Beliefs and desires are linked with one another by an obvious kind of duality. Everyone assumes that a person's beliefs may be, to some extent, ascertained by seeing how he acts: for we suppose that he will do the things which he believes will fulfil his wishes, and avoid doing what he thinks will frustrate them. Similarly, his desires may, to some extent, be ascertained by seeing how he acts; for if we know what he thinks about the results of (...) his actions, we suppose that the expected results are desired or not disliked and those believed unlikely are disliked or undesired. We have three variables, conduct, desire, and belief; and we suppose that there is some equation by which the first in conjunction with one of the others implies a determinate value of the third. In principle, it should be equally possible to deduce conduct from belief and desire. But that seems of less metaphysical interest, because we think of actions as open to observation in a straightforward way, whereas we are more dubious about beliefs and desires. (shrink)
In “Intrinsic value: some comments on the work of G. E. Moore” , I argued that Moore implies that intrinsic value is measurable, but has never suggested any method of measuring it. In this note I shall outline a method which is derived, but not exactly copied, from some of the analyses which economic theorists have given of “cardinal utility”. A survey of some of the economic discussions is given by S. A. Ozga in “Measurable utility and probability”.