Wikipedia 10K Redux by Reagle from Starling archive. Bugs abound!!!
One may certainly hold more stubbornly at least to those notions which serve to give context, comprehensibility, and utility to other notions. Care should be taken, however, not to exclude new information which is not inherently self-contradictory. To the degree that new data conflicts with a large body of mutually consistent observations (note: observations - not beliefs) one may more safely devalue it, but that a new idea was not suggested by one's interpretation of earlier data is not the same as to say that one may safely discard it because it comes as a complete surprise. Rather, "surprise" information, instead of being automatically viewed with skepticism, may be be held in a bullpen while it is compared with other cross-verifying observations, and, if not inherently inconsistent with old data, accepted conditionally for purposes of comparison with new input in order to see what it does or does not explain. People call this "having an open mind." -- AyeSpy