In this fairly short reading, from Theorrhoea and After (1999), which condenses views already set out at greater length in Newton's Sleep, Tallis begins by drawing on a story of Tolstoy's: an aristocratic lady is at the opera weeping at the imagined tragedy on the stage, while outside her faithful old coachman is freezing to death. Tallis's purpose here is to open a consideration of the moral value of art, and in particular of the novel. One might argue, as many critics have done, that by showing how people are corrupted by others, how they influence one another, and so on, the novels of Henry James, say, increase our sensitivity to the reality of the lives of other people, and by deepening our imaginative grasp of their lives open us up to more profoundly human relationships. Tallis rejects this view. There is no empirical evidence for it — indeed, what would constitute such evidence? — and all claims to that effect must be based on an a priori assumption, and, as Tallis argues, there are many reasons for thinking this assumption unacceptable. He wants, in short, to argue that claims for the value of art based on claims for its ability to influence behaviour for the better are without foundation. The value of art lies elsewhere, in the kingdom of ends: while it may not improve life, a great novel will introduce those who are its readers to a greater awareness of the complexity of the meanings that the world may bear, and so widen and deepen their experience of life. The appreciation of literature — and appreciation is inseparable from evaluation — is a matter of responding to the way disparate themes vital to human concerns are drawn together and unified.