The following essay was delivered at the MLA convention in Washington D.C. in December 1996 at a panel on the subject, 'Anthologizing Romantic-Era Writing: Shaping the Canon for the Commercial Marketplace'. It was chaired by Susan Wolfson, and the other participants were Jack Stillinger, Richard Matlack, Peter Manning, and Anne K. Mellor. The background to the paper is the publication of a number of teaching anthologies of the Romantic period in recent years, including my own Romanticism: An Anthology.
The title of this session is, especially for the literary scholar, particularly salutary. Scholars haven't traditionally need to think about the commercial marketplace, and there remains the suspicion that it's improper for them to do so. You have to remind yourself that teaching texts, arguably the most commercial manifestation of academic work, have been with us for centuries. At school, Wordsworth and Coleridge read Vicesimus Knox's enormously popular Elegant Extracts in Prose and Verse, the Norton anthology of its day. 
What makes the present situation distinct from any in the past is the peculiar way in which, in the last few years, anthologies have become a principal means of canon reform. Of course, they have always been one means of questioning or revising, however lightly, the existing canon. One thinks of the appearance of Wordsworth's The Ruined Cottage (MS D) in the Oxford anthology in 1973,  or that of the Two-Part Prelude some years ago in the Norton. But these were comparatively minor innovations. The discovery - or, in some cases, rediscovery - of the work of women writers, combined with a shift in the importance granted the work of philosophical and political writers, has generated an imperative for canon reform that is focussed on teaching anthologies.
All of which begs, it seems to me, a vitally important question yet to be answered - or, in fact, asked: should the anthologist provide teachers with texts of works already taught, or is it more properly their job to give teachers works that they (the anthologist) thinks should be taught? Put another way, is the teaching anthology the most appropriate vehicle for canon reform? Is it right to regard the classroom as the laboratory, and our students as the guinea-pigs, on which to try new canonical configurations and new critical ideologies?
I don't think, incidentally, that there's a yes or no answer to this question - all I'm saying is that I find myself faintly perturbed by the fact that no one has yet seen fit to ask it, let alone provide an answer. All the more odd, it seems to me, because in one way or another it must be one of the most urgent questions confronting the academic world today. Perhaps it is in the interests of publishers, who care only about profits, to discourage such enquiry; it is hardly in either our interests or that of our students to ignore it.
One related factor makes this question all the more important. Publishers have not been slow to recognize that the demand for teaching texts that reflect current critical trends has opened a gap in the market. Traditionally scholars have taken their projects to publishers; in this case, the plethora of new teaching anthologies is market-led. Publishers have seized an opportunity to take custom away from market leaders like Norton. And taking custom away from someone else is what this pastime is all about - at least from the publisher's point of view. In the case of Harcourt Brace they have decided to forestall competition from other publishers by competing with themselves. Alongside David Perkins' excellent English Romantic Writers (1967), which has served university students for many years, they have now published Anne K. Mellor and Richard Matlak's British Literature 1780-1830 (1996), which is designed, in due course, to drive Perkins out of print. Moral: if you want to keep up with the opposition, join the opposition. That, I suppose, is about as cut-throat as it gets.
Commercial considerations invariably count for more in decisions concerning textbooks than in those concerning monographs - of which, at an average of 200 pages, most university presses will count themselves lucky to sell 500 copies. At over five times the number of pages, the teaching anthology has to sell thousands just to break even. But there's no point in going to the trouble of editing, typesetting, printing, and distributing such a work if it's going to do no more than that; to make it worth anyone's while it has to turn a profit. Sales of 10-15,000 have to be a minimum target, and this raw fact of the business means that it's vitally important to provide an enormously large readership with exactly what they want - and at a lower price than the opposition. This is the game which, like it or not, I and my co-panellists are playing.
My anxiety has nothing to do with the propriety of all this; it is, rather, to lend a certain urgency to the question I asked earlier. What I'm saying is that commercial considerations will inevitably play their part in determining the contents of each different anthology: scholarly or pedagogical factors cannot be the sole determinants. This is nothing new; I take it that Norton has revised the contents of its anthology for years partly on the basis of what it thinks will sell. I'm really asking you, the audience, and the readers of these books, whether you want editors to give you texts of the works you already teach, or those of works which we believe should be there?
The question is, in a sense, naive. The answer lies in that most chilling statements of all - sales figures (which leads me to add that one of the most illuminating questions of all, were any member of the audience shameless enough to ask it, would be to ask each of the editors on this panel the details of their most recent figures  ). Chilling not just because of the obvious implications such numbers have for each member of this panel, chilling not just because there's a great deal of money at stake, and Jack Stillinger might not be able to afford his Gucci loafers in 1997, or I my two-week holiday in St Tropez - but because, particularly in the ruthless world of academic publishing, people's jobs are on the line.
You, who teach from these anthologies will answer the question I pose in the most direct manner possible, either by adopting, or not. And that, finally, is why I and my fellow editors are lined up before you like the identification parade in The Usual Suspects. Is the market sufficiently large to accommodate all the anthologies available? It seems doubtful, and not all of them may outlast the millennium. But before any of these worthy volumes drop out of print I'd like at least to pose the question to you the readers. Is the anthologist most properly the servant of an ideology, a market, or of the classroom?
See my Wordsworth's reading 1770-1799 (Cambridge: Cambridge University Press, 1993) 82-3.
The Oxford Anthology of English Literature: Romantic Potry and Prose, ed. Harold Bloom and Lionel Trilling (Oxford and New York: Oxford University Press, 1973) 130-42.
No one did, alas.