When I ponder the future of history in this digital age, few things worry me more than the fate of books—and, with them, the kind of reading that books enable and require.

Hard though it may be to believe, history is one of the few remaining book-based disciplines in the modern academy. The natural sciences long ago committed themselves to journals as their principal form of communication. The social sciences have followed suit, and even most of the humanities increasingly rely on journal articles to convey their most important ideas.

It's not that colleagues in other disciplines don't occasionally publish books, and certainly not that historians don't publish journal articles. But books don't count for nearly as much in disciplines where one's productivity is now measured by the number of lead-authored articles one publishes per year that have high citation rates in high impact factor journals. (If certain phrases in the previous sentence seem like Greek to you, you have some catching up to do to appreciate how profoundly the academic research enterprise has been transformed since Eugene Garfield published his path-breaking article on "Citation Indexes for Science" in 1955. There's a really important work of history—a book, I hope—waiting to be written about the intended and unintended consequences that Garfield thus set in motion over the past half century.)

I once listened in considerable horror as a distinguished senior colleague in a neighboring discipline gave a workshop for graduate students in which he urged them to plan their dissertation research so as to generate what he called "least publishable units" (LPUs) so as to maximize the number of articles they generated. He then explained how they should deploy these LPUs into targeted journals whose prominence would increase the impact of citation rates that could be manipulated in still other ways. When promotion and tenure hinge on LPUs, citation rates, and impact factors (to say nothing of one's ability to win major grants), as they now do in the journal-based disciplines, young scholars and scientists do not launch careers by publishing books, and their mentors accordingly discourage them from wasting time on such publications.

In history, on the other hand, journal articles still serve mainly as stepping stones on the way to book-length monographs which are the capstones of research projects requiring many years to complete. Mentors in our discipline often caution young scholars not to publish too many articles lest these steal the thunder from the book that is yet to come. At many institutions, a published book is still the sine qua non for promotion and tenure, and articles are merely the icing on that indispensable cake. As a result, historians look like pretty strange beasts to colleagues in most other parts of the academy who have serious doubts about whether books are worth publishing at all.

History's longstanding commitment to books is thus partly at risk because the cultural norms of the academy have been shifting away from monographs for many decades. When deans and divisional committees hire and evaluate faculty members, historians must now offer 101-level tutorials about our publishing expectations because they are so alien to most of our colleagues. Myriad subtle feedback loops compound the problem by translating cultural norms into political economic pressures. The same deans who no longer understand book disciplines like history also have trouble understanding why their institutions should continue to support university presses or the books they publish. Everyone now knows that outlandish subscription rates costing many thousands of dollars for access to a single scientific journal are putting enormous pressures on library budgets, forcing institutions to reduce their purchases of other publications (like books). But we don't always recognize that the monopoly conditions enabling journals to charge such high rates are expressions of the power they now wield over the entire research enterprise. Reforms designed to end the influence of the commercial journal publishers by requiring free online publication of open-source articles do nothing to slow the movement away from books, which require sales if the editorial and marketing infrastructure required to create them is to survive at all. With fewer libraries buying fewer books, and fewer university presses publishing them, the challenge of getting published at all makes it ever harder for history to sustain its commitment to this form of communication.

Challenging as these academic changes may be, it is the digital revolution beyond the walls of the academy that represents by far the greatest threat to books and book reading—as well as their most likely salvation. As I noted in the first of these columns, the Internet represents a transformation in human knowledge no less sweeping than Gutenberg's invention of moveable type. Just as the printing press made obsolete the manuscripts of medieval Europe, so too have we seen "old media" give way to "new media" with astonishing rapidity. First to go were the encyclopedias, with Britannica losing ground to Encarta, and both losing ground to Wikipedia, far more quickly than anyone would have thought possible. The recording industry has never fully recovered from the invention of MP3 and online file sharing. Newspapers are reeling from competition with online news outlets and struggling to survive their loss of advertising revenue to the likes of Craigslist and eBay. (Did anyone realize how important the humble want ad was to sustaining local journalism in the United States?)

The net result of these myriad changes is that the average American—including the average historian—is spending ever increasing amounts of time absorbing information from screens as opposed to pages. I'm not sure the total amount of reading has diminished—we're devoting so many hours to reading words on screens that even television viewing appears to be declining in consequence—but the nature of what we read has changed quite radically. Images, sounds, and videos have become more and more tightly integrated with texts to a degree inconceivable before the advent of digital tools. HTML and the Web have created nearly infinite links among texts that are no longer trapped within the static confines of the printed page. We now take it for granted that we can leap from text to text as our curiosity requires, with many clicks per minute carrying us hither and yon in pursuit of our ever-shifting interests.

The interconnectedness and frictionless movement enabled by the Web have made reading a far less linear experience than it once was, just as the inventors of hypertext intended. Readers arrive at a web page via millions of possible corridors mapped by search engine algorithms. They quickly scan its contents to answer questions or seek products or find useful links, and then scurry on to the next stop in their textual ramblings. The time they spend on any particular page is typically measured in seconds rather than minutes, and this places a premium on content that is arresting, efficient, and above all concise. The tendency of the Internet has generally been toward ever smaller units of discourse. The most effective blogs are typically one to three paragraphs in length, and Twitter—with its 140-character "tweets"—currently stands as the extreme symbol of this movement toward brevity.

The challenge for books, then, is not just that most are still printed on paper when the world is moving so quickly and irresistibly toward pixels on screens. Imperfect though they still are, e-book readers (about which I will have much more to say in my next column) may yet enable books to survive in the digital age. The much bigger problem is that the long arguments and narratives on which the best history writing has always depended require many pages—many screens—to be absorbed, understood, and appreciated. More important still, they require well-stocked minds with the patience and discipline to pay attention for many hours to complicated webs of actors and actions, causes and effects, events and contexts, ideas and meanings, without which we cannot hope to make sense of what happened in the past or why it mattered. Good history needs time and space to be grasped in all its richness. If journal articles aren't long enough to do the job, then what are we to do if blogs and text messages and tweets are the media our audiences prefer to read?

Please do not misunderstand me. I embrace and celebrate the digital age. I believe historians should use blogs and tweets, Wikipedia entries and YouTube videos, web pages and Facebook postings, and any number of other new media tools to share our knowledge with the wider world. But I also celebrate complicated arguments that need space to develop and patience to understand. And I love long stories that can only unfold across hundreds of pages or screens. What I most fear about this new age is its impatience and its distractedness. If history as we know it is to survive, it is these we most need to resist as we practice and defend long, slow, thoughtful reading.

William Cronon (Univ. of Wisconsin–Madison) is the president of the AHA.