The direction-of-transmission inference turns on three coincident facts. (1) Timing: the NB1 case was diagnosed 28 April, after the mink outbreak (19-20 April) was established, so exposure occurred during peak animal shedding rather than before. (2) Exposure route: viral RNA was present in airborne inhalable dust inside the sheds during exactly this window, giving a concrete high-dose aerosol route from mink to a person working among them. (3) Genotype: preliminary sequencing placed the worker’s virus with the mink-farm lineage rather than an unrelated community lineage — a virus nested within the animal clade points to the animal as source, since a community-acquired infection would be expected to carry an independent human genotype. None of the three alone is decisive (the sequence evidence is preliminary and not fully presented), but jointly they make mink-to-human transmission the better-supported explanation than re-acquisition from the community.
Step 6 verdict — approved (checked)
Reconstruction. Premises: (i) the NB1 worker was diagnosed 28 April, after the mink outbreak (19-20 April) was established, i.e. during peak animal shedding; (ii) SARS-CoV-2 RNA was present in airborne inhalable shed dust in that window, a concrete high-dose aerosol route; (iii) preliminary sequencing nested the worker’s virus within the mink-farm lineage rather than an unrelated community lineage. Load-bearing step: a likelihood comparison — the three facts are jointly more expected if mink infected the worker than if the worker was independently community-infected.
Evaluation. The move is a valid abductive/likelihood inference. Premise (iii) is the fulcrum: an independent community infection would carry an independent community genotype, so a virus nested in the animal clade is more expected under mink-to-human than under community re-acquisition; (i) and (ii) supply the timing and route that make the animal route physically available. I checked the natural direction-reversal defeater — the worker infected the mink (human-to-mink), which would also match genotypes — but premise (i) blocks the worker being the per-farm index (illness onset after the outbreak was already underway); and a further-upstream human who caught it from mink still routes back to a mink source. The statement is correctly hedged to “more plausibly / better-supported,” not decisive (it flags the sequence data as preliminary), so no surviving defeater forces a weaker form. Traced myself; author-blind.