The need for epistemic security is best satisfied by a scientific rather than a traditional empiricism. The emphasis on human observations can be transcended by the dilution argument and the overlap argument, but scientific empiricism requires that we know how instruments work, a position argued with reference to Galileo and Hacking. The three instrumental criteria of accuracy, precision, and resolution are examined. Instruments detect properties and it is argued that objects are clusters of properties.
Keywords: accuracy, dilution argument, empiricism, epistemic security, instruments, observations, overlap argument, precision, properties, resolution