[...] In Stakes¹ (let’s call it), the text’s success is suspiciously scanned, with the expectation, perhaps hope, that the piece of literature under scrutiny will turn out to be productively unsuccessful. In Stakes², the text’s success is anxiously searched for, with the assumption that the piece of literature’s lack of success cannot be productive for reading, but simply renders the book not worth picking up. The first way of reading is non-evaluative, at least at the level of craft or technique; the second is only evaluative, and wagers everything on technical success, on questions of craft and aesthetic achievement. Stakes¹ presumes incoherence; Stakes² roots for coherence. [...]

Not to think about literature evaluatively is not to think like a writer—it cuts literature off from the instincts and ambitions of the very people who created it. But to think only in terms of evaluation, in terms of craft and technique—to think only of literature as a settled achievement—favors those categories at the expense of many different kinds of reading (chiefly, the great interest of reading literature as an always unsettled achievement). To read only suspiciously (Stakes¹) is to risk becoming a cynical detective of the word; to read only evaluatively (Stakes²) is to risk becoming a naïf of meaning, a connoisseur of local effects, someone who brings the standards of a professional guild to bear on the wide, unprofessional drama of meaning.