Galaxy spins are webbed into the filaments: low-mass galaxies align their spin axes parallel to their host filament, high-mass perpendicular, with alignment amplitudes and coherence lengths exceeding tidal torque predictions, part of the spin-correlation family that overshoots the theory by factors of 10 to 20 at 30 to 100 Mpc. Filaments themselves rotate, solid-body-like at 110 km/s (Tudorache et al.), giving the web an angular momentum budget torquing struggles to supply.
In ΛCDM, spin is acquired late and locally through tidal torques during collapse, a perturbative mechanism with fast-decaying correlations, further randomized by mergers. The model has no channel by which spins across tens of megaparsecs, and the filament hosting them, can share an axis: large-scale spin coherence has nowhere to come from.
SCT replaces tidal acquisition with inheritance: a collision with impact parameter deposits angular momentum J = μ(b × v_rel) into the debris field (P31), and every structure condensing from that field, the filament and the galaxies strung along it, inherits a share of the same vector, with the scaling J proportional to M^(5/3) observed across seven decades of mass (P32). Spin-filament alignment is then not a correlation to be generated but a family resemblance: filament rotation and galaxy spins are partitions of one parent J, which is why the alignment is coherent over the filament's full length and present from the earliest epochs.
The mass transition falls out of the partition dynamics: low-mass galaxies condense from the stream's bulk shear and carry the parent axis directly (parallel), while the most massive systems assemble at nodes where flows from multiple directions meet, their final spin set by the collision of inheritances rather than any single stream (perpendicular tendency). Rotating filaments stop being a puzzle and become the parent reservoir made visible, the smoking gun the inheritance picture requires and torque theory must explain away.
This is the same J inheritance behind satellite-plane co-rotation, cluster rotation scaling, and the quasar polarization alignments. There is no need to invoke spin-flip bookkeeping or anisotropic-accretion patches on a mechanism whose amplitude is short by an order of magnitude.
The registered kills: a systematic MeerKAT/SKA survey of fifty or more filaments finding bulk rotation amplitudes and member spin alignments consistent with IllustrisTNG-level tidal torque predictions removes the inherited excess. Structurally, the inheritance gradient must be present: spin coherence strongest along a filament's main strand, weaker in branches, most disordered at junctions where inherited vectors compete; finding no such gradient in well-resolved systems falsifies the partition picture.