Bootstrap classification of Regge-slope spectral densities for unitary dual-resonance S-matrices

Tree-level dual-resonance (Veneziano-type) amplitudes realize crossing symmetry, Regge behavior, and infinite resonance towers in closed analytic form, but they fail exact partial-wave unitarity because their absorptive parts live only on isolated poles. de Rham, Tolley, Wang, and Zhou (arXiv:2607.24922, 2026) constructed the first genuinely unitary dual-resonance S-matrices by smearing Veneziano kernels over a continuous Regge slope with a spectral density $\phi(y)$, and posed as the central open problem the bootstrap classification of the admissible smearing data. This record states that source-faithful problem: characterize the Regge-slope spectral densities $\phi(y)$ compatible with crossing symmetry, analyticity, locality, and full partial-wave unitarity. The same Discussion notes related facets of the program — the singularity structures realizable within the admissible class (more general smearing data can generate multiple families of second-sheet Regge trajectories and nonresonant absorptive backgrounds), the various Regge cuts in the complex angular-momentum plane that such data induce, and the relation of Regge-slope smearing to standard eikonalization — with the characterization question as the posed open problem. The problem is open: the source provides a single explicit family in $D=10$, and no admissibility characterization, multi-trajectory construction, Regge-cut classification, or smearing-eikonal analysis exists.

Background

Scattering amplitudes at strong coupling are constrained by a small set of first principles, and the modern S-matrix bootstrap program (Paulos, Penedones, Toledo, van Rees, and Vieira, JHEP 12, 040 (2017)) seeks to chart the space of two-to-two amplitudes satisfying all of them simultaneously. For four-point scattering of identical scalars the principles are: crossing symmetry, meaning the amplitude is a single analytic function of the Mandelstam invariants $s$, $t$, $u$ (the Lorentz-invariant kinematic variables built from the external four-momenta, with $s+t+u$ equal to the sum of the external squared masses, hence zero for massless scalars) invariant under their permutation, so that one function describes all channels; analyticity in the complex $s$ and $t$ planes away from physical cuts and poles; locality, imposed here as polynomial boundedness at large $s$ for fixed $t<0$ (the property underlying dispersion relations and encoding causal propagation); and unitarity. In the partial-wave expansion $M(s,t)=\sum_\ell (2\ell+1) a_\ell(s) P_\ell(\cos\theta)$, where $P_\ell$ are Legendre polynomials and $\theta$ the scattering angle (written in $D=4$ form; in general $D$ the Legendre polynomials are replaced by the Gegenbauer partial waves $P_\ell^{(D)}(\cos\theta)$ used in the source paper, whose normalization defines $a_\ell(s)$), exact (full) partial-wave unitarity for identical scalars is the nonlinear inequality $|a_\ell(s)|^2 \le 2\,\mathrm{Im}\, a_\ell(s)$ for every allowed spin $\ell$ (even, for identical scalars) and every $s>0$ (normalization as in the source paper); since $\mathrm{Im}\, a_\ell > 0$ away from thresholds, a unitary amplitude must possess absorptive branch cuts in $s$, not merely isolated poles.

Dual-resonance models are the oldest analytic realization of crossing together with Regge behavior. Veneziano's 1968 amplitude $V(s,t)\propto \Gamma(-\alpha' s)\Gamma(-\alpha' t)/\Gamma(-\alpha' s-\alpha' t)$, with $\alpha'$ the Regge slope (the inverse string tension; a linearly rising Regge trajectory $\alpha(x)=\alpha' x$ organizes the resonances by spin), is crossing-symmetric, behaves as $M\sim s^{\alpha(t)}$ at large $s$ and fixed $t$ (Regge behavior), and exhibits an infinite tower of evenly spaced zero-width resonances — the tree-level open string. Its fundamental shortcoming is unitarity: as a meromorphic function (analytic except for poles), its absorptive part is supported only on the poles, so full partial-wave unitarity fails. The same holds for later meromorphic generalizations such as the Coon amplitude (a $q$-deformed Veneziano amplitude with geometrically spaced resonance poles) and the bespoke dual-resonance amplitudes of Cheung and Remmen (Phys. Rev. D 108, 086009 (2023)), and for the Veneziano amplitude itself, whose uniqueness within a meromorphic dual bootstrap was proven analytically by Wan and Zhou (arXiv:2605.11084, 2026) from dispersive sum rules (positivity constraints obtained by writing dispersion relations for the amplitude) plus string monodromy or splitting input. The meromorphic (zero-width) side of the dual program is therefore comparatively well understood; the unitary, non-meromorphic side was empty until 2026.

de Rham, Tolley, Wang, and Zhou (arXiv:2607.24922, 2026) filled this gap with an explicit construction: smear the triple-crossing-symmetric sum of Veneziano kernels $W_y(s,t,u)=V_y(s,t)+V_y(u,t)+V_y(s,u)$ over a continuous range of Regge slopes $y\in[\alpha',\infty)$ with a spectral density $\phi(y)$,

$M(s,t)=\int_{\alpha'}^{\infty} dy\,\phi(y)\,W_y(s,t,u).$

The smearing converts the zero-width pole towers into physical absorptive cuts while preserving the underlying crossing and Regge organization, and unitarizes the amplitude through a mechanism distinct from eikonalization (the standard high-energy, small-angle resummation, mandatory for long-range gravitational exchange, that restores unitarity through impact-parameter profiles, i.e. the transverse-distance dependence of the scattering). For one explicit family of positive weights — with large-$y$ tail $\phi(y)\sim y^{-p}$, $1<p\le 2$, and a pair of complex-conjugate poles of $\phi$ in the $y$-plane at $y_0\pm i\Gamma$ — in $D=10$ spacetime dimensions with identical massless scalars, they show: the $y$-plane poles generate one family of second-sheet (i.e. continued through the unitarity cut) Regge trajectories of resonances with finite decay widths, while holomorphic, singularity-free pieces of $\phi$ would generate nonresonant absorptive backgrounds; analyticity and locality hold, established through Landau-singularity/pinch analysis (locating the kinematic configurations where the $y$-integration contour is pinched between singularities of the integrand, which is where the smeared amplitude could become singular) and monodromy analysis (tracking how the integral changes when the kinematic variables wind around the integrand's branch points) of the smearing integral; and exact partial-wave unitarity holds in the asymptotic regimes accessible analytically — e.g. the threshold analysis forces $p\le (D-4)/2-1$, i.e. $p\le 2$ in $D=10$ — and is verified numerically on sampled finite-$(s,\ell)$ grids. What the construction does not provide is any characterization of the admissible class of smearing data; the source poses exactly this as the open bootstrap program.

Problem Statement

The problem, as posed by the Discussion of arXiv:2607.24922, is to turn the Regge-slope smearing construction into a bootstrap classification of unitary dual-resonance S-matrices. Concretely: characterize the class of Regge-slope spectral densities $\phi(y)$ — by necessary and sufficient conditions — for which the smeared dual amplitude

$M(s,t)=\int_{\alpha'}^{\infty} dy\,\phi(y)\,W_y(s,t,u)$

(with $W_y(s,t,u)=V_y(s,t)+V_y(u,t)+V_y(s,u)$ and $V_y(s,t)\propto \Gamma(-ys)\,\Gamma(-yt)/\Gamma(-ys-yt)$ the Veneziano kernel of Regge slope $y$, in the normalization of arXiv:2607.24922; setting of the source: four-point scattering of identical massless scalars in $D$ spacetime dimensions, $D=10$ as the concrete baseline) satisfies simultaneously: (1) crossing symmetry; (2) analyticity in the maximal-analyticity sense, i.e. analyticity in the complex $s$ and $t$ planes away from physical cuts and poles; (3) locality, as polynomial boundedness at large $s$ for fixed $t<0$; and (4) full unitarity, i.e. exact partial-wave unitarity $|a_\ell(s)|^2\le 2\,\mathrm{Im}\,a_\ell(s)$ for all $s>0$ and all even $\ell$ (with $a_\ell(s)$ the partial-wave coefficients in the normalization of the source paper, in which identical scalars project onto even partial waves).

The overarching objective — charting the admissible space of consistent strongly coupled S-matrices of dual-resonance type — is posed by the source as the characterization (i) of the admissibility conditions on the smearing data $\phi(y)$, holding simultaneously across the threshold, Regge/asymptotic, and finite-$(s,\ell)$ regimes. The same Discussion notes three related facets of the program, as directions that more general smearing data can realize rather than as jointly required deliverables: (ii) explicit constructions of the singularity structures the admissible class can realize, in particular multiple families of second-sheet Regge trajectories and nonresonant absorptive backgrounds, beyond the single-trajectory examples of the source; (iii) the classification of the Regge cuts (branch cuts, as opposed to isolated Regge poles, in the complex angular-momentum plane) these data induce, with their physical interpretation in strongly coupled theories; and (iv) the relation between Regge-slope smearing and standard eikonalization as crossing-symmetric unitarization mechanisms. Component (i) is the core on which (ii)-(iv) hinge: it fixes the admissible class whose singularity structure and mechanism-comparison the rest of the program describes. The $D$-dependence of the admissible class (including $D=4$, relevant for hadron scattering) is part of the program's natural scope, with the source's $D=10$ analysis as baseline: a necessary-and-sufficient characterization of $\phi(y)$ at $D=10$ resolves (i), while the extension to general $D$ (including $D=4$) is optional additional scope. Progress may be reported on any of the related facets (ii)-(iv); the classification question (i) is the definitive statement of the problem. Resolving (i) requires conditions on $\phi(y)$ that are both necessary and sufficient for (1)-(4); a necessary-only or sufficient-only result counts as partial progress on (i), not as its resolution.

References

1. C. de Rham, A. J. Tolley, Z.-H. Wang, and S.-Y. Zhou, Unitary Dual-Resonance S-matrices, arXiv:2607.24922 [hep-th] (2026), DOI: 10.48550/arXiv.2607.24922, https://arxiv.org/abs/2607.24922
2. G. Veneziano, Construction of a crossing-simmetric, Regge-behaved amplitude for linearly rising trajectories (title as published; the misspelling is in the original 1968 publication), Il Nuovo Cimento A 57, 190-197 (1968), DOI: 10.1007/BF02824451, https://doi.org/10.1007/BF02824451
3. C. Cheung and G. N. Remmen, Bespoke dual resonance, Physical Review D 108, 086009 (2023), DOI: 10.1103/PhysRevD.108.086009, https://doi.org/10.1103/PhysRevD.108.086009 (arXiv:2308.03833)
4. S.-L. Wan and S.-Y. Zhou, Analytic Bootstrap of the Veneziano Amplitude, arXiv:2605.11084 [hep-th] (2026), https://arxiv.org/abs/2605.11084
5. L.-Q. Shao and A. Vichi, Analytic Boundaries of Infinite-Spin-Tower Amplitudes from Hidden Zero, arXiv:2607.27300 [hep-th] (2026), https://arxiv.org/abs/2607.27300
6. J. Jeong, Partial Waves for Multipositivity, arXiv:2608.02719 [hep-th] (2026), https://arxiv.org/abs/2608.02719
7. M. F. Paulos, J. Penedones, J. Toledo, B. C. van Rees, and P. Vieira, The S-matrix Bootstrap III: Higher Dimensional Amplitudes, JHEP 12, 040 (2017), arXiv:1708.06765 [hep-th], https://arxiv.org/abs/1708.06765
8. Q. Chen and Y. Yin, Exact All-Level Positivity of the Bosonic Veneziano Amplitude, arXiv:2608.25183 [hep-th] (2026), https://arxiv.org/abs/2608.25183
9. H. Gomez, R. L. Jusinskas, and S. P. Kashyap, A field-inspired derivation of open string amplitudes, arXiv:2608.02754 [hep-th] (2026), https://arxiv.org/abs/2608.02754