Address given to Publishers Marketing Association University at the Marriot Marquis in New York City on June 1, 2005 by Otto Barz, Publisher of YBK Publishers and George Ernsberger, its Editor in Chief.

    This talk describes the structure of a traditional book publishing company and defines the roles played by each of the major divisions of responsibility in the average publishing house.

PMA University presentation, June, 2005

    Book publishing has changed dramatically in the past twenty years and spectacularly in the past five. Twenty years ago the new kid on the block was Desktop Publishing. Five years ago the most important change ever in publishing--then called the information superhighway, the Internet—was beginning to be felt. Today the Internet is a vast sales platform for books. Also five years ago, another fledgling technology, print-on-demand, began to change publishing; its affect is still to be felt.

     Desktop publishing was going to change everything in book publishing. It didn’t. Actually, it’s a misnomer. Desktop publishing should really be called desktop typesetting. It totally changed the face of the typesetting industry, but it hardly changed publishing at all. It has done little to change the way publishers think about or sell their books. Everyone is familiar with Quark XPress, but did it change book publishing? Not one bit. Typesetters today are vastly different entities because of it, but the process of publishing a book—making it public--has not changed.

     About five years ago one aspect of publishing really did change things; Amazon came along. It changed the way books are sold. Not the way they’re chosen, not the way they’re edited, not the way they’re typeset, and not even the way they reach the buyer, except that more books are shipped through the mails than ever before. What has changed is how books are sold. Amazon and the Internet are, together, the single largest bookselling venue in the world.

     The world—there’s another new concept. American publishers have always thought of themselves as worldwide organizations, but they can’t get their product out of this country cost effectively. However, American publishers are on the brink of becoming worldwide marketers. I’ll tell you more about this later if someone asks about how this will come to pass during the question and answer period.

     That’s some quick history. Let’s change gears and look at the structure of a book publisher. The first thing that happens is that a decision is made to publish a given book. Depending on the size of the house, that decision could come from a publishing committee, the publisher or an acquiring editor. Next, someone edits the book. That person makes it flow better. The book is rewritten to a greater or lesser degree, depending on the author’s technical writing skill. That editor, called a copyeditor, makes it consistent and grammatically correct. Fact checking is another function of the copyeditor, as is preparing the manuscript to send to a typesetter. Preparation for the typesetter is really a collaboration between the copyeditor and the designer. For instance, the copyeditor determines how important one kind of heading is over another and how headings will relate to each other, but the designer determines what those headings will look like.

     So, what’s a designer? Someone has to decide how big the type will be; how big the margins will be; how many pages will be required. Should the page number be at the top? How big? What should the chapter opening page look like? The title page? That all-important selling tool, the book jacket or cover—what should it look like? That’s what designers do and their ways haven’t changed much since before Gutenberg. As few as ten years ago designers sketched the look of a page with pencils on tissue paper. They drew the actual page as nearly representative as possible by tracing the fonts. These drawings were beautiful artistic renderings. When finished they often looked better than the actual printed pages. They were very costly and they were very inflexible—you don’t say, “Hey, how would that look in 10-point instead of 11?” Sample pages are now prepared using Quark or another typesetting program and show exactly what the final page will look like. They can be changed in an instant. However, that’s about the only technological change there’s been in page design. Yes, a lot has changed in cover and jacket preparation through the use of Photoshop, but not about the pages themselves.

    Okay, so someone has chosen the book we’ll publish. Someone has dusted it up and made it read right. And, to bring us to this point, someone’s decided what it should look like. Let’s get it into typeset pages.

    Typesetting changed little from the hot metal methods that evolved between Gutenberg’s time and the late 1900s when phototypesetting came into being. In that process, light is flashed through a film matrix exposing the character directly onto film and then the film was used to create a printing plate. That method hit its stride in the 70s and 80s with machines in the 80s getting an assist from largish computers. Then, through the 80s and 90s phototypesetters were replaced by personal computers when desktop publishing took over. And, for now, that is where typesetting remains. One person with a PC does the work that once needed enough equipment to fill a gymnasium and a dozen people to operate it.

    Nowadays book manuscripts are mostly prepared using Microsoft Word. The Word document is poured into a typesetting program like Quark XPress. Once there, the XPress operator creates style sheets—Quark’s translation of the designer’s intention for what the type pages should look like. After styles are applied the pages begin to look as the designer had intended them to. However, the computer is not yet able to deal with the nuances readers take for granted—unbalanced pages; an incomplete line at the top of a page—a widow; a lot of hyphenated line endings in a row; tables that break in an awkward manner, each of these and numerous other requirements must be adjusted manually by the operator (yes, the computer does claim to be able to deal with these things, but, often, not very well).

    Each step, from manuscript preparation to the typeset page, can introduce error. To find these the proofreader is called in (and, usually, the author too) and each page is read top to bottom. The changes called for are manually inserted by the XPress operator.

    Finally we are ready to print--provided, of course, that each of these steps has also been taken for the cover and/or jacket.

    Here is where major technological change is taking place. Shall we take the traditional path or the print on demand path?    

    The traditional path says, print at least 2000 copies of the book. (We’re talking about a garden variety  book, say 256 pages, 6 x 9, printed in black ink.) When printed we will put them in a warehouse and announce to the world that they’re ready to be bought. Wholesalers place their orders with the publisher. Booksellers place their orders with the wholesaler. The books are ultimately shipped to shelves, where they’ll rest for a while. Months later (sometimes years) the unsold books are returned to the publisher who refunds the full purchase cost including return shipping! It sounds crazy, and it is crazy for smaller publishers, but that method has worked for close to 100 years for the larger American book publishers..

    The print on demand pathway is way different. The pages are sent off to the press-- a computer database that sits ahead of a xerographic printer that prints pages in quite the same way as the computer printer on your desk. The pages are collated and bound into a cover—one book at a time and looking no different from any book you’ve ever seen before! The immediate thought comes to mind: “Oh my, how expensive is that?--printing one book at a time! Well, not much at all, and that’s what’s revolutionary. And, when weighed against the litany of costs in the traditional path, the cost savings are significant. Paying for inventory that gets thrown away is costly.

    This suggests a real change coming to book publishing. Actually, it’s not coming, it’s here and it’s being used in much greater volume than most people are aware. But this is just the beginning.

    Later, when we get into the question and answer section, which will be most of the program—an hour or so—don’t forget to ask how world distribution is going to occur. You’ve probably got a good guess by now, but let’s get on with our presenters. We’re going to keep them in the order in which a book goes through a publishing house. You’ve just heard from the publisher. The next thing that happens is what the editor does in choosing what is to be published, and in overseeing its preparation prior to hand-off to production/design. I am pleased to present my colleague at YBK Publishers, our editor-in-chief and resident expert-in-almost-everything, George Ernsberger, who will, in turn, introduce this team’s production/design expert, Mayapriya Long of Bookwrights Design, who will be followed by our marketing and distribution maven, Susan Doerr of Consortium Book Sales and Distribution. Each will give you their take on the changes that have or are occurring in the technologies they work with. I hope you’re writing down questions for later.

George Ernsberger on the editorial functions as affected by new publishing technologies.

    My opening remarks were very informal indeed; here’s what I meant:

    First, let me acknowledge—so I can then ignore them—some categories of texts for which the new technologies have, as the forecasters have been forecasting for some years, made books obsolete:

    I may be overlooking some, and it might also be that more categories will be found that will do well in one of those nonbook forms.

    But for narrative texts—texts to be read, rather than sampled, that require sustained attention, that tell a story or make an argument—the great technology is type on paper, lots of uniform sheets of it bound together—books. This is nothing to do with nostalgia. Easy on the eyes, lightweight, inexpensive books just work better than any alternative. The breakthrough to something new that works better than books for this purpose may be decades off or it may be minutes off—no sense committing to a forecast since we can’t imagine what it’ll be—but it’s off; it isn’t here.

Old is new again

    There is a new technology that we believe is very meaningful to the publishing of books, and that’s P.O.D.—the ability to print one book at a time economically. We do indeed do publishing differently as a result of the ability to print on demand, rather than create and manage inventory. And yet the truth is that we do it only a little differently, and only with the aim in mind of producing, by this different technology and methodology, something not different after all. It’s our belief that the readers for us to reach, and the authors for us to serve, are book people. Not some other, newer things that might be called data consumers and data generators. So what I have to talk about today is books.    

    Both readers and writers have things in mind as part of the definition of that word. We believe—publishers have always believed—that readers have more things in mind than they know they have; they know in their neurology, maybe you could say, as much as in their minds, some things that only real book production people know in their minds. They are learning things, and making judgments, about the book they’re holding in hand even before they begin reading: by its cover, obviously, but then by the pages: type fonts, heads—kinds of subheads, numbers of them, no heads—running heads—page density—all of that is telling them things about the book in hand: prose or something else? just to start with; but then: fiction or non? book to read, book to study, book to look things up in? All before they begin to read.

    And beyond that, people know—again, not necessarily consciously—they know when the book they’re holding is ill-made—slovenly, first, in a physical, production sense—and not only typos but loose lines, widows, unbalanced pages. A little later they may have a similarly subconscious reaction (or a conscious one, of course) to slovenly thinking or writing. They don’t always care; but they know, with some part of their brains, and at some level they do care about at least some of that.

    So we do our best to make them comfortable with our books in all those ways—to make them confident that what they’re holding, and trying to get something out of, is a real book—and can be trusted as one trusts books. And I don’t mind saying that our best is pretty good—we know pretty well what we’re doing because we’ve been doing it since just a few weeks after Gutenberg got his ISBN numbers—or so it sometimes seems to us—anyway, since long before POD.

    We want readers to believe that—but first, we want authors to believe it. We want real authors, and so we want to convince authors that we make real books. So—to risk becoming repetitive—we do that in the same way that publishers have always done it, except where POD allows us to do it differently.


    Editors have two functions, essentially: selecting, and then working on the manuscripts at the beginning of the publishing process.

    That first function—selecting—is done differently because we can, and not so much because we have to. We haven’t come across a book that an old-fashioned publisher—we call them print-run publishers—would publish that we wouldn’t. There may be such a category—anybody know one? but I don’t know what it would be. I mean, of course there are books that belong in big commercial publishers with their promotion/marketing capabilities, and there are books that belong in scholarly or university-press publishers—but even those books can sometimes be overlooked and find not only a home but a sort of incubator, in a P.O.D. house. Our contracts are not exclusive—and not only that, but the printer files that we produce might well be of use to another publisher.

    We can choose books of much more limited sales potential, since we’re not creating and managing physical inventory; but we still choose books that fit into recognizable publishing categories because we want our list to look like a list of books. You know, I suppose, that you can’t choose books—recognize and discriminate among and prefer some—according to any recipe. You’ve got to learn to like the kinds of books that you can publish successfully—to recognize by feel, I guess you’d say—viscerally—which ones you’re seeing have the right stuff. The publishing process has always been a feedback loop. Even without thinking about it editors learn to like stuff that will work in the market—and then in our publisher’s particular place in it.

    With POD publishing, that involves a different variety of stuff. And not only do you have to learn to like still different things, you can now go back to liking stuff that you once gave up because it wouldn’t sell as much as was required in print-run houses. Maybe some genre you gave up liking because it was too special—poetry, say, or science fiction criticism, or, maybe the history of custom-made motorcycles. Maybe some specialized sort of scholarship (we’ve just published a history—or an ethnohistory, the authors call it—of the Navajos and Puebloan peoples of the southwest). We’ve published—reprinted, actually—a book of psychoanalytic theory that’s used in classes in several training centers. We know people who teach and study in some of those centers, so we knew about the book and its continuing sales—which were only barely enough for the author’s previous, print-run publisher to keep it in print—keep space for in his warehouse—and he wasn’t terribly sorry to give it up once the author decided he’d rather have us do it. So that sells a few dozen books a year—and we know it will continue to—and that’s fine, in this publishing model. Hundreds rather than dozens would be finer, but dozens can be profitable—and in this model, beginning with this much smaller initial cost to produce a book, it’s almost as good to have several dozen titles, each selling dozens of copies, as to have a dozen titles, each selling many hundreds.

    But you do have to have them sell in the dozens, at least. At our level of advertising and promotion—essentially, zero—it’s easier said than done to get a book to sell even a few dozen copies; and in fact it’s entirely possible to have a perfectly decent book sell not even one dozen copies over years of life. (You can trust me on this—we’ve achieved that result.) So an author has to come to us with both a manuscript and a real, thought-through selling/marketing plan that convinces us that his book has a chance of selling two or three hundred copies in some reasonable expanse of time. Even that, as I say, isn’t automatic—it isn’t even easy—but it’s at a whole other level of magnitude from what the print-run houses have to think toward.


    I’m going to talk only a little about the second editing function—the  “editing” side of editing—because technology can’t make very much difference, there. It is possible—and more and more being done—to do certain kinds of editing—copyediting, essentially—on screen, in word-processing programs. I can do that and have, but it puts you in a peculiar, abstracted, somehow chilly relationship with the text—with the language—and with the author, too. It also, at least in Word, creates very messy, hard-to-manage—and hard to typeset from—files. There’s probably a way to work past that, to develop a more productive relationship with that process—I did that as a writer, changing from typewriter to word processor. Took months, but it was finally worth it. But so far I have found no compelling reason to do that with editing, so I can’t say much about it. Check with me in a year or so.


Click here to return to the home page.