Go ahead. Keep working on your book. But before you finish, I hope you\’ll consider a modest proposal. Your book will never be entirely yours—it can\’t be. Because the book you\’re writing needs an important collaborator: the reader.

Today our paradigms for scholarly writing may never have seemed less paradigmatic, our disciplinary affirmations never less comforting. Yet professionalism remains tied to the idea of confirmation through writing.

It\’s unnecessary to rehearse the well-known conditions of scholarly production, circa 2013. A declining body of readers purchasing physical books, a growing number purchasing digital books. Publishing houses that once issued 1,500 copies of a certain kind of book now issue 300. The figure 300, which comes up a lot in conversation about print runs, has an unintended echo in the number of Spartans holding back the massed forces of Persia. Cue Gerard Butler.

And yet there\’s a paradox here. Even if scholarly publishing may be in trouble, readers—even scholarly readers—aren\’t. In fact, readers may never have been more important in the ecology of knowledge production.

Publishers must listen more today—and communicate more. The once vital archipelago of independent bookstores is now all but a memory. Its replacement—the Internet\’s biblio-rhizome—fosters an increasingly sophisticated culture of digital sampling.

\”Look inside\” was once the carnival barker\’s pitch outside the sideshow\’s tent. One imagines a 19th-century Parisian gasping in one of the shopping arcades to which Walter Benjamin devoted such attention: \”Look!\” \”Inside!\” In our modern-day digital universe, the invitation is a familiar direction to engage, however briefly, with the contents of a published book. With a click, the reader is offered the table of contents or the copyright page or the \”first pages\” (so knowing, so deliberately incomplete). It\’s a manufactured frustration that the publisher hopes will result in a sale, but that decision—as always in a commercial economy­—rests with the purchaser.

The \”Look inside\” feature of listings like those on Amazon is one proof that we have fully and completely entered the Age of the Reader. It is, of course, possible that we never left it, that readers have always defined the possibilities of writing. But the conditions of scholarly writing depend in new ways on the reader as arbiter and recipient.

You might think that having all that power makes things easy for our reader. But it doesn\’t. There\’s too much to read, to know, to search for, to copy and paste; too much to wade through to locate the right scholarly writer with the right scholarly message. Even a successful quest will often end in disappointment.

New reading conditions make old questions even more important: Whatever we write is for readers, or it\’s nothing at all. Do we write as if we mean to welcome readers? And what do we want to happen after the handshake and the hello?

The traditional model of scholarly production is based on scrupulous research methods and well-documented results. But that same model, especially as applied to the job of writing a first scholarly book, has also tacitly endorsed a principle of precise and modest display. Nothing too flashy, nothing too speculative. The reader\’s presence has respectfully been acknowledged, but not often actively engaged.

That\’s how we\’ve done things these past 30 years. But our writing protocols need to change now, in part because the digital environment increases potential visibility and exposure. Let me propose here a controlling metaphor: We\’ve been writing snow globes.

Every semester I conduct writing and publishing seminars with faculty at other institutions. I often ask if anyone collects snow globes. They can be delightful little objects, those glass spheres sitting contentedly on window ledges and dressers and desks. But they\’re not complex: an Eiffel Tower, a little house, a tiny snowman positioned on a white field, each encased in what looks like water. A small quantity of flaked material lying around Frosty\’s feet is easily distributed through the liquid with a simple shake. Turn the snow globe upside down, and you create a moment of theater.

The academic book—especially that first academic book—is often conceived of as a snow globe. It\’s carefully constructed to be a perfect little world, its main purpose to be admired. There\’s a glass wall that separates the contents from the reader. That construction is not accident.

Within the realm of the snow globe, every authority on the subject has been cited or pacified. Look inside and find a perfect, tidy, improbable world where no questions are asked, or invited. Scholarly books, especially first ones, are a paranoid genre—their structure assumes that someone is always watching, eager to find fault. And they take every precaution against criticism.

As we fear, so we write. Fearful writing is different from covering the bases. It\’s building a glass wall around one\’s project so that the reader can look at but can\’t disturb the pleasant scene within.

It\’s no surprise that the snow-globe book doesn\’t do much. It can\’t. There\’s nothing to grab onto, no space for the reader to get into the text and texture of the argument. There may not even be an argument, or a thesis, or a claim, or even a very good, almost entirely coherent half-an-idea.

Academe has been in the snow-globe business for years. The problem here is not the specificity of research but the intention of the finished product. Inward-looking, careful to a fault, our monographs have been content to speak to other monographs rather than to real, human readers.

What happens if we imagine writing—and I include scholarship here—as a tool rather than as an artifact? As a mechanism instead of a brainy objet d\’art? What if a scholarly book were, in other words, a machine?

To propose a machine model for academic writing invites a series of objections: The beauty of prose might be downgraded; the discursive quality of exploratory writing might suddenly have no place. There might be something suspiciously STEM-y about the machine model, as if the humanities and the narrative social sciences were trying to mimic science, the struggle of writing reducible to formula.

That\’s not what I\’m suggesting. But the machine model and the snow-globe model do point in different directions. The snow globe invites admiration. The machine waits to be deployed. The snow globe can brighten a windowsill or hold down a small stack of mail, but we\’d be hard-pressed to call what it does work. A machine, on the other hand, exists to do something or to allow you to do something with it.

That should mean that writing has consequence. Walk away from the snow globe, and nothing much in your life has changed (though you may find that you have lingering thoughts of Santa or your last vacation). Walk away from the book-as-machine, and something about your view of the world should be different.

That difference can take place only if the machine has been well made, and if you\’ve been an attentive reader. The book-as-machine should trouble or excite or inspire or even confuse. (Sowing confusion is, within a limited compass, a reasonable goal for a writer, just as long as it isn\’t the only goal.) The book-as-machine requires that the scholarly writer imagine a problem or concern that will engage the reader, making the investment of reading time worthwhile.

This is not the same as having a thesis or an argument. Those are author-centered positions. They\’re about what the writer thinks. The book-as-machine turns the spotlight onto a problem to be solved, and the reader for whom the problem is genuine, and genuinely interesting.

What I\’m suggesting is very much in line with one of the oldest teaching methods: posing questions for further discussion. It might be easy to feel superior to those final pages in popular editions of novels for reading groups (\”Do you think Newland Archer made the right decision?\”), but such questions ask the reader to think with and through the text.

An academic writer, of course, is unlikely to print up a set of discussion questions at the end of a manuscript. Yet implicit in the machine model is that the writer openly acknowledges that the book enables collective action. Reader, can you apply my theory to your own field? Can you take this book\’s idea and go further? Can you take what the writer provides and build what the writer could never have imagined? This is imagining one\’s writing as activism­—not necessarily political, but activism in the sense that it causes action in others.

So how can writing be, in a good sense, a mechanical contrivance? To consider writing as a machine for changing readers is to acknowledge that the power to persuade isn\’t restricted to the political stump or the pulpit or the agora. Something more needs to be at stake than a new adjustment to a theory or a sequence of facts.

I\’m advocating for a riskier, less tidy mode of scholarly production, but not for sloppiness. I\’m convinced, though, that the scholarly book that keeps you awake at night thinking through ideas and possibilities unarticulated in the text itself is the book worth reading. It may be that the best form a book can take—even an academic book—is as a never-ending story, a kind of radically unfinished scholarly inquiry for which the reader\’s own intelligence can alone provide the unwritten chapters.

Let every writer reflect on Rilke\’s famous line: \”Du musst dein Leben ändern.\” You must change your life. Books are life-changing for writers­—but often only for the scholars who write them. In the new order of scholarly production, let\’s double down on Rilke\’s dictum: You must change their lives, too.

Author Bio: William Germano is dean of the faculty of humanities and social sciences at Cooper Union. A new edition of his From Dissertation to Book ­will be published in the fall by the University of Chicago Press. You can follow him on Twitter @WmGermano.