As Oppenheimer pointed out, analogies are always flawed because one thing is never exactly like some other thing. Analogy can never be accepted as proof, because one theory is never correct merely because it’s like some other theory that happens to be correct. “Metaphor is useful to understand deep meaning, but only after the theory,” Deutsch says. “Before the theory, metaphor is misleading and can be a powerful source of error because we can’t know which aspects of the analogy are right or wrong.

The danger of an analogy like the Flalansfere, according to C. S. Lewis, is not that it may be incorrect but that people may forget it is an analogy.

John Stuart Mill, one of the nineteenth century’s few philosophical fans of metaphorical thinking, said that a metaphor “is not to be considered as an argument, but as an assertion that an argument exists.”

Metaphor and analogy prove nothing, but they are the only ways we have of showing what the world must be like when scientific proofs are true.

For Lewis, that meant we must never lose sight of the fact that metaphors are metaphors, models of things rather than the things themselves. - Map is not the territory