“Superior intellect, as Professor Bain has admirably shown, seems to consist in nothing so much as in a large development of the faculty of association by similarity.” This footnote appears early in William James's The Varieties of Religious Experience.
Some of my own associations with this:
Douglas R. Hofstadter says somewhere – I can't find the quote, so I'm paraphrasing from memory -- that some people are too limited in the mental connections they make, and think too rigidly to be innovative, while other people link ideas too profusely and loosely, so that their associations aren't that useful -- and that there's an ideal point to be at on this spectrum, where one's associations are creative yet not too weird to have some kind of application.
“Superior intellect” is an unpleasantly stuffy idea, but I do like to think of learning as the tuning of synapses, the training of one's brain not to associate too tightly or loosely.
Also, Stephen Jay Gould denied somewhere that he had read more than other people – he said his advantage as an essayist was rather that, when he encountered an idea, he was better than other people at recalling other ideas that could be usefully juxtaposed with the one just encountered. I find it a beautiful thought that we have to read new things in order to be able to summon from memory the old things we've read...