O-26 - The natural reservoir and the index case exposure route were not identified by the 1976 Ebola investigation
Note: the source-node summary’s ‘asymptomatic seropositive villagers with no hospital/known-case contact’ detail is NOT established by this 1976 report (no such community serosurvey is reported in it); only the hospital-amplification and unresolved-source findings are extracted.
Link to original
Why this is evidence
That a clearly natural outbreak (1976 Ebola) left its animal reservoir and index source unidentified is a base rate showing that failure to find SARS-CoV-2’s reservoir (O-27) is weak lab evidence — raising P(no reservoir found | H-41) and blunting the inference toward the lab members.
Likelihood
# E-58 (HC-1) — lone edge, one observation O-26 "a confirmed natural zoonosis (1976 Yambuku Ebola) left its
# animal reservoir and index exposure route unidentified". A BASE-RATE / precedent datum: it establishes that
# reservoirs commonly go unfound even for outbreaks known to be natural, which makes SARS-CoV-2's own missing
# reservoir (O-27, priced separately at E-59) consistent with natural origin. It therefore FAVOURS H-41 over
# the lab bloc {H-42, H-43}; it does NOT distinguish H-42 from H-43 (neither needs a trade reservoir, so the
# precedent carries no differential pull between them). Members in HC-1.hypotheses order [H-41, H-42, H-43,
# H-44]. Anchored on H-41 = 1; lab members priced as ratios to it (rule 7). A-42 (approved) reinforces the
# datum: the visible Yambuku cluster is an iatrogenic-amplification/ascertainment artifact and the true
# spillover was left permanently unresolved by the outbreak record — i.e. a genuine natural zoonosis whose
# reservoir the data never located. NOTE: this edge points OPPOSITE to E-59 on the same "was a host found"
# question; each is priced on its own and the runner composes them.
lik_ebola_H41 = 1.0 # anchor: under natural zoonosis the "no reservoir found" state is exactly what this
# precedent shows to be normal, so O-26 sits most comfortably here
lik_ebola_H42 = 0.8 # ~0.8x: under an unmodified-collection leak SARS-CoV-2's missing host is explained by
# there being no trade reservoir at all, so the reservoir-unfound precedent adds little
# — modestly less favoured than under H-41
lik_ebola_H43 = 0.8 # identical to H-42: an engineered leak likewise needs no trade reservoir, and O-26 does
# not discriminate the two lab members
lik_ebola_H44 = 0.9 # residual, unconstrained: some unlisted/hybrid origins involve natural spillover and
# some don't — middling, near-neutral
t_ebola = 0.85 # cap = trust_score of S-88 = 0.85 (International Commission 1978, canonical first Ebola
# investigation, hospital-amplification finding never overturned in 45 yr). At cap: the
# raw datum — that the 1976 reservoir/index source was never identified — is solidly
# established (the excluded "seropositive villagers" detail is already stripped in O-26,
# so no measurement weakness remains in what is stated). The one-Ebola-case reference-class
# generalisation is an inference-strength matter, priced in the modest ratio above, not a
# data-reliability dock.
evidence("HC-1", ["O-26"], [lik_ebola_H41, lik_ebola_H42, lik_ebola_H43, lik_ebola_H44], t=t_ebola)