Abstract
Many writers have called the term ‘normal’ highly ambiguous both in and out of medicine, especially between descriptive and normative meanings. But careful analysis shows that its ambiguity is much less than usually supposed. In fact, all correct nontechnical uses of ‘normal’ mean “typical” in some way – either typical, at least typical, or at most typical – and therefore express no value judgments except by contextual implication. The distinctive, purely medical use, as the opposite of ‘pathological’, is just a specialization of the second meaning, to at-least-typical biological part-function. As statisticians have often warned, one must not confuse this uniquely medical use with a general use formerly applied to clinical tests, in the term “normal range.” That term is misleading because the reference ranges of clinical variables entail nothing about pathology, for three reasons: besides resting on an arbitrary choice of a 95% central range, they are derived from apparently healthy populations, and the variables’ connection to underlying biological function can be very indirect. So the term “clinical abnormality” is best restricted to a diagnosed or diagnosable pathological condition. If so, true clinical normality contrasts with theoretical normality in some interesting ways: it may or may not correlate with disease severity; it is individual-relative; and it is partly determined by value judgments.