Spin-Waves without Spin-Waves: A Case for Soliton Propagation in Starling Flocks

Andrea Cavagna, Guido Cimino, Javier Cristín, Matteo Fiorini, Irene Giardina, Angelo Giustiniani, Tomás S. Grigera, Stefania Melillo, Roberto A. Palombella, Leonardo Parisi, Antonio Ponno, Mattia Scandolo, Zachary S. Stamler · 2025

Read Spin-Waves without Spin-Waves: A Case for Soliton Propagation in Starling Flocks by Andrea Cavagna, Guido Cimino, Javier Cristín, Matteo Fiorini, Irene Giardina, Angelo Giustiniani, Tomás S. Grigera, Stefania Melillo, Roberto A. Palombella, Leonardo Parisi, Antonio Ponno, Mattia Scandolo, Zachary S. Stamler (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-19665.txt.

Introduction

When a starling flock turns, the turn sweeps through the birds like a wave that barely fades. Andrea Cavagna and co-authors wrestle with a puzzle in that motion. Theory says such crisp waves should come with a wavy pattern in the flock's correlations, yet sharp new measurements show a smooth, fading pattern instead.

The two facts seem to clash: underdamped waves in the motion, overdamped shape in the correlations. The authors resolve the clash by adding a classic nonlinear term, of Fermi-Pasta-Ulam-Tsingou type, to the flock dynamics. With that term, the model produces both behaviors at once. The paper argues that what travels through the flock behaves less like a spin wave and more like a soliton, a self-holding pulse. The result reframes how collective turns carry information.

Catalog record

CREATORAndrea Cavagna, Guido Cimino, Javier Cristín, Matteo Fiorini, Irene Giardina, Angelo Giustiniani, Tomás S. Grigera, Stefania Melillo, Roberto A. Palombella, Leonardo Parisi, Antonio Ponno, Mattia Scandolo, Zachary S. Stamler
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE64 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-19665

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Spin-Waves without Spin-Waves: A Case for Soliton Propagation in Starling Flocks Andrea Cavagna, Guido Cimino, Javier Cristín, Matteo Fiorini, Irene Giardina, Angelo Giustiniani, Tomás S. Grigera, Stefania Melillo, Roberto A. Palombella, Leonardo Parisi, Antonio Ponno, Mattia Scandolo, Zachary S. Stamler Abstract Collective turns in starling flocks propagate linearly with negligible attenuation, indicating the existence of an underdamped sector in the dispersion relation. Beside granting linear propagation of the phase perturbations, the real part of the frequency should also yield a spin-wave form of the unperturbed correlation function. However, new high-resolution experiments on real flocks show that underdamped traveling waves coexist with an overdamped Lorentzian correlation. Theory and experiments are reconciled once we add to the dynamics a Fermi-Pasta-Ulam-Tsingou term. Experiments on starling flocks [ 1 ] show that collective turns propagate across the group as linear waves with very little damping (Videos S1-S2) and with a speed of propagation that is higher the larger the polarization; these waves are not accompanied by significant density fluctuations, as the interaction network remains nearly unchanged during the turn. Moreover, birds turn along equal-radius – rather than parallel – trajectories, which allows them to keep their speed within a physiological range [ 2 ] . To explain this phenomenology it was introduced in [ 1 , 3 ] the Inertial Spin Model (ISM): at equilibrium the ISM belongs to Model G universality class [ 4 ] , hence in its low temperature phase it sustains phase waves even in absence of density fluctuations (second sound). In the ISM the velocity \mathbf{v}_{i} is turned by the spin \mathbf{s}_{i} , while the spin is acted upon by the alignment force, \displaystyle\dot{\mathbf{v}}_{i} \displaystyle=\frac{\partial H}{\partial\mathbf{s}_{i}}\times\mathbf{v}_{i}\ , \displaystyle\dot{\mathbf{s}}_{i} \displaystyle=-\mathbf{v}_{i}\times\frac{\partial H}{\partial\mathbf{v}_{i}}-\frac{\eta}{\chi}\mathbf{s}_{i}+\bm{\xi}_{i}\ , (1) where the pseudo-Hamiltonian is given by, H=\sum_{i}\frac{\mathbf{s}_{i}^{2}}{2\chi}+\frac{J}{4v_{0}^{2}}\sum_{ij}n_{ij}(\mathbf{v}_{i}-\mathbf{v}_{j})^{2}\ . (2) The adjacency matrix n_{ij} depends on time due to the self-propulsion equation, \mathbf{r}_{i}=\mathbf{v}_{i} (whence ‘pseudo’); the noise has correlator, \langle\xi_{i}^{\mu}(t)\xi_{j}^{\nu}(t^{\prime})\rangle=2\eta T\,\delta(t-t^{\prime})\,\delta_{ij}\delta^{\mu\nu} , and the velocity modulus is fixed, |\mathbf{v}_{i}|=v_{0} . The parameter J is the alignment coupling constant (or stiffness), while \chi is the (generalized) inertia and \eta is the nonconservative friction. The ISM’s overdamped limit is the Vicsek model [ 5 ] . In the low-temperature symmetry-broken phase the velocities predominantly point in the same mean direction, \mathbf{v}=(1/N)\sum_{i}\mathbf{v}_{i} , and we can expand in the small transverse fluctuations, \bm{\pi}_{i} , writing, \mathbf{v}_{i}=\mathbf{v}_{i}^{L}+\bm{\pi}_{i} , where |\bm{\pi}_{i}|\ll|\mathbf{v}| and |\mathbf{v}_{i}^{L}|=\sqrt{v_{0}^{2}-\bm{\pi}_{i}^{2}} , with \bm{\pi}_{i}\cdot\mathbf{v}=0 [ 6 ] . By expanding linearly in \bm{\pi}_{i} the homogeneous equation of motion for the velocity (see Supplemental Material – SM) we obtain, \chi\,\ddot{\bm{\pi}}_{i}+\eta\,\dot{\bm{\pi}}_{i}+J\sum_{j}\Lambda_{ij}\bm{\pi}_{j}=0\ , (3) where \Lambda_{ij}=-n_{ij}+\delta_{ij}\sum_{k}n_{ik} is the discrete Laplacian. By assuming that n_{ij}(t) varies slowly compared to the velocity’s relaxation time [ 7 ] and by going to the continuum limit, \Lambda_{ij}\to-a^{2}\nabla^{2} (where a is the mean interparticle distance), we can derive the quasi-equilibrium dispersion relation of the ISM, {\omega}(k)=-i\gamma\pm\sqrt{(J/\chi)a^{2}k^{2}-\gamma^{2}}\ , (4) where \gamma=\eta/2\chi is the reduced friction. In the underdamped regime the frequency develops a real part, which for (J/\chi)a^{2}k^{2}\gg\gamma^{2} becomes \omega(k)\approx\pm c_{s}k , with c_{s}=a\,\sqrt{J/\chi} ; hence, the system sustains linearly propagating waves with speed c_{s} . Simulations confirm that within this deeply underdamped phase the change of direction of one particle elicits a collective turn of the entire group, while in the overdamped phase, (J/\chi)a^{2}k^{2}<\gamma^{2} , damping and sub-linear diffusion cause disgregation of the flock, with no collective turning; these results are evident from the acceleration curves of the particles (Fig.1) and even more vividly from the corresponding Videos S3-S4. The speed of propagation in real flocks was found to be in line with the ISM prediction: in the ordered phase the polarization \Phi=|\mathbf{v}|/v_{0} , is connected to the transverse fluctuations by the relation, (1-\Phi)\approx\langle\bm{\pi}^{2}/v_{0}^{2}\rangle/2\sim T/J (the prefactors depend on the specifics of the lattice [ 6 , 8 , 9 ] ), whence we obtain c_{s}\sim 1/\sqrt{1-\Phi} , a prediction in fair agreement with experiments on starling flocks [ 1 ] . Figure 1: Simulations - ISM. Upper panels: underdamped regime. a) The initiator is turned by \pi/2 , eliciting a collective turn: the accelerations of all the particles show coherent turning, with little damping (Video S3). Inset: traveled distance of the perturbation (from an intercept on the phase) vs time indicates linear propagation. b) The velocity correlation function has clear spin-wave peaks at \omega_{SW}=\pm c_{s}k (SM). Lower panels: overdamped regime. c) The initiator fails to elicit a collective turn; the accelerations of the other particles crash due to damping (Video S4). d) The velocity correlation function is Lorentzian, with half-width frequency \omega_{L}\sim k^{2} (SM). Simulations are performed in d=3 . The existence of a real part of the frequency in the underdamped regime has another consequence, namely that the correlation function of the velocity fluctuations develops two spin-wave (SW) peaks at \omega_{\mathrm{sw}}=\pm c_{s}k ; conversely, in the overdamped phase, c_{s}^{2}\,k^{2}<\gamma^{2} , the correlation function is quasi-Lorentzian (L), with half-width frequency, \omega_{\mathrm{L}}=(J/\eta)a^{2}k^{2} . In the Supplemental Material we perform a linear analysis of the out-of-equilibrium active field theory corresponding to ( 1 ), in which n_{ij}(t) is not assumed to be constant, hence including density fluctuations, and show that also the active case conforms to the standard spin-wave phenomenology. Simulations of the ISM confirm this scenario (Fig.1). Note that the association of spin-wave peaks and traveling waves does not depend on the specific way propagating modes emerge, whether it is through inertia, as in the ISM, or through the velocity-density coupling giving rise to sound modes as in the Vicsek model [ 10 , 11 , 12 ] : whenever the frequency develops a real part, there should be both spin-wave/sound propagation and spin-wave/sound peaks; this is the core of linear response theory. Experimental evidence on starling flocks starkly disagrees with this scenario, though. To calculate the correlation function accurately enough to resolve the potential spin-wave peaks one needs a small frequency step, which requires experimental acquisitions with time duration significantly longer than [ 1 ] ; to this end we developed a new panning 3D stereo system and run new experiments (Fig.2a-d and SM). Surprisingly enough, the correlation functions computed from this new dataset (an example in Fig.2f, more in SM) do not show any trace of spin-wave peaks; instead, they have a very much Lorentzian central peak at \omega=0 , with half-width frequency \omega_{\mathrm{L}}\sim k^{2} (SM), both clear hallmarks of an overdamped system. And yet, phase perturbations do propagate linearly across the group, eliciting collective turns (Fig.2e). So, starling flocks develop traveling spin-waves under perturbation, without spin-waves in the spontaneous fluctuations. How is that possible? Figure 2: Experiments - Starling flocks. a-d) In the new experimental setup micrometric rotors (c) are used to pan the stereo camera system and track flocks for longer time intervals (up to 12.5 sec vs 3 sec of the old experiments – SM). e) In turning flocks phase perturbations propagate linearly with negligible dissipation, as in the underdamped phase of the ISM (Fig.1a), and yet: f) The correlation function has a quasi-Lorentzian form, with no spin-wave peaks, and an half-width frequency scaling as \omega_{\mathrm{L}}\sim k^{2} (see SM), as in the overdamped phase of the ISM (Fig.1d). There are two possible ways out of this apparent paradox: either some further out-of-equilibrium effect radically changes the spin-wave prediction, or the relationship between response to

Read the complete text