The following essay argues that philosophers must take into account essential possibility and a posteriori necessity, and not merely logical possibility and logical necessity when constructing counterexamples. It will be argued that the notion of essences outlined below constrains the kinds of stories that can be used as counterexamples to theories. Essential necessity demands some attention be paid to the scientifically discoverable structure of the entities in question. Logical possibility does not demand such attention be paid to the structures of things, but merely to the formal definitions of the terms involved. In effect, I argue that philosophers would be right to regard crazy sounding logical possibilities as crazy, and that the notion of essential possibility I isolate explains why. Logically possible stories are crazy because they fail to take account of the natures of the things involved. It will be argued that the latter view is better because it conforms to the way things actually are—not to how we would like things to be.
The first chapter will examine two standard accounts of logical possibility, and in light of the deficiencies of these accounts provide a partial taxonomy of possibilia. Logical possibility (and impossibility) will be defined and contrasted to physical possibility (and impossibility), and essential possibility (and impossibility). The second chapter will examine Kripke's theory that names and natural kind terms are rigid designators; they signify kinds and causal origins that are both discovered by science and necessary a posteriori. The third chapter will examine the moral twin earth argument of Terrance Horgan and Mark Timmons; this will illustrate the thesis of this essay that essential necessity constrains thought experiments in a more useful way than logical possibility is able to do.