In a Hamiltonian quantum field theory, the Tamm-Dancoff method approximates the infinite coupled equations for the free-field occupation amplitudes of an interacting state by truncating them at a fixed maximum particle number. In 1953 Dyson showed that iterating the truncated equations generates a divergent vacuum self-energy contribution that cannot be removed by renormalization, and proposed a modified ('new') Tamm-Dancoff method built from amplitudes between the interacting state and the physical interacting vacuum, claiming that the resulting equations close, are free of vacuum-self-energy-type divergences, and remain tractable by the standard particle-number truncation. The one-page letter offered no worked example, convergence analysis, or comparison with alternatives. The problem the letter opens is whether Tamm-Dancoff-type particle-number truncation of a Hamiltonian relativistic field theory can be made finite, consistently renormalized, convergent, and quantitatively reliable in a genuinely interacting theory; no resolution is known. Dyson's own follow-up implemented the new equations only in a lowest approximation, found that self-energies still require renormalization and that nonphysical singularities appear, and produced no numerical results. The 1950s community abandoned the equal-time program; the modern light-front Tamm-Dancoff revival removes the vacuum problem kinematically but replaces it with truncation-induced divergences that require sector-dependent renormalization, and controlled demonstrations exist only in super-renormalizable 1+1-dimensional models.

Background

In a quantum field theory written in Hamiltonian (Schrödinger) form, an energy eigenstate of the interacting fields can be expanded on the Fock space of the noninteracting theory — the direct sum of all states with definite occupation numbers of free-particle normal modes — so that the state is described by amplitudes a(N)a(N) for finding a prescribed set NN of free particles. The Schrödinger equation becomes an infinite set of coupled linear equations for these amplitudes, linked through matrix elements of the interaction Hamiltonian H′H'. The Tamm-Dancoff method, introduced by Tamm (1945) and Dancoff (1950), truncates this infinite set by discarding all amplitudes involving more than a fixed maximum number of particles, leaving a finite problem solvable by standard matrix methods; its nonrelativistic descendant, the particle-hole Tamm-Dancoff approximation, remains standard in nuclear physics and quantum chemistry, where the relativistic vacuum problem does not arise.

Dyson (1953) identified a structural defect of the method when applied to relativistic field theory. Iterating the truncated equation once produces a sum over intermediate states ∑MH′(MN)H′(NM)/(E−EM)\sum_M H'(MN)H'(NM)/(E-E_M) that is violently divergent, because for a typical interaction H′H' creates three particles subject to only one momentum-conservation relation, leaving two unconstrained momenta. This vacuum self-energy divergence cannot be eliminated by renormalization — Dyson called it 'a real inadequacy in the Tamm-Dancoff method'. Physically, every state of the interacting fields contains many particles continually created and annihilated in the vacuum; restricting the wave function to a fixed particle number imposes an artificial correlation between vacuum fluctuations at far-distant points, which then acts spuriously on real particles. Dyson's letter proposed a modification: work not with a(N)a(N) but with amplitudes a(N′,N)=⟨Ψ0∣C(N′)A(N)∣Ψ⟩a(N',N)=\langle\Psi_0|C(N')A(N)|\Psi\rangle, where Ψ\Psi is the interacting state under study, Ψ0\Psi_0 is the physical (interacting) vacuum, and A(N)A(N), C(N′)C(N') are products of free-field annihilation and creation operators for the sets NN, N′N'; the 'minus particles' of N′N' are those present in the physical vacuum but absent from Ψ\Psi. From the two Schrödinger equations one obtains (ϵ−EN+EN′)a(N′,N)=⟨Ψ0∣[C(N′)A(N),H′]∣Ψ⟩(\epsilon-E_N+E_{N'})a(N',N)=\langle\Psi_0|[C(N')A(N),H']|\Psi\rangle with the observable energy difference ϵ=E−E0\epsilon=E-E_0. Dyson claimed (i) that the commutator expands, by Wick's theorem — the standard result reducing products of free-field operators to normal-ordered products times c-number contractions — into a closed set of homogeneous linear equations for the a(N′,N)a(N',N), tractable by the standard Tamm-Dancoff truncation; (ii) that because one particle created or annihilated by H′H' must belong to the sets NN or N′N', the offending three-particle matrix elements are absent, so 'divergences of the unpleasant vacuum self-energy type can no longer appear'; and (iii) that the minus particles restore the symmetry between emission and absorption, bringing the method closer to covariant four-dimensional methods. The letter presented no worked example, no convergence discussion, and no comparison with alternative methods; both load-bearing steps (algebraic closure, and absence of divergent matrix elements) were asserted but not demonstrated.

The question the letter opens — whether a Tamm-Dancoff-type particle-number truncation of a Hamiltonian field theory can become a controlled, convergent, quantitatively reliable nonperturbative method with the vacuum difficulty under control — has a long subsequent history. The modification became known as the new Tamm-Dancoff method and was applied to meson theory in the 1950s-60s. The modern revival quantizes the theory on the light front (evolution in light-front time x+=t+zx^+=t+z, in which all longitudinal momenta are nonnegative so the Fock vacuum is an eigenstate of the full Hamiltonian, kinematically removing Dyson's vacuum problem), but there the truncation itself generates ultraviolet and infrared divergences that must be canceled by counterterms depending on the Fock-space sector in which they act (sector-dependent renormalization), and vacuum physics re-enters through zero modes, the field components of vanishing longitudinal momentum constrained by operator equations. Two reference points frame the comparison Dyson never made: the Bethe-Salpeter equation, the covariant bound-state equation derived from field theory, and lattice regularization, the main controlled nonperturbative alternative.

Problem Statement

Determine whether Tamm-Dancoff-type particle-number truncation of a Hamiltonian quantum field theory can be made a controlled nonperturbative method, as Dyson's 1953 letter proposed. Concretely, with respect to the vacuum-relative ('new') Tamm-Dancoff formulation of Dyson (1953), or its documented descendants such as the light-front Tamm-Dancoff method: (1) Formal closure and finiteness — verify or refute, for a specified class of interactions and quantization, that the vacuum-relative amplitude equations close into a self-consistent set of homogeneous linear equations (including a self-consistent treatment of the vacuum-sector amplitudes) and are free of vacuum-self-energy-type divergences; these are the two claims the letter asserted without proof. (2) Renormalization — construct, or prove impossible, a consistent renormalization scheme for the particle-number-truncated equations, including whatever sector-dependent or cutoff-dependent counterterms the truncation requires, such that observables become independent of the truncation cutoffs in the converged limit. (3) Convergence and reliability — implement the scheme in a genuinely interacting relativistic field theory (beyond models that are both 1+1-dimensional and super-renormalizable — 'super-renormalizable' meaning that only finitely many Feynman diagrams are ultraviolet-divergent — and beyond anharmonic-oscillator systems) and demonstrate, or refute, convergence of physical observables as the truncation particle number and ultraviolet cutoff are increased, and assess quantitative reliability against exact benchmarks (soluble models, lattice calculations) and against alternatives such as the Bethe-Salpeter equation; earlier low-order Fock-space truncations in such theories, such as the two- and three-body scalar-Yukawa solution of Karmanov et al. (2016), are prior work on this item. Items (1)-(3) are conjunctive requirements for an affirmative answer to this question; a rigorous refutation of any one of them for a specified class answers it in the negative. A rigorous negative result — a proof, under explicitly stated assumptions, that no finite, consistently renormalized, convergent Tamm-Dancoff-type truncation exists for a specified class of interacting field theories — would settle the question in the negative for that class.

References

  1. I. Tamm, 'Relativistic interaction of elementary particles', J. Phys. (U.S.S.R.) 9, 449 (1945)
  2. S. M. Dancoff, 'Non-Adiabatic Meson Theory of Nuclear Forces', Phys. Rev. 78, 382 (1950), DOI 10.1103/PhysRev.78.382, https://doi.org/10.1103/PhysRev.78.382
  3. G. C. Wick, 'The Evaluation of the Collision Matrix', Phys. Rev. 80, 268 (1950), DOI 10.1103/PhysRev.80.268, https://doi.org/10.1103/PhysRev.80.268
  4. E. E. Salpeter and H. A. Bethe, 'A Relativistic Equation for Bound-State Problems', Phys. Rev. 84, 1232 (1951), DOI 10.1103/PhysRev.84.1232, https://doi.org/10.1103/PhysRev.84.1232
  5. F. J. Dyson, 'The Use of the Tamm-Dancoff Method in Field Theory', Phys. Rev. 90, 994 (1953), DOI 10.1103/PhysRev.90.994, https://doi.org/10.1103/PhysRev.90.994
  6. F. J. Dyson, M. Ross, E. E. Salpeter, S. S. Schweber, M. K. Sundaresen, W. M. Visscher, and H. A. Bethe, 'Meson-Nucleon Scattering in the Tamm-Dancoff Approximation', Phys. Rev. 95, 1644 (1954), DOI 10.1103/PhysRev.95.1644, https://doi.org/10.1103/PhysRev.95.1644
  7. M. Cini and S. Fubini, 'Non perturbation treatment of scattering in quantum field theory', Nuovo Cim. 12, 142 (1954), DOI 10.1007/BF02782911, https://doi.org/10.1007/BF02782911
  8. P. T. Matthews and A. Salam, 'Covariant Fock equations', Proc. R. Soc. Lond. A 221, 128 (1954), DOI 10.1098/rspa.1954.0010, https://doi.org/10.1098/rspa.1954.0010
  9. W. P. Silin and W. J. Fainberg, 'Die Tamm-Dancoff-Methode', Fortschr. Phys. 4, 233 (1956), DOI 10.1002/prop.19560040602, https://doi.org/10.1002/prop.19560040602
  10. A. N. Mitra and R. P. Saxena, 'Nucleon-antinucleon interaction in the new Tamm-Dancoff approximation', Proc. R. Soc. Lond. A 257, 59 (1960), DOI 10.1098/rspa.1960.0134, https://doi.org/10.1098/rspa.1960.0134
  11. W. C. Henneberger, 'Hermiticity of Hamiltonians and existence of eigenstates in soluble field theories', Nuclear Physics (1963), pp. 321-327, DOI 10.1016/0029-5582(63)90097-X, https://doi.org/10.1016/0029-5582(63)90097-X
  12. H. Stumpf, F. Wagner, and F. Wahl, 'Zur Konvergenz der einzeitigen Tamm-Dancoff-Methode beim anharmonischen Oszillator' (On the convergence of the single-time Tamm-Dancoff method for the anharmonic oscillator), Z. Naturforsch. A 19, 1254 (1964), DOI 10.1515/zna-1964-1102, https://doi.org/10.1515/zna-1964-1102
  13. C. Schwartz, 'A study of some approximation schemes in quantum mechanics', Annals of Physics (1965), pp. 277-291, DOI 10.1016/0003-4916(65)90018-7, https://doi.org/10.1016/0003-4916(65)90018-7
  14. B. Jouvet and E. Tirapegui, 'Zero-dimensional physics, stairs, continued fractions and convergence problems', Il Nuovo Cimento A (1968), pp. 69-106, DOI 10.1007/BF02824925, https://doi.org/10.1007/BF02824925
  15. C. A. Uzes and J. H. Henkel, 'Level-Dependent New Tamm-Dancoff Method', Phys. Rev. D 8, 1067 (1973), DOI 10.1103/PhysRevD.8.1067, https://doi.org/10.1103/PhysRevD.8.1067
  16. R. J. Perry, A. Harindranath, and K. G. Wilson, 'Light-Front Tamm-Dancoff Field Theory', Phys. Rev. Lett. 65, 2959 (1990), DOI 10.1103/PhysRevLett.65.2959, https://doi.org/10.1103/PhysRevLett.65.2959
  17. B. van de Sande and S. Pinsky, 'Renormalization of Tamm-Dancoff Integral Equations', Phys. Rev. D 46, 5479 (1992), arXiv:hep-th/9210147, DOI 10.1103/PhysRevD.46.5479, https://arxiv.org/abs/hep-th/9210147
  18. S. Pinsky and B. van de Sande, 'Spontaneous symmetry breaking of (1+1)-dimensional φ^4 theory in light-front field theory (II)', Phys. Rev. D 49, 2001 (1994), arXiv:hep-ph/9310330 (1993), DOI 10.1103/PhysRevD.49.2001, https://arxiv.org/abs/hep-ph/9310330
  19. K. G. Wilson, T. S. Walhout, A. Harindranath, W.-M. Zhang, R. J. Perry, and S. D. Glazek, 'Nonperturbative QCD: A Weak-Coupling Treatment on the Light Front', Phys. Rev. D 49, 6720 (1994), arXiv:hep-th/9401153, DOI 10.1103/PhysRevD.49.6720, https://arxiv.org/abs/hep-th/9401153
  20. E. A. Ammons, 'A Wilson Renormalization Group Approach to Light-Front Tamm-Dancoff Scalar Field Theory', Phys. Rev. D 50, 980 (1994), arXiv:nucl-th/9407028, DOI 10.1103/PhysRevD.50.980, https://arxiv.org/abs/nucl-th/9407028
  21. K. Harada, A. Okazaki, and M. Taniguchi, 'Six-body Light-Front Tamm-Dancoff approximation and wave functions for the massive Schwinger model', Phys. Rev. D 52, 2429 (1995), arXiv:hep-th/9502102, DOI 10.1103/PhysRevD.52.2429, https://arxiv.org/abs/hep-th/9502102
  22. K. Harada and A. Okazaki, 'Perturbative Tamm-Dancoff Renormalization', Phys. Rev. D 55, 6198 (1997), arXiv:hep-th/9610020 (1996), DOI 10.1103/PhysRevD.55.6198, https://arxiv.org/abs/hep-th/9610020
  23. A. H. Rezaeian and N. R. Walet, 'Linked-cluster Tamm-Dancoff Field Theory', Phys. Lett. B 570, 129 (2003), arXiv:hep-th/0306159, DOI 10.1016/j.physletb.2003.07.005, https://arxiv.org/abs/hep-th/0306159
  24. V. A. Karmanov, Y. Li, A. V. Smirnov, and J. P. Vary, 'Nonperturbative solution of scalar Yukawa model in two- and three-body Fock space truncations', arXiv:1610.03559 (2016), https://arxiv.org/abs/1610.03559
  25. S. Głazek, 'Renormalization group approach to quantum Hamiltonian dynamics', Int. J. Mod. Phys. A 30, 1530023 (2015), DOI 10.1142/S0217751X15300239, https://doi.org/10.1142/S0217751X15300239
  26. W. Heisenberg, 'Bemerkungen zur Neuen Tamm-Dancoff-Methode', in: Scientific Review Papers, Talks, and Books / Wissenschaftliche Übersichtsartikel, Vorträge und Bücher (Springer, 1984), pp. 901-903, DOI 10.1007/978-3-642-61742-3_66, https://doi.org/10.1007/978-3-642-61742-3_66
  27. G. Grimm, B. Hailer, and H. Stumpf, 'State Representations for Renormalized Energy Equations in Quantum Field Theory', Annalen der Physik (Leipzig) 503, 327 (1991), DOI 10.1002/andp.19915030504, https://doi.org/10.1002/andp.19915030504
  28. R. H. Dalitz and F. J. Dyson, 'Renormalization in the New Tamm-Dancoff Theory of Meson-Nucleon Scattering', Phys. Rev. 99, 301 (1955), DOI 10.1103/PhysRev.99.301, https://doi.org/10.1103/PhysRev.99.301