Friday, December 04, 2009

Updating fallacy

Just now catching up on my reading of the book blogs, which I shamefully neglected during my ten-day stay in California. I am tickled to find that, while I was gone, Andrew Seal gave another demonstration of how not to do literary criticism. For reasons that are still not clear to me, he dusted off an old essay by the feminist scholar Nina Baym, which he found in a school anthology, and decided to use it as a “still serviceable model” for complaining about the exclusion of women from American fiction.

This is not, of course, the procedure of a rational inquiry. Seal announces that he is “updating” Baym’s essay, because it excites him and “feels largely on target.” Instead of examining its “bracing rush of argument,” then, he accepts its validity as a given, and bracingly rushes to his real concern: confirming its findings (“yep, guess we still do that”) by applying them to eight randomly selected novels written by men during the past decade. Seal is numb to the irony of treating as canonical—that is, established by the authority of republication in an anthology—an essay that reproaches the “canon” of American literature.

But what if Baym’s argument is false? If so it follows that “think[ing] through its major claims . . . in light of the American fiction of the past ten year[s]” is sound and fury. The first duty in reading literary criticism is to think through its claims, not by industriously extending them to further corroborating examples, but by scuffling to falsify them. If and only if the claims resist falsification can they be validly extended to texts undreamed of in the original critic’s philosophy. A critical argument may “generat[e] enthusiasm and a feeling of recognition” and yet be utterly false.

Does Baym’s argument hold up?

At least the argument is clear. “As late as 1977,” Baym declares, “that canon [of American literature] did not include any women novelists” for the simple reason that literature is read “always through the perspective allowed by theories,” and until very recently “theories of American literature” posited “a literature that is essentially male.”[1]

Neither major nor minor premise is true. Baym is not really interested in whether they are true, because she offers no argument in support of either. They are for her axiomatic: claims that are so obvious they need no further proof. In literary criticism, however, there are no axioms, because there is always something else from which a critic’s assertions follow. To treat a controversial claim as axiomatic is merely to explore the familiar contents of an ideology, to invite colleagues to rehearse the well-practiced tenets of a party.

Consider the claim that, as late as 1977, the American canon did not include any women novelists. It is impossible to determine what the word canon refers to. Course syllabi? Examination lists? Baym’s notes from lectures she heard as an undergraduate? Who knows? The vagueness of the referent liberates Baym from having to defend the claim.

For if she were referring to literary history, the claim would be demonstrably false. In his History of American Literature (1896), Fred Lewis Pattee devotes an entire chapter to “Woman in Literature” in which he discusses Helen Hunt Jackson at some length, describing her Ramona as a “matchless work of art,” and also includes longer or shorter sections on Harriet Prescott Spofford, Elizabeth Stuart Phelps, Sarah Orne Jewett, Mary Noailles Murfree, and Frances Hodgson Burnett, while glancing at Rose Terry Cooke, Jane G. Austin, Mary E. Wilkins, Mary Hallock Foote, Alice French, Rebecca Harding Davis, Louise Chandler Moulton, Blanche Willis Howard, Mary Hartwell Catherwood, and Margaret Deland. Elsewhere in the book he dilates upon Catherine Maria Sedgwick and Harriet Beecher Stowe, while paying less attention to Louisa May Alcott. In a book intended as a school text, Pattee lists the following novels by women as “required reading”: Uncle Tom’s Cabin (1852) and Stowe’s later Oldtown Folks (1869)—a book described in a 1953 article in College English by Ruth Suckow as an “almost lost American classic”—along with The Story of Avis (1877) and Jack the Fisherman (1887), both by Elizabeth Stuart Phelps, and Ramona (1884).

Except for Stowe, none of Pattee’s canonical women novelists received much attention in the twentieth century. Other women displaced them. The main figures were Edith Wharton, Willa Cather, and Ellen Glasgow. Critical discussion of them began in the ’twenties. Elizabeth Monroe singled out all three for praise and extended treatment in The Novel and Society in 1941, and Alfred Kazin did the same the next year in On Native Grounds. More women entered the “canon” in short order. Eudora Welty was first given the serious critical treatment in the ’fifties, Flannery O’Connor at the end of the decade (the first articles were on Delta Wedding and Wise Blood). Kate Chopin’s Awakening was rediscovered in 1956 by Kenneth Eble, and began attracting widespread scholarly notice in the late ’sixties. Alcott was firmly entrenched in American literary scholarship by the ’seventies.

Baym’s major premise is either too vague to qualify as a truth-claim or is factually untrue. What, then, of her minor premise? Is it the case that literature is read “always through the perspective allowed by theories”?

Again, the imprecision of the claim makes it difficult to know exactly what Baym means. If she means that apriori assumptions always precede the reading of a literary text, who would quarrel? But clearly she means something more:

There are . . . gender-related restrictions that do not arise out of cultural realities contemporary with the writing woman, but out of later critical theories. These theories may follow naturally from cultural realities pertinent to their own time, but they impose their concerns anachronistically, after the fact, on an earlier period. If one accepts current theories of American literature, one accepts as a consequence—perhaps not deliberately but nevertheless inevitably—a literature that is essentially male.Theories are collective, cultural, subsuming. They begin with a hypothesis; they employ a common vocabulary; they depend upon “the idea of Americanness” or “the idea of the best”; that is, they are an explanation of “some qualitative essence.” They are, in short, full-blown and widely accepted interpretations that are just waiting for new texts to be worked in.

But is it true that this is how unfamiliar books are read? A common cultural interpretation of an entire literature precedes the encounter with a new text, and reading it involves—perhaps not deliberately but nevertheless inevitably—making sense of it in the terms of the common cultural interpretation? Pretty clearly, Baym is advancing some literary version of a Kuhnian scientific paradigm (“universally recognized scientific achievements that for a time provide model problems and solutions to a community of practitioners”).[2] And her concept of “the perspective allowed by theories,” which defines the reading of literature, suffers from many of the same problems as Kuhn’s concept of a paradigm.

For young critics like Seal, the suggestion that there are any problems at all with Kuhn’s paradigm is not merely heretical, but absurd. The concept of the paradigm is axiomatic in a certain broken-backed style of criticism wildly popular in academic departments of English, where Seal memorized his errors a few years ago. How is it, though, that (in Baym’s words) “we never read American literature directly or freely, but always through the perspective allowed by theories,” but somehow we enjoy unmediated access to this very assertion? Or is it only American literature that is filtered through a ready-made perspective? And are we really sure that such a perspective even exists?

No doubt many sweeping interpretations of American literature are similar, but does this similarity leave us no choice but to believe in the hocus-pocus of a single common theoretical perspective? The truth is that Baym’s argument is non-falsifiable because it twirls around in circles: the “basic American story” is found to be antagonistic to women because antagonism to women is defined in advance as the “basic American story.” That’s just what makes it a “theory” or “perspective” or “paradigm.”

More to the point, how did someone like Nina Baym come to shake off the prevailing theories of American books, which encouraged her to accept a literature that is essentially male, in order to adopt a different paradigm in which women novelists are included? Is it possible that she arrived at an interpretation of the American novel that differed from the prevailing theories? On her own showing, though, these theories do not allow a different interpretation. How then was she able, all on her own, to shift the paradigm?

The plain truth that new books are not read “through the perspective allowed by theories,” and if they were, Andrew Seal would not have to study the perspective in order to extend it, rather woodenly, to eight novels by males in the last ten years. He would, instead, read the novels in its terms unconsciously, without any intellectual freedom, as if no other terms were even conceivable.

But that is not how we read. We read new books against a lifetime of reading books, and as we open a new book we toss a net of expectation over its unread portion, which we pull back and adjust as we go along—as the book is converted from expectation to memory. By this means we are able to avoid such stupidities as saying, for instance, that Delphine Roux, the French professor in The Human Stain, is an “encroaching, constricting, destroying” woman who proves that Philip Roth “conform[s] to the American myth,” according to Nina Baym, whereby men are “beset” by women, but “struggling to break free” of them—even though the agent of Coleman Silk’s freedom is Faunia Farley, another woman conveniently ignored (because inconvenient to his theory) by the theory-besotted Andrew Seal.

[1] Nina Baym, “Melodramas of Beset Manhood: How Theories of American Fiction Exclude Women Authors,” American Quarterly 33 (Summer 1981): 123–39.

[2] Thomas S. Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press), p. x.


R. T. said...

You say: "We read new books against a lifetime of reading books, and as we open a new book we toss a net of expectation over its unread portion, which we pull back and adjust as we go along—as the book is converted from expectation to memory."
I say: Isn't this a bit like converting Harold Bloom's concept of anxiety of influence (in Bloom's usage, relevant to authors) and extending it to reading? We are, after all, repositories of all texts we have previously encountered, and either consciously or unconsciously we are "anxious" about new texts as we read them and attempt to match them against our critical judgment based on earlier textual encounters. Perhaps Seal's error--to the extent that there is one--can be found in what you seem to suggest is his treatment of Baym in isolation without awareness (or acknowledgement) of Baym's antecedents and descendants. This tends to be a naive approach to criticism in which the critic finds secondary sources that agree with his or her critical disposition and goals, and then conspicuously ignores either the flaws in the sources or the arguments from other sources, all in the name of furthering the already decided critical position. Hence, the critic denies--either consciously or unconsciously--the anxiety of influence that ought to be acknowledged and confronted.

D. G. Myers said...

The “application model” of criticism, in which a dominant figure’s conclusions are extended to new texts, is indeed naïve.

But the reality that I describe owes little to Harold Bloom. It depends upon Augustine’s account of reading a Psalm (in Book 11 of the Confessions):

“I am about to repeat a Psalm that I know. Before I begin, my expectation is extended over the whole; but when I have begun, how much soever of it I shall separate off into the past, is extended along my memory; thus the life of this action of mine is divided between my memory as to what I have repeated, and expectation as to what I am about to repeat; but ‘consideration’ is present with me, that through it what was future, may be conveyed over, so as to become past. Which the more it is done again and again, so much the more the expectation being shortened, is the memory enlarged: till the whole expectation be at length exhausted, when that whole action being ended, shall have passed into memory.”

J. V. Cunningham quotes and discusses this passage in “Poetry, Structure, and Tradition” in the Collected Essays, pp. 143–44.

As I understand it, the “anxiety of influence” applies to poets, not necessarily their readers.

R. T. said...

Yes, I agree that Bloom's concept applies to poets (though I think by inference he includes all authors), but I dared to borrow and expand the concept (which might surprise Bloom) and apply it to readers and critics because I believe the same dynamics are at work.

D. G. Myers said...

Ah. Well, maybe so. But I don’t really see anything anxious about the reading procedure that I describe. If there is any anxiety involved, as a matter of fact, I’d locate it in the unfamiliarity of a new text.