Warren in the Spider's Web

Since his death in 1987, Robert Penn Warren has suffered the peculiar fate of American writers who fit uneasily, if at all, into the schools we concoct to house our national genius. These iconoclasts (others who come to mind are John Dos Passos, who never quite fills the mold of a naturalist or a modernist, and James Merrill, whose formal labyrinths outdid new formalism before it had a chance to be new) find their reputations torn between hagiography and desecration. Fortunately, a few of Warren's literary works have survived this tug of war, and we can enter the next phase of critical review. The essential phase, one might add, because if a writer cannot belong to a school, he needs to be part of the connective tissue of the canon, or we will surely slough him off. Two of the books under review here delineate the web of influence that Warren helped spin, while the third goes deep into Warren's material in order to chart the tension between that material and the demands of the literary world itself.

Keith Perry's The Kingfish in Fiction sets Warren among his near-contemporaries Sinclair Lewis, Hamilton Basso, John Dos Passos, and Adrian Locke Langley, all of whom used the Southern demagogue Huey P. Long as the model for a protagonist, with varying degrees of success. By "success" I mean here the novel's coherence as a work of art, though what Perry means is sometimes unclear. He claims to be determining the "'aesthetic significance' of the convergences and divergences" between the historical Huey Long and his fictional counterparts, but sometimes seems engaged in a game of "Gotcha!" Discussing Sinclair Lewis' It Can't Happen Here, for instance, he writes, "Long and Windrip have their differences, of course, but those differences are more matters of degree than class," and in Dos Passos' Number One, "Crawford's resumé, appearances, and personality all resemble Long's." Details in All the King's Men, on the other hand, vary in ways that Perry is quick to point out: "The shirt is the same, and the man inside it is roughly the same size, but the suit in the [historical] photo, though cheap, is not seersucker, the tie not striped." Perry's goal is nonetheless larger than the fact-or-fiction game. He is attempting to discover what use these writers made of the facts or the myth of Long, as a way of exploring the relation of historical life to narrative art.

Perhaps because the authors Perry examines are less read these days, he expends an inordinate amount of time on their backgrounds, especially in the cases of Sinclair Lewis and John Dos Passos. That Lewis' novel It Can't Happen Here was a natural outgrowth of books like Elmer Gantry and Ann Vickers may shed light on the book's political outlook, but the lengthy summation of Lewis's political maturation does not go very far toward clarifying either how the novel was structured to "prevent . . . a 1936 Huey Long presidency" nor why it fell often into "protracted burlesque." Similarly, Perry works hard to establish Dos Passos's social-protest credentials, but the effect is to distract us from his central argument that Dos Passos's novel Number One presents a character who not only is closer to "what the real Huey Long seems to have been" but also embodies "threats . . . [that] in no way died with Long."

By contrast, Perry's discussion of the more minor authors Hamilton Basso and Adrian Locke Langley takes a refreshingly candid look at flawed novels whose characters nonetheless convey a...