“The French term revenant (someone who revient, who comes back)” writes Miguel Tamen, implies “that a phantasm, a vision, is indeed a revision.” Thus, he suggests, there is something ghostly, something at once unheim-lich and ephemeral, about every vision and revision of literary history. The history of literary-historical projects is the story of revenant-like returns, repetitions: “In every revision, the past gets apprehended through visions. Deceptively varied as the latter might be, however, each and every vision consists primarily in the ascribing of mistakes to previous visions.” It is sobering, especially to a critic who aspires to say something new, to make visible something hitherto unseen, to be reminded that there is a heimlich side to his or her novel views – that “the practice of literary history, while often presenting itself as producing definitive visions of well-established topics, cannot help being a revisionary practice.” And it is a little disquieting, but humbling, to think that, as Tamen reflects, “all revisions get revised, as my own truly true ghost descriptions will become, at the most, someone else's ghost.”

Writers, I find, know better than anyone else where their weaknesses lie. Instead of a retrospective or punchy anecdotal ending, I shall conclude with a brief look at one reason why my readings of the texts I have tackled are open to revision – how these readings are made ghostly because of the blind spot, or blind spots, that they generate, and which are both inevitable and deliberate. They are inevitable, I think, because every project of this kind is structured around omissions; one can never say everything one would like to say (and very often one should not).