Skip to content

CPT Theorem Variants and Their Hypotheses

In the Wightman setting, the CPT theorem asserts the existence of an antiunitary operator implementing total spacetime inversion together with charge conjugation and the representation-dependent action on field indices. Scalar, charged, and spinorial cases share the analytic proof but not one universal component formula. Algebraic, curved-spacetime, and boundary versions are separate theorems whose geometric and positivity assumptions must be stated anew.

Required background. Analyticity, CPT, and spin–statistics gives the common proof mechanism; Jost points, edge-of-the-wedge, and locality gives weak local commutativity; and Clifford algebras and Pin/Spin groups supplies the representation data for spinorial fields.

Helpful background. CPT: hypotheses, content, and limits gives the physical theorem statement and convention choices.

Consider a finite collection of fields on Minkowski space satisfying temperedness, positive-metric Hilbert-space positivity, Poincaré covariance, the spectrum condition, a cyclic invariant vacuum, and local commutativity with the appropriate grading. Then there is an antiunitary operator Θ\Theta leaving the vacuum invariant and implementing the CPT transformation on the cyclic field domain. For a field multiplet,

Θϕa(x)Θ1=bCabϕb(x),\Theta\phi_a(x)\Theta^{-1} =\sum_b C_a{}^b\,\phi_b^\dagger(-x),

where CC includes intrinsic phases and the matrices required by the Lorentz representation. This schematic formula must be specialized before component signs are used. For a neutral Hermitian scalar it reduces to Θϕ(x)Θ1=ηϕ(x)\Theta\phi(x)\Theta^{-1}=\eta\phi(-x); for a charged scalar it exchanges ϕ\phi and ϕ\phi^\dagger; for spinors it also acts on spinor indices.

The theorem is proved at the level of vacuum distributions. Spectrum support gives tube analyticity, complex Lorentz covariance reaches total inversion, and local commutativity reverses the field order at Jost points. Analytic uniqueness extends the relation, and reconstruction turns the correlator identity into an antiunitary operator. The theorem and this proof are given in Streater and Wightman 2016, §§ 4-1–4-3, pp. 134–145; Jost’s original formulation isolates weak local commutativity in Jost 1957, pp. 409–416.

Under the other Wightman assumptions, weak local commutativity at Jost points is not merely sufficient: the Jost result relates it to CPT invariance. “Weak” refers to a vacuum-correlation identity on the Jost set, not to approximate locality.

Antiunitarity resolves an apparent positive-energy paradox. If U(a)=eiPaU(a)=e^{iP\cdot a} and ΘU(a)Θ1=U(a)\Theta U(a)\Theta^{-1}=U(-a), then

ΘiΘ1=iΘPμΘ1=Pμ.\Theta i\Theta^{-1}=-i \quad\Longrightarrow\quad \Theta P^\mu\Theta^{-1}=P^\mu.

Thus CPT maps a positive-energy particle to an antiparticle of the same four-momentum; it does not create a negative-energy state. The field argument is inverted while the antiunitary conjugation reverses the phase in the translation exponential.

Vacuum functions consequently obey an inversion, conjugation, and reversed-order relation with the relevant field matrices. Omitting the reversed order is a common mistake: antiunitarity alone complex-conjugates matrix elements, while locality and analytic continuation supply the reordering.

Write the free complex field as

ϕ(x)=d3p(2π)32Ep(a(p)eipx+b(p)eipx),\phi(x)=\int\frac{\mathrm d^3\mathbf p}{(2\pi)^3\,2E_{\mathbf p}} \left(a(\mathbf p)e^{-ip\cdot x} +b^\dagger(\mathbf p)e^{ip\cdot x}\right),

with normalization factors absorbed consistently into aa and bb. Define Θ\Theta antiunitarily by

ΘΩ=Ω,Θa(p)Θ1=ηb(p),Θb(p)Θ1=ηˉa(p).\Theta\Omega=\Omega,\qquad \Theta a^\dagger(\mathbf p)\Theta^{-1} =\eta\,b^\dagger(\mathbf p),\qquad \Theta b^\dagger(\mathbf p)\Theta^{-1} =\bar\eta\,a^\dagger(\mathbf p).

Because coefficients are complex-conjugated, substitution gives Θϕ(x)Θ1=ηϕ(x)\Theta\phi(x)\Theta^{-1}=\eta\,\phi^\dagger(-x) after a consistent phase choice. The same map exchanges charged one-particle states while preserving their momentum and norm. Applying it to products reproduces the conjugated, inverted, reversed vacuum correlators. This is the first application needed before interpreting charged matter on dynamical gauge fields and matter; the interacting gauge-theory complications are not assumed here.

An independent check applies CPT twice. On the scalar field, Θ2ϕ(x)Θ2=η2ϕ(x)\Theta^2\phi(x)\Theta^{-2}=|\eta|^2\phi(x), so unit normalization requires η=1|\eta|=1. The value of Θ2\Theta^2 on a full multiplet can include representation and internal-symmetry information; it should not be inferred from the scalar formula.

  • Neutral versus charged fields. Neutral fields may map to themselves; charged fields map to conjugate multiplets. Charge conjugation is part of the combined transformation even when CC, PP, or TT separately fails.
  • Spinorial fields. The complex Lorentz cover and the finite-dimensional spin representation determine additional matrices and phases. A scalar formula is not a spinor theorem.
  • Algebraic formulations. CPT may be expressed as an anti-automorphism or modular-geometric symmetry of local observable algebras. Such results replace point fields by net assumptions and have their own hypotheses.
  • Curved spacetime or boundaries. There may be no global inversion isometry and no translation spectrum cone. Locally covariant CPT statements compare appropriately orientation-reversed theories; they are not the Minkowski theorem applied word for word.

As an adversarial example, put the theory on a half-space with a boundary condition not invariant under xxx\mapsto-x. Translation covariance, the full Lorentz group, and the global Jost geometry used above are absent. Failure to obtain the Wightman CPT operator is therefore expected and does not refute the theorem.

Show from antiunitarity that ΘPμΘ1=Pμ\Theta P^\mu\Theta^{-1}=P^\mu is compatible with ΘU(a)Θ1=U(a)\Theta U(a)\Theta^{-1}=U(-a).

Solution

Apply Θ\Theta to U(a)=eiPaU(a)=e^{iP\cdot a}. Antiunitarity sends ii to i-i, so the result is ei(ΘPΘ1)ae^{-i(\Theta P\Theta^{-1})\cdot a}. Equality with U(a)=eiPaU(-a)=e^{-iP\cdot a} for all aa gives ΘPμΘ1=Pμ\Theta P^\mu\Theta^{-1}=P^\mu.

  • Jost, Res. 1957. “Eine Bemerkung zum CTP-Theorem.” Helvetica Physica Acta 30: 409–416. Digitized article.
  • Streater, Raymond F., and Arthur S. Wightman. 2016. PCT, Spin and Statistics, and All That. Princeton Landmarks in Physics. Princeton University Press. DOI.