A many-worlds interpretation is of quantum mechanics tells us that the linear equations of motion are the true and complete laws for the time-evolution of every physical system and that the usual quantum-mechanical states provide complete descriptions of all possible physical situations. Such an interpretation, however, denies the standard way of understanding quantum-mechanical states. When the pointer on a measuring device is in a superposition of pointing many different directions, for example, we are to understand this as many pointers, each (...) in a differentworld, each pointing in a different determinate direction. We ask here whether such talk makes any genuinely intelligible sense of the term world. We conclude that it does not. (shrink)