Drawn from lectures delivered to literary critics, this collection is more deaconly and theoretical than Kermode's last book on the legitimacy of interpretation, The Genesis of Secrecy. But it unshirkingly continues--with admirable steadfastness and without reactive hysteria--a reasoned case for hermeneutics, for a literary criticism that doesn't (as the new French and Yale-ites seem to do) throw out the baby with the bath-water. Kermode will not apologize for the sharp eye he keeps peeled for "secrecy, an inexplicable consonance" in fictional narratives: "We know it is there because the disturbance in the prose signals its presence. It is so arranged that we . . .may ignore it if we choose; but it is there, and has to do with expansion, with the reconciliation of art with honorable rest." This persistence of secrecy in texts, says Kermode, puts the Derrida/Barthes mania fora reader-made or "lisable" text to some disadvantage; all novels, he stresses, share a "property of narrative: it accepts and exploits the pluralities that arise from their situations." As examples, he goes far and wide to analyze Ford's The Good Soldier, Conrad's Under Western Eyes, Green's Loving, even Arnold Bennett's Riceyman Steps and the English mystery Trent's Last Case--finding each of them faux-naif, more knowing and lisable than the French school would even begin to give them credit for. Generous, patient, strenuous Kermode defangs the New French criticism of its self-congratulatory novelty; he demonstrates that all stories, as far back as the Gospels, will mold to some deconstructionist demands--but disproves the idea that any work can be totally self-referential, free-playing, and "unmeaning." This middle ground of serious criticism has so far found its clearest, most wide-ranging advocate in Kermode; and his new book, though difficult and strictly academic, is an important one--ecumenical, level, acute--in its field.