The words 'true value' assume an existential relationship with the thing being measured. There is assumed (by the speaker of the words) to exist some aspect of the thing being measured that is independent of its relationship to the person who is interested in the result of the measurement. Yet measurement is a response to the need of an observer to know something about the real world. There is therefore an epistemological aspect to measurement. Some aspect of measurement has to do with the observer as a knower of the result of the measurement. The paper considers electrical measurement. It proposes that these relationships correspond to two distinct kinds of measurement: one has the characteristic that there does exist such a thing as a true value, though it cannot be known: in the other kind, no such thing as true value even exists. The notion of true value is very useful, and for the kind of measurement where such a thing does not exist, the paper defines an alternative that has the advantage that it can actually be known under some circumstances in calibration. © 2018 IEEE.