Ben! writes:
Is scientific knowledge the set of all source data in the literature, or is it the set of inferences and conclusions that have been generally agreed upon by the scientific community, based on those data?
I think of "knowledge" as a geodesic sphere; the joints are the data points and the beams are the inferences that connect them. No data point or inference can stand alone - i.e. hang freely in space; they are all connected.