“Whether sociology can ever become a full-fledged "science" (a description of a class of events predictable on the basis of deductions from a constant rationale) depends on whether the terms which sociologists employ to describe events can be analyzed into quantifiable observables.”