The merger record contradicts itself and the model: spectroscopic pair fractions decline beyond z of 2 to 3 (MUSE), JWST morphology finds settled disks dominating to z of 6+ with merger signatures far below expectation (Ferreira et al. 2024), while the latest JWST pair counts rise to z of 8 then plateau (Duan et al. 2025). Probes that should agree under hierarchical assembly tell three stories, none matching the simulated tempo.
The model assembles mass by merging, so merger activity must rise steeply into the past and leave fingerprints everywhere: disturbed morphologies, hot kinematics, climbing pair fractions. Early galaxies that are massive yet settled force the model to grow mass through a process the images keep failing to catch in the act.
SCT replaces hierarchical growth with collision seeding: proto-structures condense at masses set by deposition geometry (P56), with morphology and angular momentum fixed at birth by the impact parameter and inherited J (P31, P32). Mergers still occur, but they are secondary traffic among already-formed structures rather than the engine of mass growth, so the model expects exactly the contradictory record observed: modest pair fractions without a steep hierarchical climb, settled disk morphologies dominating early because disks are the birth state, and massive quiet galaxies at z of 6 whose mass arrived by deposition rather than disruption.
The probe disagreements dissolve into measurement of different things. Pair counts at high z partially trace sibling proto-structures condensing from common debris, not orbit-decaying mergers, which is why pair statistics and morphological disturbance decouple; and merger observability tuning becomes unnecessary once mergers are no longer required to carry 70 percent of mass growth. The fingerprint absence is the prediction: seeded galaxies wear their birth geometry (P22, P25), not their collision history with neighbors.
This is the same seeding physics behind the z = 14 massive galaxies, the early disks, and the morphological maturity JWST keeps finding. There is no need to invoke per-epoch merger timescale tuning to hide the assembly line.
The discriminating statistic: under seeding, mass growth must be dominated by in-place formation, so deep kinematic samples should show massive early galaxies with intact, inherited rotation rather than merger-built dispersion. If uniform JWST and Euclid pipelines instead confirm pair fractions, morphological disturbance, and merger-driven mass accretion all rising together at the simulated hierarchical tempo to z of 10, with mergers verified to supply the majority of stellar mass growth, the seeded-assembly reading fails.