Fakeon Prescriptions in Quantum Field Theory

• Fakeon prescriptions are a framework for quantizing strictly virtual degrees of freedom that propagate internally without ever reaching on-shell status.
• They replace the standard iε prescription with an average continuation, thereby preserving unitarity, Lorentz invariance, and power-counting renormalizability in complex-pole theories.
• Applied in quantum gravity and Lee–Wick models, fakeon methodologies resolve ghost issues and introduce unique nonperturbative corrections, including peak uncertainty.

Fakeon prescriptions define a mathematically consistent framework for quantizing degrees of freedom that are strictly virtual, ensuring that certain fields—while propagating internally in Feynman diagrams—never go on shell and thus do not correspond to physical asymptotic states. The fakeon, short for "fake particle," arises naturally in higher-derivative and nonlocal quantum field theories, notably in quantum gravity and Lee–Wick extensions, where complex conjugate poles appear in propagators. The fakeon prescription provides a uniformly consistent approach to amplitude construction, preserving unitarity, Lorentz invariance, and renormalizability, unlike alternative contour or analytic continuation prescriptions (Anselmi et al., 3 Mar 2025, Anselmi, 2020).

1. Mathematical Foundation and Prescription

A fakeon replaces the standard Feynman $+i\epsilon$ prescription with an "average continuation" or principal value assignment at the propagator pole. The canonical form for a fakeon of mass $m$ is

$$G_{\text{fakeon}}(p) = \lim_{\epsilon\to 0} \frac{1}{2}\left(\frac{i}{p^2-m^2+i\epsilon} + \frac{i}{p^2-m^2-i\epsilon}\right) = i\,\mathrm{P.V.}\frac{1}{p^2-m^2},$$

where P.V. denotes the distributional principal value (Anselmi, 2020). This prescription is implemented by the "average continuation" of the analytically continued amplitude $M(p^2)$ across the physical cut,

$$M_{\text{AP}}(p^2) = \frac{1}{2}\left[ M(p^2+i0) + M(p^2-i0) \right],$$

ensuring that fakeons contribute to virtual dynamics but cannot be produced on shell (Anselmi et al., 3 Mar 2025).

For loop integrals with complex-conjugate poles (e.g., in higher-derivative gravity or Lee–Wick models), the energy ($k^0$) contour is deformed in the complex plane to avoid the fakeon poles (Lee–Wick contour $\Gamma_{LW}$), while the spatial momentum integration is further deformed into a specific complex domain $D_k$ to evade extended regions of nonanalyticity and retain Lorentz invariance (Anselmi et al., 3 Mar 2025).

2. Comparison with Alternative Prescriptions

There exist four inequivalent ways to define amplitudes in theories with complex poles (Anselmi et al., 3 Mar 2025). Only the fakeon prescription simultaneously preserves Lorentz invariance, unitarity (optical theorem), and power-counting renormalizability, but sacrifices full analyticity:

Prescription	Lorentz invariance	Analyticity	Optical theorem	Power-counting renorm.
Euclidean + analytic continuation	✔	✔	✘	✔
Lee–Wick–Nakanishi (LW)	✘	✘	✔	✔
Fakeon (Anselmi–Piva)	✔	✘	✔	✔
Direct Minkowski (Feynman ±iε)	✔	✔	✘	✘

The fakeon prescription avoids nonlocal counterterms (present in Feynman–Wheeler or mixed sign prescriptions) and ensures that any degree of freedom quantized as a fakeon does not appear as an intermediate physical state—even above threshold (Anselmi, 2020).

3. Properties of the Fakeon Amplitude and Propagator

The fakeon prescription removes the on-shell discontinuity (the "iπ" from the cut) associated with ordinary particles. In physical terms, the cut propagator for fakeons vanishes, ensuring zero contribution to imaginary parts of amplitudes from fakeon lines,

$$2\,\Im\left[-i\,\Sigma_{\text{fakeon}}(p)\right] = 0 \quad \text{(for loops with a fakeon)},$$

which enforces unitarity even when fake degrees of freedom circulate inside diagrams (Anselmi, 2020).

Dressed propagators with self-energy corrections $\Sigma(p^2)$ yield, for the fakeon,

$$\widehat{P}_\chi(p^2) = \frac{i(p^2-m^2)}{(p^2-m^2)(p^2-m^2-\Sigma)+\epsilon^2},$$

in contrast to standard Breit–Wigner forms for physical particles (Anselmi, 2022). Importantly, geometric resummation of loop series for fakeons fails to admit an analytic continuation near the fakeon pole; the perturbation series diverges nonperturbatively at the resonance, a behavior encoded in the concept of "peak uncertainty" (see Section 5).

4. Connections to Quantum Gravity and High-Energy Applications

The fakeon concept is crucial in quantizing higher-derivative theories where the graviton propagator contains complex conjugate poles—a central issue in renormalizable gravitational models. By quantizing the spin-2 ghost pole as a fakeon

$$S_{\rm QG} = -\frac{1}{2\kappa^2}\int d^4x \sqrt{-g} \left( 2\Lambda_C + \zeta R + \alpha (R_{\mu\nu}R^{\mu\nu} - \tfrac{1}{3}R^2 ) - \frac{\xi}{6} R^2 \right) + S_{\rm matter},$$

the resulting theory is renormalizable, unitary, and Lorentz invariant at the perturbative level, with no physical ghost state in the spectrum. Effective couplings allow gravitational scattering amplitudes to remain under perturbative control up to extremely high energy scales (the so-called "God’s energy") (Anselmi, 2020).

Fakeon quantization has also been systematically applied in Lee–Wick gauge theories—removing interpretational ambiguities and ensuring well-defined higher-order amplitudes—and in fractional QFTs, where non-polynomial kinetic operators naturally generate complex poles (Anselmi et al., 3 Mar 2025).

5. Nonperturbative Aspects, Peak Uncertainty, and Dressed Propagators

The fakeon propagator differs fundamentally from those of physical particles or ghosts when dressing with self-energies. Whereas the physical Breit–Wigner propagator is analytic and summable in the entire complex plane, the fakeon series involves ill-defined sums and contact terms,

$$\widehat{P}_\chi \sim iZ(p^2-m^2)/[(p^2-m^2)(p^2-m^2 + i m\Gamma) + m^2 \Delta E^2],$$

with convergence only outside a "peak region" around $p^2\approx m^2$. Analytical completion fails near the resonance; instead, one must introduce a finite energy resolution $\Delta E$ in measurements, leading to the "peak uncertainty" principle: $$\Delta E \gtrsim \Delta E_{\min} \simeq \Gamma_f/2,$$ where $\Gamma_f$ is the fakeon width. This expresses the impossibility of resolving the fakeon pole arbitrarily closely, reflecting that purely virtual modes cannot be seen on shell (Anselmi, 2022).

For collider-scale resonances (large width, e.g., the $Z$ boson), standard Breit–Wigner summation is used, whereas for quantum gravity (narrow fakeon width, e.g., the spin-2 fakeon), nonperturbative corrections are negligible, and the principal value prescription suffices.

6. Implications for Unitarity, Renormalization, and Causality

The fakeon prescription ensures conservation of probability (unitarity) by eliminating negative norm states from the S-matrix. Unitarity is preserved as fakeons contribute no on-shell cuts, and all UV divergences remain local, in contrast to the Feynman–Wheeler principal value prescription, which leads to nonlocal divergences and negative cross-sections (Anselmi, 2020). The fakeon methodology thus identifies and implements a unique class of "purely virtual quanta" that can interact with ordinary fields exclusively off shell.

Lorentz invariance is restored by the explicit construction of the spatial-momentum integration deformation (domain $D_k$), avoiding the extended nonanalytic regions that break boost invariance in the Lee–Wick–Nakanishi prescription (Anselmi et al., 3 Mar 2025).

7. Summary of Physical and Theoretical Significance

The fakeon prescription establishes a rigorous theory for quantizing degrees of freedom associated with complex-conjugate poles, preserving all essential aspects of quantum field theory except analyticity. It provides a framework in which virtual modes contribute meaningfully in internal lines without inducing ghosts or runaway instabilities, uniquely enabling renormalizable and unitary models of quantum gravity, and resolving foundational issues in higher-derivative field theories and alternative gauge extensions (Anselmi et al., 3 Mar 2025, Anselmi, 2020, Anselmi, 2022).