McEwan, Ian. "A Parallel Tradition." GUARDIAN April 1, 2006.

Those who love literature rather take for granted the idea of a literary tradition. In part, it is a temporal map, a means of negotiating the centuries and the connections between writers. It helps to know that Shakespeare preceded Keats who preceded Wilfred Owen because lines of influence might be traced. And in part, a tradition implies a hierarchy, a canon; most conventionally, it has Shakespeare dominant, like a lonely figurine on top of a wedding cake, and all the other writers arranged on descending tiers. In recent years, the canon has been attacked for being too male, too middle class, too Euro-centric; what remains untouched is the value of a canon itself: clearly, if it did not exist, it could not be challenged. But above all, a literary tradition implies an active historical sense of the past, living in and shaping the present. And reciprocally, a work of literature produced now infinitesimally shifts our understanding of what has gone before. You cannot value a poet alone, T. S. Eliot argued in his famous essay, "Tradition and the Individual Talent", "you must set him, for contrast and comparison, among the dead." Eliot did not find it preposterous "that the past should be altered by the present as much as the present is directed by the past." We might discern the ghost of Auden in the lines of a poem by James Fenton, or hear echoes of Wordsworth in Seamus Heaney, or Donne in Craig Raine. Ideally, having read our contemporaries, we return to re-read the dead poets with a fresh understanding. In a living artistic tradition, the dead never quite lie down. Can science and science writing, a vast and half forgotten accumulation over the centuries, offer us a parallel living tradition? If it can, how do we begin to describe it? The problems of choice are equalled only by those of criteria. Literature does not improve; it simply changes. Science, on the other hand, as an intricate, self-correcting thought system, advances and refines its understanding of the thousands of objects of its study. This is how it derives it power and status. Science prefers to forget much of its past - it is constitutionally bound to a form of selective amnesia. Is accuracy, being on the right track, or some approximation of it, the most important criterion for selection? Or is style the final arbiter? . . . Read the rest here: http://www.guardian.co.uk/books/2006/apr/01/scienceandnature.richarddawkins.