At the request of a member, we are posting this essay by Jason Epstein, originally delivered as a speech at the Hong Kong Book Fair in 2008, and published in the Winter 2009 Authors Guild Bulletin as “Backlist Maestro: Mr. Epstein’s Dream Machine.” Editorial director of Random House from 1965 to 1995 and co-founder of The New York Review of Books and The Library of America, Epstein is now Chairman of On Demand Books.

Backlist is a publisher’s most important asset: titles that have covered their initial costs, earned out the authors’ advances, require no further investment except the cost of making and shipping the book itself and sell steadily year after year without advertising or significant sales expense. Without a substantial list of such titles a publisher cannot survive. The same can be said of a civilization, for the books that survive the test of time, books that are treasured and read year after year, are humanity’s backlist, our collective brain. I do not refer simply to the classics of our various traditions but also to more recent books, hundreds of which are published every year and join the backlist, if not permanently, at least long enough to move the process forward, and provide depth and complexity to our understanding.

Journalism is often called the first draft of history. Books are the second, third, and in some cases final draft. Books are the ongoing dialogue of the present with the past, the endless confrontation of the human mind with the problem of existence. Even those millions who may never read these books: even those who may never have heard of them could not survive if our collective backlists, our racial memory, the wisdom of our species were to disappear.

Since I joined the publishing business as an editorial assistant in 1951, I have been obsessed with the preservation and distribution of backlist, for I understood from the beginning two important truths about our business: the first is that publishing is not really a business at all, at least not a very good business. If it’s money that you want to make, go into a real business and take your chances. The second truth is that publishing is a vocation, a secular priesthood, for publishers are caretakers of our collective memory, indispensable servants to those other caretakers, poets, story tellers, librarians, teachers and scholars. The cultivation of backlist is not only our business but our moral responsibility.

In 1951, six years after the end of World War II, and two years after I graduated from Columbia College, I found a job at Doubleday, a large commercial firm with rather low literary standards. But it had large marketing and production facilities and even its own printing presses, so it was a good place to learn the business. Some six years before I joined Doubleday, Congress had passed a most enlightened law: the so-called GI Bill of Rights which offered a free college education to everyone who had served in the armed forces during World War II and qualified for admission to an accredited school. The GI Bill democratized American higher education and created, among other things, a mass market for serious books, for backlist books—a market that had never existed in such numbers before. My classmates were among the beneficiaries of this generosity, but the GI Bill did not cover the purchase of as many books as my friends and I wanted to own. We could not afford the expensive hardcover editions of the books we were eagerly reading. In those days paperbacks were mainly drug store thrillers and romances and reprints of last year’s best sellers—among them a copy of Norman Mailer’s The Naked and the Dead which I still own—but mostly they were ephemera. Deep backlist, however—what is today called publishing’s long tail —was available if at all only in hardcover.

So six months after I joined Doubleday as an assistant to an assistant I suggested to our editor in chief that it might be a good idea to publish, in well-made paperback editions, on acid free paper, the kinds of backlist books my classmates were reading but could not afford at hardcover prices. My idea was to stock these titles as paperbacks permanently in bookstores. Today we take so-called trade paperbacks for granted: their annual sales are in the billions of dollars. But in those days most American bookstores didn’t carry paperbacks. Doubleday’s editor in chief, a wonderful man named Ken McCormick, agreed that we should test a preliminary series of 12 titles. I selected the titles, negotiated for their rights, chose the format, supervised the cover design, worked with the sales department to develop a marketing plan, met the major booksellers, and in this way learned at first hand at the age of 21 that the mechanics of book publishing were essentially unchanged from the time of Gutenberg 500 years ago. This is largely true 60 years later. Publishers still order an edition from a printer, store it physically in a warehouse and deliver copies physically to bookstores where they await customers, and at the end of six months unsold copies are shipped back to the publisher and destroyed or resold as remainders. Even at Doubleday in the 1950s I wondered whether there was a more efficient way to publish books. It would take 30 years and the advent of digitization for me to find the answer. But I am getting ahead of myself.

I decided to call my new series Anchor Books in honor of the famous Aldine colophon, the mark of the House of Aldus, a great Venetian printer of the 15th century. Their colophon was a dolphin wrapped around an anchor: the dolphin for brilliance, the anchor for depth. The first series of Anchor Books was a huge success, not because the titles themselves were so desirable— individually some of the more esoteric ones might not have sold at all—but because the idea of what we were trying to do seemed right to my generation of college educated readers the way digitization seems right to young people today. Within a year or two the publishing industry had re-created itself. Every publisher now had a catalog of so-called quality paperbacks. What was called the paperback revolution changed the landscape of our industry as publishers searched their dormant backlists for likely titles to sell in the new paperback format. From a lowly assistant to an assistant, I had become a revolutionary hero. I decided to continue in that role.

From this experience I learned two lessons: first that book publishers are slow to seize the opportunities that history provides, for how else could a lowly editorial assistant, with six months’ experience, have recognized a perfectly obvious major new market for books that my older colleagues had not seen at all? And second, though I had taken a significant first step, there would be a long way to go before a truly comprehensive backlist of important books in all disciplines and languages could be assembled and distributed to readers wherever they might be. But this is what I dreamed of doing. A glaring example of the problem was our own American literature: a few individual works of America’s major writers—Moby Dick, The Scarlet Letter, The Wings of the Dove, The Red Badge of Courage—were now available as trade paperbacks, but the complete or collected works of American writers were not in print. This was a national disgrace, and with the encouragement of the great critic and advocate for American literature, Edmund Wilson, I decided to correct it.

By this time I had left Doubleday to join Random House, where I was offered an unusual arrangement. I would become an editor, and eventually editorial director, but I was also free to create my own businesses, provided they did not conflict with what Random was doing. Unlike Doubleday, Random House was a distinguished literary house, the publisher of Joyce, Proust and Faulkner among others, as well as a brilliant list of younger writers that included Truman Capote, Peter Matthiessen and William Styron. At Random House I found a home and remained there for the rest of my career. The first thing I did was make plans for what would eventually become the Library of America: a handsomely designed, clothbound series on fine paper to include the collected or complete works of important American authors. This was not nearly as easy to accomplish as Anchor Books had been because the series I envisaged could not be published profitably. Inventory expense for slow-moving, expensive hardcover titles would be prohibitive. I would have to create a nonprofit corporation with foundation support. This took rather longer than I had expected—25 years in fact—but I persisted. Today the Library of America is a national treasure, the permanent repository of nearly the entire backlist of classic American literature.

But by this time—the mid 1980s—publishers’ backlists had begun to shrink ominously. The reason was not hard to find. Starting in the 1960s and accelerating year by year thereafter, American families had been leaving the cities for the suburbs. With the great exodus to the suburbs, the downtown bookstores, with their vast, carefully catalogued backlist inventories in which readers felt that they might find whatever they wanted, began at first slowly and then rapidly to disappear, until today, from what had once been a thousand or so fully stocked major independent bookstores in America, only 50 or so, if that many, remain. Instead, suburban readers were now buying their books in shopping malls where the chain bookstores were paying the same rent as the shoe store next door and therefore needed the same turnover in the same limited space. This meant a radically compressed inventory of celebrity biographies, popular novels, picture books, trinkets and so on, leaving almost no space for backlist. To make matters worse, the Supreme Court ruled in 1979 that unsold inventory could not be expensed for tax purposes. Publishers responded by destroying their slow-moving backlist titles and deducting their cost as losses.

By the 1980s this shift from city to shopping mall turned the book business upside down. Whereas before we depended upon our backlists for year- in, year-out stability, now we were forced by the shopping mall chains to concentrate on seasonal bestsellers, mostly popular fiction, self-help titles offering financial, spiritual or diet advice, and celebrity and political gossip that might sell hundreds of thousands, even millions of copies in the first year or so, but relatively little thereafter. These bestsellers set the sales target for the following year, which publishers had to exceed to meet their ever-increasing costs. For the right to publish these potential bestsellers, not all of which by any means lived up to their promise, publishers could no longer count on their authors’ loyalty but were forced to bid against one another in auctions held by agents of popular writers, so that the eventual cost of acquisition often became more than the publisher could hope to earn back in royalties or even in gross sales, but we had no choice. The mall stores had no room for backlist; they needed turnover and publishers had to supply it. Soon smaller houses that could not afford to play this high-risk game merged with larger ones, and when the larger ones inevitably staggered they were bought up in turn by European conglomerates, which is where things stand at the moment.

You may recall that I said at the beginning that publishing is not much of a business, if it’s a business at all. This held true for the small independent firms like Random House when I joined it in 1958, when its total sales were barely six million dollars, and it is no less true for today’s conglomerates whose sales are in the billions. When the financial officers of these conglomerates decide to look for a better use for their capital, one may wonder if they will find a buyer for their American assets. What happens if they do not is hard to predict, but the possibility of chaos before order sets in again should not be discounted.

In the mid-‘80s, when I first became aware of this dangerous shrinkage of backlist, I created the Reader’s Catalog, a catalog the size of the Manhattan telephone directory from which readers could order books by telephone (the Internet would not be in wide use for another two years) from an annotated backlist of 40,000 titles: a traditional backlist bookstore in the form of a direct-to-consumer catalog and the precursor to online book selling. The Reader’s Catalog was an instant success. We sold several hundred thousand catalogs and many more hundreds of thousands of books to readers all over the world, proving once again that if backlist is available people will buy it. But the cost of handling individual orders proved prohibitive. The average order was $36 and the cost of packing and mailing, even after customers paid their share, was devastating. Three years after I launched the Reader’s Catalog, Amazon and Barnes & Noble started their own Web-based catalogs. I chose not to do this, knowing from my experience with the Reader’s Catalog that losses would be enormous, and they were. Instead, I looked ahead to the next stage. I wondered if there was a way to sell a large, even limitless backlist, without physical inventory. By the late 1980s the solution was in plain sight.

Digitization was now a practical possibility. This meant that publishers could store their digital inventory at almost no cost title by title and transmit it anywhere in the world as easily and cheaply as e-mail. Now multilingual backlists could be expanded almost without limit and sold to a radically decentralized market wherever Internet connections exist rather than merely to specific retail outlets. Readers in the United States could, in theory if not immediately in fact, access the entire contents of the Beijing Library while readers in China could access all the books in the New York Public Library. Digitization meant the end of the 500-year-old Gutenberg era: no more physical inventory, no more warehouse, no more shipping and billing, no more returns, and theoretically limitless backlists for sale almost everywhere—a true revolution, orders of magnitude greater than Gutenberg’s had been. With the traditional publishing supply chain eliminated, all that was needed was a machine at the customer’s end—a kind of ATM for books as I imagined it—to which the buyer could transmit a selected digital file that the machine would then automatically print, bind and trim as a library-quality paperback. This at last was the solution I had been seeking from the time I first entered the publishing business, a worldwide, multilingual, all but limitless backlist that could serve readers wherever they happened to live—in Shanghai, New York, London, Mumbai, but also in the remote corners of the world where traditional booksellers had never penetrated.

Nine years ago I delivered three lectures at The New York Public Library, which became the basis for my book Book Business. In one of those lectures I laid out the digital future as I saw it and published that lecture in The New York Review of Books. A reader called me to say that such a machine as I envisioned had recently been invented and could be seen in its inventor’s workshop in St. Louis. I went there, saw this machine and thought, “Well there it is: the end of the Gutenberg era with its physical inventory, costly warehouses, limited marketplace and declining backlist.” The machine was not much bigger than an office copier. As I watched, it received a digital file and reproduced it, cover and all, as a library-quality paperback, identical to the factory made version, in minutes, at less than a penny a page. It was a moment I shall never forget. I told the inventor that it was much too soon to develop his prototype because book publishers, cautious as ever, were not yet ready to digitize their backlists, and that a premature start is certain to be an entrepreneurial disaster. But I also said that when publishers showed signs of awakening to this historic opportunity his machine would revolutionize the industry. I said that we would keep in touch and we did.

Three years ago I decided it was time to move. My partner and I formed a corporation that we called On Demand Books, acquired the inventor’s patents and hired him as our exclusive consultant. We then built several handmade test models of his machine and placed them in various real world settings: The New York Public Library, the World Bank Infoshop in Washington, D.C., the Alexandria Library in Egypt, the University of Alberta bookstore in Edmonton, Canada, Northshire Bookstore in Manchester Center, Vt.—one of the great surviving independents—another in the devastated New Orleans library system. Another is now in Australia and others are being readied for a large Midwestern university library and the Blackwell chain in Great Britain. A smaller version is being designed for a large American bookstore chain, and our company has formed an alliance with Ingram Book Group, the world’s largest English language book wholesaler, which has assembled a very large inventory of digital titles.

Our machine is small enough to function in a library or bookstore or school or hotel. It is as easy to operate as an office copying machine. It prints and binds a high quality perfect-bound book in minutes at point of sale for less than a penny per page and trimmed to infinite sizes between 8.5/11 and 4.5/4.5 inches. We call our machine The Espresso Book Machine because, like the coffee machine, it delivers on demand one item at a time, selected by an individual customer, quickly, practically anywhere.

Five hundred years ago, in the German city of Mainz, Johannes Gutenberg combined several existing technologies to make a machine that printed uniform pages from movable type. Gutenberg was an ardent Catholic who made a living selling trinkets at religious fairs. He hoped that his machine could cheaply produce for all the parishes of Europe a uniform Catholic prayer book to heal the factional strife that was threatening the church in Northern Europe, a problem not only for Gutenberg’s church but for his trinket business. Before the invention of the press, prayer books were beautifully handmade, illustrated volumes created one at a time by monks and well beyond the budgets of ordinary people or most parishes. Now Gutenberg’s machine could put prayer books in everyone’s hands, rather like Mao’s “Little Red Book.” Alas for Gutenberg, his press had just the opposite effect from what he had expected. By making it possible for lay people to read for themselves the word of God, he overturned the authority of the priesthood as God’s sole interpreter and all but destroyed the church in Northern Europe. Gutenberg’s printing press would force open the gates to the modern world with all its wonders and woes.

Within 30 years, presses had been established in all the major European cities. Now Europeans could read for themselves not only the holy books but an increasing flood of secular works of philosophy, science and practical knowledge.

The Protestant Reformation, the Enlightenment, the philosophical tradition that foreshadowed the French Revolution and shaped our secular, rational, humanistic modern world were the direct result of Gutenberg’s press. No wonder Islam in the 15th century banned the press, as Gutenberg’s own church might have done had it not been preoccupied by its various schisms.

Today, as the Gutenberg era ends and the digital age begins, the future is as unclear to us as the momentous future awaiting them and their successors had been to Gutenberg’s generation. To predict the digital future to any degree of specificity would be rash. Nevertheless it is already apparent that reference books of all kinds, many of which are out of date the day they are published, need never be printed and bound again but will be updated and accessed online and read on computer screens, handheld readers, BlackBerrys and devices yet to be marketed, item by item for a fee or by subscription, or free, like the amazing if occasionally frustrating Wikipedia. This transformation is already well advanced and awaits only an entrepreneur to categorize and package these materials and offer them to subscribers at reasonable fees. But the books in which cultures are embedded—the ever refreshed backlist on which civilizations depend—will I believe continue to be read in traditional formats printed from digital files and bound at multiple sites world-wide by machines like the one I first saw in the inventor’s workshop nearly 10 years ago.

There is a place in the digital future for handheld electronic readers, comparable perhaps to that for audio devices, but their costs will have to be much reduced and their design simplified before they achieve the economy, durability, portability and convenience of a printed book, especially for those millions of new readers in tomorrow’s worldwide digital marketplace. A practical problem with electronic readers is copyright security. Music publishers may welcome the promotional benefits of file-sharing and in any case can do nothing to prevent it, while musicians can offset their foregone royalties with live concerts. Authors, however, cannot survive without royalties. Thus downloads to electronic readers will be constrained by software to prevent misuse. Buyers will not own the electronic book they have paid for: They cannot lend it separately from the machine to a friend, sell or give it to others or convert it to a physical book. This may sort itself out, somehow or other, but for books that embody the ancient and ongoing dialogue that constitutes civilization, the format of printed and bound books seems optimal, irreplaceable and uniquely protective of authors’ rights as well as readers’ needs. The digital world without physical books envisioned by some of our more extreme futurists seems to me an unlikely and highly undesirable prospect, a misreading of human nature and the nature of books.

As the digital marketplace evolves, the barriers to entry for publishers will fall, a process well under way in the United States, where thousands of independent publishers, including self-publishers, have already emerged. In this new marketplace traditional readers of genre fiction—women’s romances, men’s adventures, and so on—will be replaced by a new generation that will find such entertainments on the Web, along with practical advice and instruction in electronic formats yet to be conceived. As this transition materializes, publishers will increasingly have to distinguish themselves by the quality and durability of their publications, leaving ephemera to the new media. In the digital future titles will be accessed not only from such general repositories as Google and Amazon but from websites of related interest, so that books of Chinese poetry or American constitutional history will be posted and evaluated by experts on such websites as well as within annotated bibliographies provided by Wikipedia and similar reference sources. In this way filters will be created to separate books of value from the undifferentiated material that will inevitably accumulate in cyberspace, while publishers’ imprints will become increasingly meaningful as marks of quality and depth of backlist.

With each innovation, from mnemonic verse to written language to movable type to digitization, the extent of transmission and the range of content have been progressively broadened until now these extensions approach their utmost limits—the limits of the earth itself. Gutenberg put the Bible and a few religious texts in the hands of European elite. From this beginning there soon emerged the writers who gave the West the secular, experimental, skeptical, democratic culture from which our United States was hatched. None of this could have been foreseen by Gutenberg and his contemporaries. And nothing but the broadest outlines of a digital future can be seen by ourselves today. Yet it may not be unreasonable to extrapolate from our past a worldwide future of widespread literacy in which readers on all continents will one day embrace writers from all cultures as part of a common heritage, transcending but not obliterating traditional boundaries and local languages, an unimaginably vast and complex cultural transformation, both wonderful and, because we are human, terrible, as human history has been from the beginning.