Science and Literature

I found this passage of John Gardner's On Moral Fiction rather interesting:

"Good science normally makes hypotheses based on observation or probability; art deals, at its best, with what has never been observed, or observed only peripherally--darts from what is to what might have been--asking with total interest and sobriety such questions as 'What if apple trees could talk?' or 'What if the haughty old woman next door should fall in love with Mr. Powers, our mailman?' The artists' imagination, or the world it builds, is the laboratory of the unexperienced, both the heroic and the unspeakable."

Perhaps this articulates my feeling at the suggestion that literary criticism needs to incorporate more science. I say, thanks but no thanks. The methods science uses to make sense of our universe, and the conclusions it reaches, while valid and important, are simply not the same as the methods and conclusions of literature, which are also valid and important.