Complete fixed-point structure and global renormalization-group flow of the non-abelian Thirring model (the current-current-deformed Wess-Zumino-Witten theory)

Non-abelian bosonization (E. Witten, 1984) identifies free massless fermions in two dimensions carrying a compact symmetry group $G$ with the $G$-valued sigma model endowed with a Wess-Zumino term at integer level $k$, at a critical coupling where the bosonic theory becomes the Wess-Zumino-Witten (WZW) conformal field theory. Witten's derivation of the critical coupling was a one-loop renormalization-group (RG) argument reliable only at large level, and the paper left the global RG structure undetermined: whether additional nontrivial fixed points exist, and what happens at couplings beyond the neighborhood of the critical point. The open problem addressed here is the complete fixed-point inventory and global RG flow of the current-current deformation of the WZW model — equivalently the non-abelian Thirring model — for arbitrary compact group, level, and coupling: the fifty-year-old Dashen-Frishman question of an additional fixed point at coupling of order $1/c_G$ ($c_G$ the quadratic Casimir of the adjoint representation of $G$), the strong-coupling infrared behavior, and the conjectured exact truncation of the beta function in the 't Hooft limit. Exact all-loop results exist only in the semiclassical regime $c_G/k \ll 1$, and recent large-$c_G$ conformal perturbation theory reaches only quartic order in the 't Hooft coupling, so the problem remains open.

Background

Setting. In two space-time dimensions, fermionic field theories admit exact rewriting rules ('bosonization') mapping fermion bilinears to bosonic operators with identical correlation functions. The non-abelian version was established by E. Witten in 1984, building on A. M. Polyakov and P. B. Wiegmann's work on two-dimensional non-abelian Goldstone bosons: $n$ species of free massless Dirac fermions transforming under a compact group $G$ are equivalent to the $G$-valued sigma model in two dimensions supplemented by a topological Wess-Zumino term (a quantized term constructed by extending the field $g$ into an auxiliary third dimension), provided the Wess-Zumino coefficient is the integer $n$ (the 'level' $k$) and the coupling takes the critical value $\lambda^2 = 4\pi/n$. At this value the bosonic theory is a conformal field theory, the Wess-Zumino-Witten (WZW) model; its conserved currents generate two chiral Kac-Moody (affine Lie) algebras at level $k$, and these currents are precisely the bosonized fermion bilinears. The WZW model was subsequently developed into an exactly solvable conformal field theory: V. G. Knizhnik and A. B. Zamolodchikov derived the differential equations for its conformal blocks, D. Gepner and E. Witten developed the arbitrary-level theory on group manifolds, and H. Leutwyler and M. Shifman systematized its perturbation theory.

The deformation and its RG flow. Adding to the WZW action the current-current operator $\mathcal{O} = J^a \bar J^a$ (the product of left- and right-moving currents, summed over the adjoint index $a$) with coupling $\lambda$ yields, by non-abelian bosonization, the non-abelian Thirring model: fermions with a four-fermion current-current interaction. (The deformation parameter $\lambda$ of this modern convention vanishes at the WZW point; it is a different coupling from the sigma-model coupling $\lambda^2$ of the original paper, whose critical value there is $4\pi/n$.) The renormalization-group flow of $\lambda$ (its 'beta function' $\beta(\lambda) = d\lambda/d\ln\mu^2$, with $\mu$ the energy scale) determines the infrared (long-distance) fate of the theory: points where $\beta = 0$ are scale-invariant limits ('fixed points'), and A. B. Zamolodchikov's c-theorem supplies a function (the C-function) that decreases monotonically along any two-dimensional RG flow. Witten's argument for the critical coupling was one-loop and reliable only at large level $|k|$ (where the whole action at the fixed point is multiplied by a large parameter, as recalled in the modern literature by A. Nahum's citation of the result); whether additional fixed points or more complicated infrared behavior exist at generic $k$ and strong coupling was left open. On the fermionic side, R. F. Dashen and Y. Frishman had claimed in 1973 an additional scale-invariant point of the $u(N)$-symmetric Thirring model at coupling of order $1/N$, in bosonic language order $1/c_G$; here $c_G$ denotes the quadratic Casimir of the adjoint representation of $G$, defined by $f^{acd}f^{bcd}=c_G\,\delta^{ab}$, with $c_G=2N$ for $SU(N)$ in the normalization standard in this literature. This claim was contested by C. Destri and H. J. de Vega.

Established partial results. D. Kutasov found a weak-strong coupling duality of the non-abelian Thirring model; D. Karabali, Q.-H. Park and H. J. Schnitzer proved exact bose-fermi equivalence with Thirring interactions at large $N$ without settling the fixed-point question. O. A. Soloviev's affine-Virasoro construction (1993) argued for the existence of the Dashen-Frishman conformal points, but it is not a controlled construction of a scale-invariant point at specified $G$ and $k$, and it does not settle the fixed-point question either. The modern integrability literature gives the strongest control: K. Sfetsos's $\lambda$-deformation (the all-loop effective action of the deformed WZW model) and the analysis of G. Itsios, K. Sfetsos and K. Siampos show that in the semiclassical regime $c_G/k \ll 1$ the exact beta function is $\beta(\lambda) = -(c_G/2k)\,\lambda^2/(1+\lambda)^2$, describing a flow from the WZW ultraviolet fixed point at $\lambda = 0$ toward $\lambda \to 1$, together with the duality $\lambda \to 1/\lambda$ with $k \to -k$ (shown in that work to leading order, and established in the exact form $k \to -k - c_G$ by G. Georgiou, E. Sagkrioti, K. Sfetsos and K. Siampos (2020)). R. Corso B. Santos, C. A. Hernaski and P. R. S. Gomes (2023) showed within the large-level expansion that the bosonized theory exhibits three fixed points of which only two are physical, and that the infrared fixed point is strictly unachievable along the RG flow because of an emergent gauge invariance there, with an associated discontinuity in the C-function. S. Li and K. Siampos (2026) computed in the complementary regime $c_G \gg k$, by conformal perturbation theory in the 't Hooft coupling $\tilde\lambda = (c_G/k)\lambda$ (the limit $\lambda \to 0$, $c_G/k \to \infty$ at fixed $\tilde\lambda$), that no additional fixed point of order $k/c_G$ in the deformation parameter $\lambda$ (equivalently, at 't Hooft coupling $\tilde\lambda$ of order one) exists up to quartic order in $\tilde\lambda$; the Dashen-Frishman point is quoted in the same parameter $\lambda$, at order $1/c_G$. Neither analysis determines the global structure at arbitrary group, level, and coupling.

Problem Statement

Determine the complete fixed-point structure and global renormalization-group flow of the two-dimensional current-current deformation of the Wess-Zumino-Witten model, $S = S_{\mathrm{WZW},k}(g) - (\lambda/\pi)\int d^2z\, J^a \bar J^a$ — equivalently, by non-abelian bosonization, of the non-abelian Thirring model — for arbitrary compact Lie group $G$, integer level $k$, and coupling $\lambda$, over the whole coupling space including the strong-coupling region beyond the neighborhood of the conformal points. The problem has three linked deliverables sharing the single objective of complete quantitative control of this RG flow; deliverable (3) feeds (1) and (2), since an exact (provably truncated) beta function fixes both the fixed-point inventory and the flow. (1) Fixed-point inventory. Establish whether the WZW point $\lambda = 0$ (with its free-fermion equivalents where a free-fermion realization exists, and its images under the duality) is the unique nontrivial RG fixed point, or whether additional scale-invariant points exist — in particular settle the Dashen-Frishman question of a fixed point at coupling of order $1/c_G$. Settling the fixed-point inventory for general compact $G$ and integer $k$ requires either an explicit construction of such a fixed point at any specific $G$ and $k$ with $c_G \neq 0$ (with its operator spectrum; the exactly marginal abelian case $c_G = 0$ is excluded), or a proof of absence for all compact $G$ and all $k$. (2) Global flow and strong coupling. Determine the infrared fate of the flow at strong coupling: whether the flow approaches the $\lambda \to 1$ limit suggested by the semiclassical analysis, whether the singularities at $\lambda = \pm 1$ are artifacts of the $c_G/k \ll 1$ expansion, how the flow terminates in view of the strictly unachievable infrared fixed point identified at large level, and what the complete flow diagram (stable and unstable manifolds, duality orbits, endpoints) is for arbitrary $G$ and $k$. (3) Exact truncation and operator algebra. Prove or refute the conjecture that in the 't Hooft limit $c_G \gg k$ the exact beta function and anomalous dimensions truncate at finite order (currently verified only to quartic order in the 't Hooft coupling $\tilde\lambda = (c_G/k)\lambda$), and, if the conjecture holds, determine the symmetry algebra — operator product expansions (the algebras governing products of local operators at short distance), three-point functions, and equal-time current commutators at large $c_G$ — that would underlie such a truncation. Results confined to the semiclassical regime $c_G/k \ll 1$, to finite perturbative order at large $c_G$, or to the large-level limit are already in the literature; a determination valid for arbitrary $G$, $k$, and coupling is required, or a rigorous argument reducing the general case to the established regimes.

References

1. E. Witten, 'Non-abelian bosonization in two dimensions', Commun. Math. Phys. 92 (1984) 455-472. DOI: 10.1007/BF01215276; https://doi.org/10.1007/BF01215276 (INSPIRE record: https://inspirehep.net/literature/14087)
2. R. F. Dashen and Y. Frishman, 'Thirring model with U(n) symmetry: scale invariant only for fixed values of a coupling constant', Phys. Lett. B 46 (1973) 439-442. DOI: 10.1016/0370-2693(73)90161-5; https://doi.org/10.1016/0370-2693(73)90161-5
3. A. M. Polyakov and P. B. Wiegmann, 'Theory of nonabelian Goldstone bosons in two dimensions', Phys. Lett. B 131 (1983) 121-126. DOI: 10.1016/0370-2693(83)91104-8; https://doi.org/10.1016/0370-2693(83)91104-8
4. A. M. Polyakov and P. B. Wiegmann, 'Goldstone fields in two dimensions with multivalued actions', Phys. Lett. B 141 (1984) 223-228. DOI: 10.1016/0370-2693(84)90206-5; https://doi.org/10.1016/0370-2693(84)90206-5
5. V. G. Knizhnik and A. B. Zamolodchikov, 'Current algebra and Wess-Zumino model in two dimensions', Nucl. Phys. B 247 (1984) 83-103. DOI: 10.1016/0550-3213(84)90374-2; https://doi.org/10.1016/0550-3213(84)90374-2
6. A. B. Zamolodchikov, 'Irreversibility of the Flux of the Renormalization Group in a 2D Field Theory', JETP Lett. 43 (1986) 730-732 (Pis'ma Zh. Eksp. Teor. Fiz. 43 (1986) 565-567). INSPIRE record: https://inspirehep.net/literature/240292
7. D. Gepner and E. Witten, 'String theory on group manifolds', Nucl. Phys. B 278 (1986) 493-549. DOI: 10.1016/0550-3213(86)90051-9; https://doi.org/10.1016/0550-3213(86)90051-9
8. C. Destri and H. J. de Vega, 'Jacobian subtleties and the complete path-integral bosonization of the non-abelian Thirring model', Phys. Lett. B 208 (1988) 255-260. DOI: 10.1016/0370-2693(88)90426-1; https://doi.org/10.1016/0370-2693(88)90426-1
9. D. Kutasov, 'Duality off the critical point in two-dimensional systems with non-Abelian symmetries', Phys. Lett. B 233 (1989) 369-373. DOI: 10.1016/0370-2693(89)91325-7; https://doi.org/10.1016/0370-2693(89)91325-7
10. D. Karabali, Q.-H. Park and H. J. Schnitzer, 'Thirring interactions, non-abelian bose-fermi equivalences and conformal invariance', Nucl. Phys. B 323 (1989) 572-584. DOI: 10.1016/0550-3213(89)90124-7; https://doi.org/10.1016/0550-3213(89)90124-7
11. H. Leutwyler and M. Shifman, 'Perturbation theory in the Wess-Zumino-Novikov-Witten model', Int. J. Mod. Phys. A 7 (1992) 795-841. DOI: 10.1142/S0217751X92000387; https://doi.org/10.1142/S0217751X92000387
12. O. A. Soloviev, 'Dashen-Frishman conformal points for fermionic Thirring model from affine-Virasoro construction', Mod. Phys. Lett. A 8 (1993) 3071-3078. DOI: 10.1142/S0217732393002026; https://doi.org/10.1142/S0217732393002026
13. J. F. Bennett and J. A. Gracey, 'Three-loop renormalization of the SU(N) non-abelian Thirring model', Nucl. Phys. B 563 (1999) 390-436. DOI: 10.1016/S0550-3213(99)00570-2; https://doi.org/10.1016/S0550-3213(99)00570-2
14. D. B. Ali and J. A. Gracey, 'Four loop wave function renormalization in the non-abelian Thirring model', Nucl. Phys. B 605 (2001) 337-364. arXiv:hep-th/0104207. DOI: 10.1016/S0550-3213(01)00214-0; https://arxiv.org/abs/hep-th/0104207
15. K. Sfetsos, 'Integrable interpolations: From exact CFTs to non-Abelian T-duals', Nucl. Phys. B 880 (2014) 225-246. arXiv:1312.4560. DOI: 10.1016/j.nuclphysb.2014.01.004; https://arxiv.org/abs/1312.4560
16. G. Itsios, K. Sfetsos and K. Siampos, 'The all-loop non-Abelian Thirring model and its RG flow', Phys. Lett. B 733 (2014) 265-269. arXiv:1404.3748. DOI: 10.1016/j.physletb.2014.04.061; https://arxiv.org/abs/1404.3748
17. G. Georgiou, E. Sagkrioti, K. Sfetsos and K. Siampos, 'An exact symmetry in $\lambda$-deformed CFTs', JHEP 01 (2020) 083. arXiv:1911.02027. DOI: 10.1007/JHEP01(2020)083; https://arxiv.org/abs/1911.02027
18. A. Nahum, 'Note on Wess-Zumino-Witten models and quasiuniversality in 2+1 dimensions', Phys. Rev. B 102 (2020) 201116. arXiv:1912.13468. DOI: 10.1103/PhysRevB.102.201116; https://arxiv.org/abs/1912.13468
19. R. Corso B. Santos, C. A. Hernaski and P. R. S. Gomes, 'Bosonization, duality, and the C-theorem in the non-abelian Thirring model', JHEP 07 (2023) 172. arXiv:2305.11113. DOI: 10.1007/JHEP07(2023)172; https://arxiv.org/abs/2305.11113
20. K. Önder, 'RG flows in 2d chiral gauge theories', Phys. Rev. D 108 (2023) 085009. arXiv:2305.09054. DOI: 10.1103/PhysRevD.108.085009; https://arxiv.org/abs/2305.09054
21. S. Li and K. Siampos, 'Non-Abelian Thirring model at large N', arXiv:2607.17246 (2026); https://arxiv.org/abs/2607.17246
22. S. Onoda, 'Lattice chiral non-Abelian gauge symmetry via bosonization', arXiv:2606.12358 (2026); https://arxiv.org/abs/2606.12358