The Web of Disorderly Erudition:
Electrifying the Eighteenth-Century Classroom

Jack Lynch

Delivered 3 April 1998 at ASECS in Notre Dame

Sit at my feet, children, and I'll tell you a tale of yore. The year was 1994. The world was a very different place back then -- younger, somehow more innocent. And out on the intellectual frontiers, a brave race of pioneers was staking out new territory in our late twentieth-century Wild, Wild West, the World Wide Web. I know. I was there.

It sounds absurd, I know, to look back wistfully at the early days of a four-year-old phenomenon. But the world of electronic textuality is changing rapidly, and I think we're already entering at least a second phase of Web literacy, one that brings new problems we didn't expect to face so soon. It's time to take stock.


Commentators say we're witnessing major changes in the forms of literacy -- McLuhan's wisdom on medium and message is often invoked -- and although predictions about the future of technology are almost always embarrassingly wrong, I'll wager we are in fact in for a sea change in the ways we teach and do our research. The most obvious locus of change is the sheer volume of materials on the Web. My pages on eighteenth-century resources comprise over sixteen hundred links -- a staggering number. Some of these sixteen hundred items are remarkably valuable: the Blake Archive at Virginia, for instance, will soon be recognized as the standard edition of Blake's works, and a number of other chronologies, topical encyclopedias, and bibliographies are as good as anything in print. The thousands of electronic texts from the eighteenth century, ranging from short lyrics to entire novels, include the non-canonical works many of us are trying to teach, and can help to free us from the tyranny of seventy-five-dollar anthologies that contain only half of what we want.

That's very exciting -- but sea changes often result in intellectual seasickness. One problem is a familiar one, given new urgency by the Internet. Because I maintain a number of literary Web pages, I routinely receive requests from strangers asking for help with their work. At its best, this is a really exciting development, and what many people mean when they talk about the Internet's democratization of scholarship. At its worst, though, it means lazy and dishonest students have a vastly expanded collection of resources for practicing their laziness and dishonesty. Strangers write every week asking me not to provide tips on their projects, but to write them (often appending messages like "this is due Thursday, please hurry"). Some get indignant when I refer them to books: they're not looking for things to read; they're looking for Web sites to pinch. The term-paper shops advertised in the back of Rolling Stone have nothing on their cyberspace equivalents.

But the biggest concern for teachers of literature in the electronic age is authority. Many of us will rejoice in a democratic medium like the Web, where anyone with twelve bucks a month can be a publisher with a larger potential audience than Gutenberg ever imagined. How liberating to live in a world where everyone's voice is potentially equal! But I worry about a medium in which everyone's voice is treated as being actually equal. We may frolic in the liberation from the institutional structures that silence less powerful voices, but let's remember that we make our livings by discriminating worthwhile scholarship from uninformed charlatanry. I'm a great fan of scholarly democracy, but draw the line at scholarly anarchy.

This isn't unique to the Net; there are great mountains of rubbish in print, too. But the problem is exacerbated on the Web, where there are not yet institutional mechanisms to distinguish the worthy from the worthless. Books are old and familiar; we know what to do with them. They say you can't judge a book by its cover. Poppycock -- you can judge most books by their spines. The print world offers institutional mechanisms which tell us that the material between these two covers has been subjected to certain kinds of review. The sober blue cloth of a Clarendon edition bespeaks one kind of authority; a watercolor of a lusty and pouting nymphette in Fabio's arms suggests quite another. We've all learned that publication in The Journal of English and Germanic Philology means something quite different from publication in The Weekly World News. But electronic texts?

An example. In the early stages of a large electronic edition of Frankenstein, the editor, Stuart Curran, and I discussed the best way to get the text of the novel in electronic form. We of course planned to edit it carefully, but wanted to know whether it would be easier to type the entire thing from the original, or to start with the public domain text on the Web and correct it. We settled on the latter as more efficient.

Had we known then what we know now, we would never have started with a text of unknown provenance on the Web. When we subjected it to a collation (through oral reading) with the original, we discovered an average of around thirty mistakes per page, which works out to something like one out of thirteen words. Most were minor enough -- silently regularized orthography and punctuation, for instance. But some words were inadvertently changed (as when Victor Frankenstein called on the names of his dead brother, where in the original text he calls on his manes), some words and sentences were transposed, and some words, sentences, and even paragraphs were missing altogether. Where this text came from, we have no idea. Perhaps many of the errors appeared in the print edition from which the electronic edition was taken, probably a late nineteenth-century edition with no authority. And many other errors were doubtless introduced in the process of scanning the text. (Some optical character recognition software packages boast of 99% accuracy, which sounds impressive, until you realize that an average of twelve hundred-fifty letters per page means over a dozen mistakes per page.) Whatever the source of the errors, though, this text was so bad that I could never in good conscience send one of my students to it.

No sheriff has yet arrived to clean up the Wild West of the World Wide Web, and the outlaws still rule Dodge. The Web so far offers no such mechanisms to allow readers to guess at the authority of the information found there. Most people in this room have been sufficiently trained to make an educated guess about the authority of a Web site after a few minutes' perusal: we look for the traditional hallmarks of scholarly respectability such as citations, standard editions, editors whose names we know, and so on. But how are we to guide our students through this maze? Russ Hunt put it well on C18-L recently, wondering how to deal with "the loonies and incompetents and freakish interest groups whose sites my students keep finding."


As it happens, these were central preoccupations in the eighteenth century, which was chock full of its own loonies and incompetents. The age of Johnson might have found much to agree with in McLuhan's famous mantra, and those who examined the history of the last major revolution in the technologies of the word were very conscious of the cultural transformation brought about by the technology. Thomas Warton, for instance, waxes rapturous over "an admirable invention, which was of the most singular utility in facilitating the diffusion of the antient writers over every part of Europe: I mean the art of printing." Johnson concurs. To "a gentleman [who] maintained that the art of printing had hurt real learning," Johnson replied, "If it had not been for the art of printing, we should now have no learning at all." And once the press had been invented and there was learning to be had, scholars began the difficult work of establishing reliable texts, which became almost an obsession in eighteenth-century Britain. Textual criticism is only one of many concerns we cyberjockeys of the 1990s share with the critics of the 1720s. But it is probably the most illuminating, and is certainly the easiest to discuss.

The common wisdom is that Shakespeare was first edited in the eighteenth century; Rowe's edition of 1709 is traditionally the first "real" edition. Of course that's not true; four folios and dozens of quartos appeared in the seventeenth century, and they all had editors of a sort, albeit anonymous ones. What we mean when we say that Rowe was Shakespeare's first editor is that he was the first one to do his hand-wringing over textual fidelity in public. This isn't the place for a thorough consideration of the eighteenth-century quest for scholarly authority, but it is worth noting that one of the great desiderata of the age was a means of distinguishing bad texts, which seemed to exist by the thousand, from good ones. In other words, this quest took the form of a preoccupation with rejecting bad sources, sniffing out the loonies and incompetents just as we find ourselves doing now. Lewis Theobald is an important figure here. He entered the Shakespeare quarrel in 1726 with Shakespeare Restored, an attack on Pope's edition. His stated goal is "retrieving, as far as possible, the original Purity of his Text, and rooting out that vast Crop of Errors" introduced by such incompetent editors as Pope.

Samuel Johnson is another big name in this history, because he was the first to apply the "stemmatic" method of textual criticism developed by Poliziano to modern vernacular texts. Poliziano's most important contribution to classical scholarship was genealogical philology, in which he established the grounds on which promising, even tempting, textual witnesses should be disregarded as sullied -- the eliminatio codicum descriptorum. Before Poliziano, critics accorded equal authority to all textual witnesses; it became a matter of simply counting up the number of manuscripts in which each reading appeared. (An aside: World Wide Web search engines such as AltaVista work on the same principle, ranking "hits" according to the number of times a phrase appears in each page.) But this sort of indiscriminate approach to texts was inadequate for Poliziano. Anthony Grafton explains the complexities of his stemmatic method succinctly:

Given three sources A, B, and C, all of which agreed on a given point; if B and C depended entirely on A for their information, should they be considered to add any weight to A's testimony? Poliziano insisted that they should not. ...

For him, the object is no longer ... simply to amass evidence, but to discriminate, to reduce the number of witnesses that the scholar need take into account.

Johnson is the first to apply this method to modern authors, and the first to reject derivative witnesses of an English work, as in his comments on Theobald's edition:
In his enumeration of editions, he mentions the first two folios as high, and the third folio as of middle authority; but the truth is, that the first is equivalent to all others, and that the rest only deviate from it by the printer's negligence. Whoever has any of the folios has all, excepting those diversities which mere reiteration of editions will produce.
Johnson's critical practice, as Malone and Steevens noticed, often fell far short of his critical theory, and he unhappily followed the eighteenth-century practice of using the textus receptus as copytext rather than an early folio itself. But his regard for editorial theory shows he was the first to internalize the humanistic lessons in the editing of vernacular classics.

I quoted Grafton's book on Scaliger, but his most recent book is perhaps more apropos. Grafton rehearses a curious history of the footnote, especially the eighteenth-century footnote of Bayle and Gibbon. In them, he observes, we see a new obsession with authenticating scholarship: footnotes become the means of demonstrating an author has done his homework. Footnotes amount to laying one's cards on the table, and serve to move the writer's claims to authority into the public domain, to make his or her assertions subject to outside scrutiny and reproducibility. In one of the countless learned footnotes to The Decline and Fall of the Roman Empire -- this one on the difference between citrus trees and citron fruit -- Gibbon criticizes one of his seventeenth-century scholarly forebears precisely for being unable to master this sort of scholarly authentication: "Salmasius," he says, "too often involves himself in the web of his disorderly erudition." Two centuries later, we're involved in our own Web, often erudite, but far more disorderly than anything Salmasius or Gibbon might have dreamed up and with texts that would send Theobald into paroxysms.


And so to the peroration. My work on digital scholarship should make it clear that I'm all for electrifying the eighteenth-century classroom. But at the same time, I want to warn others to be careful when playing with electricity.

What to do? I don't think it's possible or even desirable to keep these potentially misleading electronic resources away from our students, who will turn to them whether we approve or not. Better than fleeing is to go on the offensive, and to teach our students to think about these issues themselves. David Oakleaf writes: "Seems to me the issue of web sources and their validity invites raising as a teaching strategy; that is, students (and teachers) could assess electronic sources as they considered the analogous problem in Swift, Pope, et al. and their reaction to Grub Street."

We teachers need to give our students a more sophisticated sense of what goes on behind the scenes in the business of knowledge production and dissemination. It's in fact surprising how little our students know about things we take for granted. Many undergraduates, for instance, even senior English majors, have no conception of where a text comes from -- that there is such a thing as textual criticism. The "Note on the Text" at the beginning of a reputable, even if not scholarly, edition of a class text is a good place to start. Perhaps we needn't worry about teaching freshmen the minutiae of the rationale of copytext, but I hope that giving them some insight into the importance of early authorial editions, variants, eighteenth-century publishing practices, and modern critical and eclectic texts will soon be a necessary part of any literature class. Let's teach them Poliziano's lesson: to weigh, rather than to count, sources.

Weighing rather than counting is a good lesson outside of editing, too. We should take the time to teach them, however superficially, about the authority of modern scholarly sources, especially since they have to work through the same issues in their own papers. Few undergraduates in my experience really understand the importance of citation; for most, it's an arbitrary means of avoiding honor-code trouble. But we should explain that citation of sources is more than a way of dodging liability for plagiary: it's a means of presenting your scholarly credentials, and engaging in a more honest dialogue with your audience.

And perhaps just as important as raising our students' consciousness is raising our own. Whether we like it or not, our students have found out how much there is on the Web, and they're going to use it. We owe it to them, and to ourselves, to know what's out there, where it comes from. Much of it is sterling scholarship, and much of it wouldn't earn a C-minus in our freshman classes -- which means we can't afford a hands-off approach, since neither out-of-hand rejection nor do-as-you-please-ism will do. We need instead to engage these issues explicitly, to give our students some idea of the grounds on which we presume to talk about the texts we teach. Only then will we be able to do justice to our two constituencies, one squarely in the eighteenth century, the other entering the twenty-first.