perception, nor to observation. Observation supplies evidence that is basic relative to a field of enquiry and can be highly non-perceptual. Theoretical knowledge is typically gained by inference to the only explanation, in which competing plausible hypotheses are falsified by the evidence. In cases where not
all competing hypotheses are refuted, scientific hypotheses are not known but instead possess varying degrees of plausibility. Plausibilities in the light of the evidence are probabilities and link eliminative explanationism to Bayesian conditionalization. Bird argues that scientific realism and anti-realism as global metascientific claims should be rejected-the track record gives us only local metascientific claims.