Cosmic flows are supposed to be curl-free. In standard perturbation theory the large-scale velocity field is irrotational on linear and quasi-linear scales, vorticity being generated only deep in the non-linear regime, in cluster cores and shock fronts, and even there it stays locally confined (Bernardeau et al. 2002; Pueblas and Scoccimarro 2009). Whatever spin galaxies carry is supposed to come from tidal torquing during collapse, a mechanism whose alignments fade beyond a few tens of megaparsecs.
The filaments did not get the memo. Simulations find significant vorticity concentrated along filament spines with halo spins aligned to it (Codis et al. 2015; Wang et al. 2021), and observation has now caught a filament rotating as a body: MeerKAT 21-cm measurements of a 1.7 Mpc filament show coherent solid-body-like rotation at about 110 km/s across its member galaxies (Tudorache et al. 2025). Tidal torque theory cannot produce ordered rotation at that amplitude on that scale, and the related spin-alignment coherences run 10 to 20 times beyond its reach. Within ΛCDM there is no budget for megaparsec-scale bulk angular momentum: the initial conditions are irrotational, gravity conserves circulation, and torquing is too weak and too local.
The standing is young but pointed, one measured filament with more coming. MeerKAT and the SKA will survey filament rotation across large samples, turning a single remarkable detection into a population statistic that either matches simulation noise or establishes bulk angular momentum as a generic property of the cosmic web.