The primary benefit touted for big-endian is "When I do a memory dump, the data looks right." But if you really believe the left side is bigger, why do you put

The primary benefit touted for big-endian is "When I do a memory dump, the data looks right." But if you really believe the left side is bigger, why do you put the smaller memory address on the left side of your dump?