)¯𝜙(𝑥(𝑝))=˜𝜙(𝑥(𝑝))¯𝜙(𝑥(𝑝))=˜𝜙(𝛬1𝑥(𝑝)).

Suppressing 𝑝, the composition symbol, and the prime on 𝑥 gives

¯𝜙(𝑥)=˜𝜙(𝛬1𝑥).

Why Is Spacetime Often Omitted?

The background spacetime of quantum field theory is Minkowski spacetime. Its high degree of symmetry makes spacetime itself easy to overlook: applying a transformation to the field configuration and then making the corresponding change of coordinates produces a result equivalent to the original description.

In fact, both the field configuration and the coordinates have changed relative to spacetime. The symmetry of Minkowski spacetime prevents that change from becoming visible. On an arbitrary spacetime, this equivalence no longer holds.

Equivalence of Active and Passive Transformations in Minkowski Spacetime

In Minkowski spacetime, appropriate active and passive transformations can be chosen to have the same final effect: the same field value is associated with the same spacetime coordinate, or equivalently, the coordinate-dependent functions used to describe the field configuration are identical.

Again consider a Lorentz transformation. The active transformation gives

˜𝜙(𝑥(𝑝))=˜𝜙(𝛬1𝑥(𝑝)).

The passive transformation gives

¯𝜙(𝑥(𝑝))=˜𝜙(𝛬1𝑥(𝑝)).

Therefore,

˜𝜙(𝑥(𝑝))=¯𝜙(𝑥(𝑝)).

Moreover,

𝑥(𝑝)=𝛬𝑥(𝑝)=𝑥(𝑝).

Thus the field-configuration function ˜𝜙 obtained from the active transformation and the function ¯𝜙 obtained from the passive transformation take the same value at the same argument. The field-configuration function alone therefore does not reveal whether it arose from an active or a passive transformation.

Active and Passive Transformations Are Generally Inequivalent on an Arbitrary Spacetime

On an arbitrary spacetime, an active transformation changes the field configuration assigned to spacetime points. Different spacetime points may have different curvature, and curvature distinguishes one point from another. An active transformation therefore changes the correspondence between curvature and field value.

A passive transformation changes neither spacetime nor the field configuration on it, so it does not change the correspondence between curvature and field value.

From this perspective, active and passive transformations cannot in general be equivalent.