Reviewed by: Versions of Pygmalion Robert D. Newman J. Hillis Miller . Versions of Pygmalion. Cambridge: Harvard UP, 1990. 263 pp. $25.00. In his sequel to The Ethics of Reading (1987), J. Hillis Miller continues to defend deconstructive criticism from charges of amorality. Like Henry James in "The Art of Fiction," Miller argues that the act of reading is part of the conduct of life and that one's moral fiber is continually tested, expressed, and reshaped with each reading engagement. Books for him are not fixed entities packed with eternal wisdom but "unexploded bombs" constantly threatening to detonate or radically alter the perceptions readers bring to them. Miller also ponders the tendency of literary theory to comment on itself, increasingly in isolation from literary texts. He contends that the task of theory is application, to produce fresh readings that reopen both the literary text and the cultural text the reader brings to it. It is to this task and an examination of its meaning that Versions of Pygmalion applies itself. Miller presents carefully detailed readings of James's What Maisie Knew and "The Last of the Valerii," Heinrich von Kleist's "Der Findling," Melville's "Bartleby the Scrivener," and Maurice Blanchot's L'arrêt de mort. His controlling metaphor for all of these readings is prosopopoeia, which he derives from the Pygmalion and Galatea story in Book 10 of Ovid's Metamorphoses. For Miller, prosopopoeia, the giving of a face, voice, or name to something inanimate, is both a necessary aspect of reading as well as a trope of mourning, an allegory of death. Like Pygmalion's creation of Galatea in the image of his desire, readers conjure ideals in the texts they read. By encouraging readers in these delusions, however, the process of reading also exposes the act of prosopopoeia as our need to read ourselves in an Other, to compose a unitary reading that masters the [End Page 818] movement of narrative and imposes itself on the flux and variety of life. In such exposures, readers can assess their substitution of linguistic for material realities and commit judgments that recognize the inevitable Galatean nature that gives them form. Miller argues that language is performative as well as cognitive, an interesting distinction he fails adequately to clarify. Using Walter Benjamin's essay, "The Task of the Translator," he shows how the fluid and subjective nature of performance informs the reading process, functioning like the cultural gaps and linguistic nuances that make exact translation an impossibility. Miller also briefly indulges himself in the ongoing debate between deconstructionist critics and those whose criticism may be labeled new historicist or Marxist. Here he claims that the rehistoricizing of literature is "born of a refusal to recognize what is inaugural in literature, inaugural precisely in its historical effects." Again Miller is arguing that theorists not use literary texts simply as a jumping off point when theorizing or historicizing, a sage reminder but one too procrustean in its assessment of new historicism. In Versions of Pygmalion, J. Hillis Miller once again proves himself an insightful guide for subtle readings of individual texts and a profound investigator of the nature of reading. His use of prosopopoeia invites psychoanalytic, particularly Lacanian, elaborations. Aside from occasional allusions, however, Miller does not take this direction. Instead, his rereadings of Ovid's fable suggest the ethical consequences, broadly and flexibly indicated in social or historical ramifications, that the refashioning that is reading consists of and enacts. Robert D. Newman Texas A&M University Copyright © 1991 Purdue Research Foundation