Poor but beautiful, a young woman living in a great capital city attracts the notice of a rich and aristocratic older man. After marrying him, however, she meets a younger man. Her husband suspects her, wrongly, of improperly dallying with the younger man and imprisons her. Later, convinced of his error, he seeks a reconciliation, but the couple separates. The husband falls ill and dies. The young woman happily marries the younger man.

Those are the bare bones of a highly conventional story. Despite its triteness, a few years ago the rare-book collection at the University of Pennsylvania Libraries paid 550 pounds -- approximately $687.50 at the prevailing exchange rate -- for the full version, told in The Husband's Resentment: or, The History of Lady Manchester. The purchase of this anonymous two-volume novel, published in London in 1776, caused no comment. It did not cost a great deal of money, nor would the book otherwise attract notice or criticism. In fact, apparently only one scholar of the English novel -- J. M. S. Tompkins -- has read it, and her 1932 discussion of the book was descriptive but unenthusiastic.

Now that fiscal constraints in academe have heightened cost-consciousness, some might consider $687.50 to be a bloated price for a novel that is almost totally unknown and whose plot is remorselessly cliched. On the other hand, one might argue that a rare-book collection should, every so often, buy a rare book. The Husband's Resentment certainly is uncommon. As far as I know, only Yale University, the British Library in London, and the State Library of Queensland in Brisbane, Australia, also own copies. [2000: According to ESTC, one more copy has since turned up. It is at the National Library of Wales.]

Nonetheless, the University of Pennsylvania did pay $687.50 for what many people might think is a piece of literary rubbish. What good is this book?

The Husband's Resentment may or may not be a better book than its neglect suggests (I argue that it is, in the Spring 1992 issue of Rare Books and Manuscripts Librarianship). But regardless of the value any particular reader or scholar may place on this book, the decision to buy it highlights the mission of research libraries in general and of rare-book collections in particular: to make books available to readers.

Questions about literary merits are abstract, at least in comparison with the hard physical fact of a work's presence on a library shelf. A book on a shelf is real: it gives readers the opportunity to discover it and to debate its worth for themselves. That is the value of acquiring The Husband's Resentment.

The book obviously is not part of the literary canon. Yet all of us know that the very concept of the canon no longer can be taken for granted. Who is to say what previously obscure or scorned book might some day be recognized as important, or at least as having some scholarly value? Today, scholars and their students are pushing the boundaries of the canon, or bypassing them altogether, by examining anew books that have not been widely read in the past -- works by women and minority writers, writers from geographically, politically, or culturally marginalized environments, and writers in genres normally ignored as "subliterary."

Where can these books be found? In libraries.

Most canonical authors in the Anglo-American literary tradition have been male, Protestant, and white; relatively well-off; politically centrist or conservative; and practitioners (as far as speculation permits) of culturally sanctioned forms of sexual behavior. Did scholars' standards for what constitutes great literature rest on an unstated, extra-literary assumption that the only people capable of producing great literature would be versions, however heightened, of themselves?

Yet as long ago as the Renaissance, the humanist and educator Juan Luis Vives was aware, as are many scholars today, that "the same book [may] seem . . . beautiful at one time and in certain places, and at another time and in other places, detestable."

The problem is, however, that instructors who question canons do so largely theoretically; for the most part, established canons tend to remain established, at least in the classroom. True, a few books by women or minority writers are added to reading lists; others drift in and out of print. New anthologies offer snippets of works by non-canonical authors, which teachers may occasionally assign. But the Modern Language Association's 1991 report, "Speaking for the Humanities," could find "no evidence that faculty members in English have abandoned traditional texts in their upper-division literature courses." As the literary critic Wendell V. Harris has written, "Academics tend to teach what they have been taught."

Of course, non-traditional scholarship has made some headway: an instructor who thinks that such writers as Susanna Centlivre, Aphra Behn, and Mary Astell have been victims of gender bias in the study of Restoration and 18th-century English literature will teach that period differently from instructors with more-conventional views. But sooner or later, students will be held responsible for knowing Pope and Swift; their need to know Behn and Astell will not be reinforced to anything like the same degree.

Libraries and bookstores tend to reflect the same conservative attitudes that have been institutionalized in college curricula. Relatively well stocked with books by 18th-century male writers, they provide far fewer works by 18th-century women. In libraries, such authors tend to be found in rare-book departments, in original or other early editions; modern scholarly editions of such works are few and far between.

Although a few modern publishers, such as Pandora Press, Virago Press, and Oxford University Press, publish the works of some early women writers, most publishers adhere to conventional canonical judgments in both the individual texts they choose to reprint and the writers and texts they anthologize. In 1987 a reviewer discussing an anthology of essays on British women novelists wrote in the Times Literary Supplement that most of the essays "require the special perspective and enthusiasm of feminist criticism to come alive." This view prevails.

Here is where libraries can live up to their mission. While scholars' arguments in favor of expanding the canon may prove largely theoretical in their impact, libraries' decisions about what to acquire and pre serve are quite concrete. The decision to spend limited funds on a first edition of either Henry Fielding's Tom Jones or The Husband's Resentment makes a real, physical difference in what is available to readers. That's not theory; for want of a better word, that's "power."

At least one instructor at Penn is considering The Husband's Resentment for classroom use. Her decision will be based on the availability of the work, persuasion that it might be worth reading, and the exercise of her own judgment. The combination offers this book a chance for a new life. Many other unknown, unread books survive on the shelves of all our research libraries. At least a few might repay the odd glance now and again, and justify the expense of acquiring and keeping them -- which is what librarians want as a result of the power they exercise.

Libraries have been exercising this kind of power for a long time. It may be easier to recognize as power when viewed from a great distance; consider the power that the librarians in Alexandria exercised centuries ago. They, together with contemporary teachers and scholars, were largely responsible for the peculiar pattern of survival of classical texts. It is not quite so easy to see that modern librarians exercise similar power as we decide what to de-acidify, what to microfilm, what to discard, and what to purchase and keep. But we do.

Once there were three known copies of The Husband's Resentment. Now there are four, increasing the likelihood that this novel will survive long enough to have a shot at entering someone's canon, at finding that ideal reader who will discern some merit in the book and convey it convincingly to others. That's an accomplishment. Not a large one, perhaps, but an accomplishment still. We should expect all our research libraries to serve scholarship in this way.

Once upon a time, the works of John Donne and Herman Melville, as well as books like Harriet E. Wilson's Our Nig, needed -- and received -- refuge in libraries. Now they are classroom classics, proving that, however slowly, ideas and canons do change.

This is not an argument for jettisoning the conventional canon. It is an argument for wide collecting, both antiquarian and current. Only those libraries with both the conventionally canonical and the laughably bad will enable the future to second-guess the past. Of course, we know garbage when we see it, just as those librarians of the last century knew it when they removed Charles Dickens from their shelves. Might our own best judgments prove to be equally suspect? Scholars, librarians, and academic administrators should realize that the distance between garbage and treasure is often smaller than we suppose. Your garbage may, just possibly, be my treasure.