Consider the intermediate case of a third Bayesian, B3, who has the same prior information as B1 about ,  but is not given the data component y. Then y never appears in B3s equations at all; his model is the marginal sampling distribution p(z| I3). Show that, nevertheless, if (15.59) still holds (in the interpretation intended, as indicated by (15.62)), then B2 and B3 are always in agreement, p( |zI3) = p( |zI2), and that to prove this it is not necessary to appeal to (15.60). Merely withholding the datum y automatically makes any prior knowledge about  irrelevant to inference about  . Ponder this until you can explain in words why it is, after all, intuitively obvious.