Wightman-Framework Models, Counterexamples, and Scope

The Wightman framework contains rigorously constructed interacting models, but its best-controlled examples are concentrated in low spacetime dimensions. Massive free and generalized free fields realize the axioms transparently; polynomial interactions in two dimensions realize genuine interaction after hard constructive work. A formal Lagrangian or a perturbative correlator list is not thereby a Wightman model, and no nontrivial four-dimensional $\lambda\phi^4$ construction presently plays the same role as $P(\phi)_2$.

Required background. The Wightman reconstruction theorem gives the existence criterion; CPT theorem variants and spin–statistics failure modes give structural conclusions; and Haag-theorem variants and proposed evasions distinguishes regulated from exact continuum representations.

Helpful background. Quantizing the real scalar field supplies the benchmark model, while distributional correlators and zero modes shows why low-dimensional correlators still require distributional and infrared care.

A model-status comparison

The following entries compare models at the level actually established, rather than assigning the same status to a construction and a formal ansatz.

Model	Common field domain and positivity	Spectrum and locality	Clustering and particles	Interaction status
Massive free scalar	Finite-particle polynomial domain in positive Fock space	Positive mass shell; Pauli–Jordan commutator is causal	Unique vacuum, mass gap, one-particle shell; clustering	Exactly constructed; Gaussian and noninteracting
Generalized free field	Fock space over a direct integral from a positive tempered spectral measure	Positive spectral support; local commutator after integrating free masses	Depends on the measure; an isolated atom gives a particle shell	Exactly reconstructed but Gaussian; no nonzero truncated functions above order two
Stable $P(\phi)_2$	Interacting Hilbert space and controlled unbounded fields obtained after removing cutoffs	Wightman covariance, spectrum, and locality established for the constructed model	Vacuum and particle properties proved in specified coupling regimes	Rigorously constructed and genuinely interacting in $1+1$ dimensions
Formal four-dimensional $\lambda\phi^4$ correlators	Usually regulated or formal; no completed positive Wightman hierarchy is supplied	Order-by-order covariance/locality does not establish an exact spectral-positive model	Perturbative pole and scattering statements are conditional on the construction	Not a nontrivial rigorously constructed Wightman model

The constructive status of the low-dimensional examples and the dimensional contrast are reviewed in Summers 2016, §§ 2–3.3, pp. 5–20. In particular, the $P(\phi)_2$ construction is not inferred from its classical polynomial. It requires cutoff Hamiltonians or Euclidean measures, stability, renormalization, infinite-volume and ultraviolet limits, positivity, analytic continuation, and verification of the axioms.

Three benchmark constructions

Massive free scalar

The finite-particle domain is invariant under smeared fields and Poincaré transformations. Its $n$-point hierarchy is Gaussian, with

$$W_{2n}=\sum_{\text{pairings}}\prod W_2, \qquad W_{2n+1}=0.$$

All truncated functions above order two vanish, and the one-particle mass shell is isolated. This model is the normalization and sign check for the chapter, not evidence that every Wightman field is Gaussian.

Generalized free field

A positive polynomially bounded spectral measure $\rho$ defines

$$W_2(x)=\int_0^\infty\rho(\mathrm d\mu^2)\,\Delta_+(x;\mu^2),$$

and Wick pairings define the hierarchy. Reconstruction gives a positive local field. If $\rho$ has continuous support, the two-point function need not obey one finite-order Klein–Gordon equation. Yet all higher truncated functions still vanish. Generalized free fields therefore show that the Wightman axioms do not imply a canonical Lagrangian, a single particle mass, or interaction; see Greenberg 1961, pp. 158–176.

Constructed P(φ)₂

For suitable real polynomials $P$ bounded below, constructive methods define the two-dimensional interaction, remove cutoffs, and obtain vacuum functions satisfying the Wightman axioms. The field domain is not justified by the free formal expression $H_0+\int P(\phi)$ alone; it belongs to the limiting interacting representation. For important regimes, a mass gap and particle/scattering properties can be established. A primary result proving the axioms and particle structure is Glimm, Jaffe, and Spencer 1974, pp. 585–632.

This three-way comparison is the exact application developed further on rigorous status, construction, and open problems.

Counterexamples to overstrong conclusions

Each modification diagnoses a different missing hypothesis or invalid converse.

• Give the spectral measure a negative component. Covariance and temperedness may survive, but $W_2(\bar f,f)$ can be negative, so Hilbert-space reconstruction fails.
• Use a symmetric positive- and negative-energy mass-shell measure. Lorentz invariance may survive, but the vacuum spectrum condition and forward-tube analyticity fail.
• Keep a positive generalized-free two-point function while choosing incompatible higher functions. Two-point positivity does not imply positivity of the full hierarchy.
• Retain positivity and spectrum but replace the local commutator by a nonlocal kernel. Reconstruction may yield a field representation, but the Wightman locality axiom and its CPT/spin–statistics proof route are absent.
• Use an indefinite metric for a covariant gauge potential. A useful gauge-fixed formulation may result, but positive-metric Wightman conclusions cannot be quoted for the potential without a physical-state or observable construction.

These examples also show that the Wightman conditions are not logically equivalent to “comes from a Lagrangian.” Reconstruction begins with correlators, and many formal Lagrangians never yield a proven positive hierarchy.

Open four-dimensional scope

The absence of a fully constructed target must be stated as a status boundary, not promoted to a no-go theorem. Rigorous triviality results constrain important classes of positive-coupling scalar limits in and above four dimensions, but the detailed hypotheses and limit procedures matter. They do not license the blanket statement that every imaginable four-dimensional interacting Wightman theory is impossible. Likewise, the absence of a construction is not proof of inconsistency.

Gauge theories sharpen the distinction. Gauge-invariant local observables may fit a positive-metric algebraic framework even when covariant gauge potentials require indefinite metric or nonlocal dressings. The Wightman point-field framework remains a stringent benchmark, not the only rigorous language for QFT.

An independent consistency check is to ask what data certify “interaction.” A nonzero truncated four-point function distinguishes $P(\phi)_2$ from both free and generalized free fields. A broad spectral density by itself does not: generalized free fields can have a continuum and still be Gaussian.

Exercises

Suppose $\rho=c\,\delta(\mu^2-m^2)+\rho_c$ with $c>0$ and $\rho_c$ supported above a threshold $M^2>m^2$. What particle information is visible in $W_2$?

Solution

The atom at $m^2$ gives an isolated positive-energy mass shell and hence a stable one-particle component with weight $c$. The continuous part begins at $M^2$ and contributes continuum spectral states. This information does not establish interaction: a generalized free field with this same measure remains Gaussian and has vanishing higher truncated functions.

References

• Glimm, James, Arthur Jaffe, and Thomas Spencer. 1974. “The Wightman Axioms and Particle Structure in the $P(\phi)_2$ Quantum Field Model.” Annals of Mathematics 100: 585–632. DOI.
• Greenberg, Oscar W. 1961. “Generalized Free Fields and Models of Local Field Theory.” Annals of Physics 16: 158–176. DOI.
• Streater, Raymond F., and Arthur S. Wightman. 2016. PCT, Spin and Statistics, and All That. Princeton Landmarks in Physics. Princeton University Press. DOI.
• Summers, Stephen J. 2016. “A Perspective on Constructive Quantum Field Theory.” arXiv:1203.3991v2 [math-ph]. Open PDF.