Abstract
It is shown that to combine readings from “points of view” configurations as assumed in Tucker and Messick's model is not to combine configurations in any simple way. Both empirical and logical consequences of the disparity are discussed.