Keep named specimens, feather chemistry, wind-tunnel measurements, phylogenetic topology, molecular clocks and COL26.8 living-name coverage separate while comparing crocodile-line and bird-line evidence.
Each step names its evidence claims, with references below. Interactive Explorer states are available in the full application; the narrative remains an editorial synthesis.
AMNH FR 30587 records an edentulous, gracile suchian. Its ornithomimid-like proportions are a convergence diagnosis, not evidence that the two branches share an immediate ancestor or identical ecology.
Effigia is a pseudosuchian outside Crocodylomorpha, so it remains event evidence rather than a false child of the crocodile navigation node.
claim:event:effigia-convergent-body-planNCSM 21558 documents a large Carnian crocodylomorph. Size and predator tier come from proxies and assemblage context, not preserved feeding behaviour.
Incomplete and immature anatomy cannot establish a direct ancestor or universal global first occurrence.
claim:event:carnufex-holotype-predatorIVPP V14010 combines a consolidated, crocodylian-like skull with a gracile terrestrial postcranium. It is one anatomical mosaic rather than a compulsory stage in a ladder.
Skull mechanics and cursoriality are functional interpretations; topology is character-matrix evidence.
claim:event:junggarsuchus-skull-consolidationA combined morphology, molecular and age model groups Gavialis with Tomistoma while moving sampled thoracosaurs outside the crown. Other analyses differ.
A preferred topology is not an observed genealogy; priors, weights and sampled characters remain visible boundaries.
claim:event:crocodylian-tip-dated-topologyWDC CSG 100 exposes theropod-like pedal and palate features alongside avialan anatomy. The mosaic supports affinity without turning one specimen into the unique first bird.
Joint posture, perching and exact placement are interpretations rather than direct observations.
claim:event:archaeopteryx-thermopolis-skeletonMelanosomes in BMNHC PH828 support a statistical map of plumage regions. Preservation, reference samples and a tail borrowed from another specimen constrain the image.
Anchiornis remains a non-avian paravian event here, not a false child of Aves or a directly observed ancestral colour state.
claim:event:anchiornis-plumage-melanosomesIVPP V13352 preserves long pennaceous feathers on forelimbs, hindlimbs and tail. The fossil records surfaces, not a flight trajectory.
Microraptor is treated as a non-avian paravian dossier, not a direct bird ancestor.
claim:event:microraptor-four-winged-holotypeMeasured forces on a scaled model support stable high-drag gliding across tested configurations. That experiment is functional evidence, not footage of extinct behaviour.
Geometry, scaling, reconstructed posture and fossil distortion bound every aerodynamic conclusion.
claim:event:microraptor-wind-tunnel-modelCT anatomy anchors NHMM 2013 008 near Galloanserae, but parsimony and tip-dated Bayesian analyses place it on different branches.
A tightly dated occurrence is not automatically the globally oldest modern bird, and crown membership remains analysis-dependent.
claim:event:asteriornis-holotype-crown-placementForty-eight representatives across neoavian orders recover major groupings and expose gene-tree conflict. This is sampled extant topology, not a complete species tree.
Incomplete lineage sorting, convergence, sampling and method choice remain explicit.
claim:event:neoavian-genome-topologyClock-like exons, broader RAG sampling and fossils infer a Cretaceous root and boundary-era radiation. The dates follow calibrations, topology and priors.
COL26.8 routes 11,071 accepted living species into this package. That is a nomenclatural coverage boundary, not fossil completeness, direct ancestry or agreement on a single crown-bird date.
claim:event:neornithes-fossil-calibrated-time-tree