Having neglected my own blogging for a couple of weeks, I am going to catch up with one big blog-o-lump.
Big and little magazines in the divided field
First, modernist little magazines. Egoist, Blast, and Little Review, in their limited circulations and often esoteric content, seem to exemplify modernism’s self-conscious choice of a specialist audience. Here if anywhere is the “subfield of restricted production.” One of the main efforts of the so-called “new modernist studies” was to complicate this view of modernism as a specialist pursuit: one argument after another challenged Huyssen’s “great divide” formulation by showing how classic modernism was fascinated by or linked to the broader currents of its time. Though periodical studies was not the main current of “new modernist studies,” the kinds of inquiry exemplified by Morrisson or the Modernist Journals Project are related to this tendency. Magazines are by their nature more heteregeneous, and perhaps also more porous, than the big books of monumental authors.
The chapter we read by Morrisson is distinctive, however, in that its attack on the “great divide” is mounted by linking high modernism not to a generalized Zeitgiest or a broad mass culture but to feminist print culture and publishing practice. The feminist “counterpublic sphere,” tied as it was to a mass political movement, had a different social base and a wider social reach than artistic experimentation. Thus Morrisson’s argument, more complex than most overcomings of the great divide, suggests that a title like The Egoist relates to consumer culture, advertising, etc. through the mediation of the feminist movement—very particularly, via Dora Marsden’s individual trajectory.
Nonetheless, the divide refuses to be closed. The fact that The Egoist modeled some of what it did on Votes for Women, or that Blast’s visual rhetoric unmistakably resembles (or pastiches) that of advertising and political placards, does not actually mean these magazines are just mass-cultural artifacts like any other. Their drive to distinction and their restricted audiences are social facts. They are not pure vehicles of pure art, but they solicit (and depend upon) an audience that knows how to discriminate between what they do and what their cultural rivals do. When Bourdieu wrote (I put it on the handout) that “there is no way out of the game of culture,” he did not, of course, mean that no distinctions exist in the cultural field. These distinctions are made through different types of relations to the marketplace, not through being in the market or (impossibly) getting outside of it.
Next, larger-circulation magazines. Actually, I find publications like the Illustrated London News, the Strand, or even the London Mercury more challenging for the theory of the divided field than the little magazines. They all take literature quite seriously, even the ILN, where books occupy only a back page, and sometimes emit apparently sincere signals of their desire for cultural distinction. Where are the signs of profiteering or the dominance of economic over symbolic capital? They do exist, but they come in disguised and ambivalent forms: in class we talked about the treatment of authors as celebrities in the Strand. Notice also the difference between the language of criticism and debate one finds in Little Review or Egoist (and the language of manifesto in Blast) and the language of appreciation or depreciation in the book review pages of the ILN (say). As for the Mercury, it is a high-brow publication, just one that lies on the wrong side of literary history. Its general atmosphere is comparable to that of the official artistic academies and other institutions of consecration Bourdieu discusses. But the difference between Mercury and the Criterion is the difference between two competitors in the restricted subfield, which nonetheless is also a difference between a less restricted and a more restricted (and more “advanced”) periodical.
Bags of words
How these divides are related to the divergences in the status of books discussed in Algee-Hewitt et al.’s “Canon/Archive” or Underwood’s Distant Horizons seems to me to invite more reflection than we had time for. Just to make once more the point I made before Ted joined our seminar session, the “archive” and the “un-reviewed” are not equivalent to “mass culture” or the subfield of large-scale production. In both “Canon/Archive” and Distant Horizons, most of the books that have lower literary status also have low sales. Only the books in the “bestseller” category, which crystallizes in the 1890s, really exemplify the “mass cultural”—and not all of those, either, since, as we know from James English, bestsellers and prizewinners overlap substantially right through the midcentury. (The Sheik is untypical, an extreme example of the “absolute bestseller.”) Neither the Lit Lab study nor Underwood’s seeks to differentiate the neglected books which were trying to be popular from those which were trying for literary status. What does “trying” mean, anyway? How would you tell? (I think you could…)
What is most impressive and most frustrating about these two computational literary inquiries—and pretty much all others too—is their focus on the features of text as the variables of interest. Though Underwood is very cautious about saying that textual differences do not cause differences in status (as we saw, the Lit Lab pamphlet is bolder), in terms of the statistical modeling, the textual features are the “independent” or explanatory variables and the differences in status are the “dependent” ones. It is pretty strange that knowing only how often a bunch of words occur in a book of Victorian poems is enough to let the computer guess whether that book got reviewed in literary periodicals. It’s even wackier that knowing something about the frequency-pattern of word pairs (“bigrams”) let the Lit Lab researchers make good guesses about whether Victorian novels would be in the Chadwyck-Healy Literature Online collection.
There are two levels of strangeness about this. First is the correlation–detectable only by computation—between “bags of words” and humanly-recognizable qualities of diction, style, and theme. Strange as this appears, literary critics have long assumed that meaning is transmitted at many levels of language, and linguists have known for a long time that word frequencies reflect differences in individual style and in “register” or genre. But it is no less strange to learn that “prestige” durably correlates to these word-frequency features. This is a fact to be explained, but it would be an error to suppose it means that there is a sorting process where literary judges (e.g. reviewers) consider all the books that are published and choose the ones with certain kinds of language for consecration (e.g. being reviewed).
Given everything we know about how books are made and circulate, this is unlikely. In talking with us, Ted gestured at Bourdieu’s theory of the habitus (“durable, transposable dispositions”) as a framework for an explanation. He implied, I think, that the ability to be reviewed (because of connections to publishers and choice of subject matter or genre) is correlated with the kinds of qualities (educational trajectory, social location, etc.) that make one write review-able books. It is interesting to think about how this hypothesis could be put to the test.
By the way, Ted is particularly exemplary in his commitment to laying bare all the details of his methods. It’s not all spelled out in the text of Distant Horizons, but he created a reproduction repository with code, data, and instructions that would, with enough patience, let you recreate all the figures from the book. I draw your attention to the tables of “metadata” (i.e. title and publication information) of the poetry and fiction volumes considered in the analysis, which can be browsed online and make for instructive reading.
Equally instructive are the outputs of the statistical models, the coefficients assigned by the model to each word. These numbers tell you whether making a given word more frequent will make the model more or less likely to predict it belongs to the reviewed set: thus the lowest-status words in the model are “command,” “maidens,” “’ll”; the highest ones are “seasons,” “sign,” “hers”—but browsing the full 2600 features will help you contemplate the true weirdness of this kind of modeling.
In fact there’s even more there than all that, since his notes include further discussions of some of the methods beyond what’s in the book. From the material for chapter 3, I call your attention to the “notebook” on the method for estimating sales using “empirical Bayes,” not so much for the numerical details as for a general sense of how one makes usable data and draws inferences from it.
Now, as for The Sheik. First of all, the whole 1921 silent film is on the Internet Archive. Leavis tells us quite a bit about the coupling between fiction and film (films of the novel, novels of the film, film magazines, etc.), and, bracketing off her scorn for the phenomenon, we can reflect on the novel’s position in a changing interwar media ecology. I am tempted to say that The Sheik anticipates what Henry Jenkins calls “transmedia storytelling,” that is, a fiction that “unfolds across multiple media platforms, with each new text making a distinctive and valuable contribution to the whole.” There’s no reason to suppose E.M. Hull anticipated the Valentino vehicle, but the circulation of her novel takes place in a cultural world defined by the possibility of such vehicles—and of the positive feedback effects that send filmgoers in search of the novel (with its more transgressive thrills, censored out of the film) and novel readers out to the film (with its visual impact). In this light, the novel’s exploitation of the chapter gap for the sexiest parts reveals something important about novel-reading as an erotic technology.
We talked a bit about The Sheik as a proto-romance novel. It has been identified as such for a long time, including by pioneering romance novelists like Barbara Cartland. (The last time I taught this course, a student gave me an abridged edition of the novel from Cartland’s Library of Love.) Nonetheless, it is anachronistic to speak in 1919 of the “romance novel” in the sense of a specific channel of book production for love stories aimed at women. The term “romance” still had its older meanings of “adventure” or “fanciful story.” The London firm of Mills & Boon, which is now totally identified with the romance (and a subsidiary of Harlequin Ltd), specialized in this category only in the 1930s. Specialized production in mystery novels also emerges in the 1930s, and “genre fiction” as a system only comes into view with the boom in paperbacks after the Second World War.
The Sheik does anticipate some later iterations of the love-story genre which were central to establishing North American romance-novel publishing: the “sweet savage romance” which boomed in the 1970s (according to Radway). Our discussion focused, rightly, on the ambivalent representation of sexual violence. I want to draw attention to two other features of the novel’s handling of sex and gender. One is that, in the happy ending, Diana and Ahmed do not get married. They just go back to sleeping together in the desert. This was, though certainly not unprecedented, still transgressive—it was, and indeed it remains, typical for novels of love to punish unmarried cohabitation or “rectify” it with marriage.
By contrast, the whitewashing origin story for Ahmed does appear to rectify his alien racial status. Yet not only does this story simply displace racial mixture back into the distant past (his Spanish mother’s “Moorish” antecedents); it also breaks down the apparent distinction between the civilized white world and the barbaric Arab desert by comparing the wife-beating Lord Glencaryll to the dignified Sheik Ahmed père. In a parallel moment earlier in the text, Diana notices a resemblance between Ahmed and her brother Aubrey: “She had seen Aubrey do similarly hundreds of times. Occidental or Oriental, men seemed very alike” (106). The exotic thrill of primitive masculinity turns out to be another version of the masculine domination any woman could find at home. But to say this is hardly to say that the book mounts a critique of patriarchy, only that the ubiquity of patriarchy supplies one of the fundamental premises on which the novel builds its fantasies of escape-by-subjection.
However, the sociology of reading enjoins us not to stop at textual interpretation. We should ask seriously what people who use the book actually use it for. Q.D. Leavis thinks she knows (“compensation for life” ), but her sense of cultural embattlement—and her commitment to the high-modernist project—prevents her from actually going and finding out. As I tried to hint in class, literary studies has still not really faced this challenge. The dominant mode of our discipline remains expert “reading,” backed by the conviction that the significance of any literary work is largely due to what expert interpretation can reveal in its text. Because we have our own ways of doing things with books, we lack a disciplinary framework for studying what people who aren’t us do with them.
I also tried to hint that Leavis’s dilemma, which pushed her into the ludicrous idea of organizing an “armed and conscious minority” of cultural critics, remains our own. There is a great divide, for scholars: the divide between our ways of handling culture and the other ways that people practice. For us it is unavoidably freighted with moral significance (we believe we do it right; we try to get students to do it too). Once again I return to Bourdieu’s maxim: “there is no way out of the game of culture.” It is not so simple as adopting a new, more righteous attitude about reading. But we can try to find a way to understand the whole space of possibilities.
But there is also a real historical difference between Leavis and us. Fiction and the Reading Public registers, with some shock, the ubiquity of pleasure reading. Everyone reads novels, it seems to Leavis (McAleer gives us the best available statistics, which are not quite so high: 55% at peak ), and this changes the nature of the problem into how to get the “reading public” to read right. But not everyone reads novels anymore. This is the fundamental point in the review essay by Wendy Griswold et al. which I have assigned for our next meeting. Though the rich industrialized world has near-universal literacy, and nearly everyone believes reading more would be good for you, only a minority now reads regularly for entertainment; the era of the reading public has passed, and we have return to an age of what Griswold calls the “reading class.” What difference this makes for literary scholars’ attitude to low-status or unliterary texts—and to the literary as such—is one of the issues I hope we will take up next time. But whereas Leavis was worrying over a majority practice, we have only to do with a minority pursuit.
As for our other readings for next time, I hope you’ll continue to try to use your sociological and historical imaginations in making sense of the texts by I.A. Richards and F.R. Leavis. Both have very interesting biographies: from the DNB entry on Leavis I entice you with this detail: “On one occasion he was in trouble with the police and university authorities because he wanted Joyce’s banned novel Ulysses to be available for study.” On Richards, come for the anecdote about his missed vocation as a mountain guide, but whoops! he ended up an English professor (#humanitiescareers). If time had permitted, I would have supplemented the excerpts from Richards’s Principles of Literary Criticism with excerpts from his later, (in)famous book Practical Criticism (London: Kegan Paul, 1930). Glance at the first pages of his discussion of his undergraduates’ responses to the poems he gave them (without authors or titles or any other context).
You can also find many issues of F.R. Leavis’s journal Scrutiny on unz.org (which is great for scans of all kinds of old periodicals, so long as you can ignore the antivaxx stuff and other lunacy on the site). Say what you will about the contents of Scrutiny; you can’t beat the title.
Edit, Dec. 1: if time permits, please look at Browning’s “Pan and Luna” (1880)—never fear, only thirteen stanzas of ottava rima—which will shed some light on the interesting diagram in Richards’s Principles (116).