Reducing poetry to a science

Is it possible to tell a good poem from a bad one? To put the question another way: are there objective, even scientific, standards for evaluating literature? Helen Vendler has no doubts. Her spiky new collection of essays begins with the insistence that it is possible to prove how one poem is ‘superior’ to another, and ‘those who suppose there are no criteria for such judgments merely expose their own incapacity’.

That’s a bold claim, but in her hands, literary criticism is a science, and anyone who disagrees with her judgments is put sharply in their place. I should know: my observation, in a book I recently edited, that the late religious poetry by the great American poet John Berryman reminds me of the 17th-century English poet George Herbert is here quoted and then smacked down (‘never in them does the histrionic Berryman sound at all like the subtle and fine-grained Herbert’).