For a market-vector hypothesis to hold, the vector species must have been present at the market to contact humans there. A systematic 2.5-year census that recorded 38 species and zero pangolins/bats establishes, to good approximation, that these taxa were not traded and thus could not have transmitted the virus to humans at the market. Scope caveat: this bears only on the market-vector role; it does not exclude pangolins or bats as an upstream natural reservoir or as the source of the virus that entered whichever species (if any) was the actual market vector.
Verdict (step 6) — corrected, checked
Reconstruction. Premises: (i) a species not physically present at the market cannot be the market-vector there (near-analytic); (ii) pangolins and bats appear in zero of the 47,381 recorded traded animals. Load-bearing step is the bridge: from “absent from the recorded stock” (ii) to “actually not physically present,” which is what premise (i) then converts into “cannot be the vector.”
Evaluation. The analytic half is sound. The bridge has a surviving undercutting defeater: a sales census records observed/legal trade, and pangolins are CITES-protected and trafficked illegally — exactly the category a vendor-facing survey would systematically under-record. So recorded absence does not entail actual absence, and the deductive “rules them out / cannot be” overreaches. The bridge does support the weaker evidential conclusion (recorded absence makes physical presence, hence the vector role, unlikely — strongest for bats, which are not sold in southern-China markets, and materially weaker for pangolins given illegal trade). Corrected the statement from strict exclusion to “very unlikely, not strictly ruled out,” which is immune to the defeater. The built-in scope caveat (upstream reservoir untouched) is retained and correct. Traced directly — checked.
Original
Because a species that was never on sale at the markets cannot be the vehicle by which a virus reached humans at those markets, the complete absence of pangolins and bats from 47,381 traded animals over 2.5 years rules them out as the Huanan-market intermediate host or vector.