Join us
feature

Surface Routines

How we read on the Web
November 28, 2008

Sign up for The Media Today, CJR’s daily newsletter.

Overload—the amount people feel compelled to know combined with the volume of information they have to sift through in order to know it—is perhaps the largest factor in the increasingly distinct difference between how people read printed material and how they read online. Faced with the reality of having two eyes, one brain, and what the latest count estimates to be one trillion Web pages, many people forego immersive reading of a handful of sites in order to skim the surface of thousands.

Although scores of academics study everything from how the number of hyperlinks on a page affects a user’s heart rate to how parents read e-books to their children, the new type of reading that the Web either drives or enables is here to stay.

Newspaper designers and editors have begun their own attempts to determine Web readers’ habits. It’s part of an effort to make newspaper Web content fit the pace and shape of the Internet, divine reader tastes, and determine how to bring an audience to their sites and make them stay awhile. But honing Web strategies can also be a process of exclusion. It has been argued that long-form narrative, in-depth analysis, and other time-consuming examples of newspapers’ strengths will not translate online—a rather dubious claim given the extremely varied Web content that exists today—but there is also little doubt that the Web is rapidly evolving, and it is impossible to predict, with any certainty, where that evolution will ultimately take us.

In 1997, Jakob Nielsen, who would later be dubbed “the guru of Web page usability” by The New York Times, posted an article on his site titled “How Users Read on the Web.” “They don’t,” Nielsen explained, citing research he had conducted that showed 79 percent of users scanned any new pages they came across, and only 16 percent read word by word. (The most common method of scanning looks something like an F-shape: two quick horizontal glances followed by a third down the left-hand side.) Nielsen is a usability expert whose goal is to narrow the gap between how people wish to use a Web page and how that Web page is actually designed. Toward that end, he works to streamline, perhaps even encourage, the behaviors that the Internet fosters, behaviors that some scientists and intellectuals fear herald a new, less desirable—even dangerous—kind of literacy.

Nielsen has found Web users to be engaged in a “ruthless pursuit” of bits of personally relevant information. To help them in this hunt, he recommends that articles written for the Web include “highlighted keywords,” “one idea per paragraph,” and “bulleted lists.” The mass of narratives and investigations that fill a newspaper are, in the eyes of Nielsen’s Web user, distractions to be hounded into the shadows.

Mario Garcia Jr., of the prolific newspaper design firm Garcia Media, has suggested that “a newspaper is something we read, but the Web is something we do.” I wondered out loud to Nielsen whether newspaper Web sites could draw an audience that is interested in something beyond all of this using and doing. “What about all the other things that people want or need?” I asked. “What about education?”

Sign up for CJR’s daily email

“The analogy with the school doesn’t work because you’ve forced the children to attend,” he said. “You can’t force anyone to go to a Web site so you have to make it more appealing to people. They’ll still get what’s good for them, but they won’t get the things that are just a little bit good for them.”

Nielsen is equal parts realist and optimist, a result of his belief that communication mediums exist in a kind of socialist harmony where each does its part to fulfill a different task. (He’s originally from Denmark, which may explain this.) He argues that we should accept the fact that the Web isn’t well suited for deep reading or narrative, but says that’s okay because other mediums, such as books or magazines, will always be there to fill that role. The flipside to his argument that different mediums needn’t fear one another, however, is a rather narrow view of what any given medium is allowed to be—the content of a declining medium (print newspapers) can latch on to a surging one (the Web), but not without being completely integrated into the new form. If Nielsen is right, narrative storytelling and investigative journalism, at least as they exist in print newspapers, will be stranded without a business model to sustain them.

Maryanne Wolf, the director of the Center for Reading and Language Research at Tufts University, thinks, unlike Nielsen, that the more scattered style of Web reading is not simply picked up when one opens a browser and left behind when it is closed. Instead, Wolf has argued that reading, an unnatural act to which we’ve forced our brains to become accustomed, in effect requires us to create a new circuit within our brains. She said that what and how we read determine the function of that circuit, and that the methodical way in which we read a book is more than an example of how our brains perform feats such as logic—it is the model by which we became logical thinkers. In other words, deep thinking is nurtured by deep reading. Call it the you-are-what-you-eat theory of cerebral input, and Wolf believes that the size of the Web is driving people to frantic, chaotic gorging.

Nicholas Carr, who wrote the much-discussed Atlantic cover story last summer employing Wolf’s findings, worries that the way a person reads on the Internet becomes the way he or she thinks in other facets of life—in a classroom, for example, or a voting booth. Elsewhere, Carr has suggested that the upshot of this shift could be the devolution of “contemplative man” to “flickering man,” at which point only the most monastic Web reader would be able to focus on deep reading within a newspaper’s Web site while the rest of the Internet idled in the background.

In 2007, the Poynter Institute released the results of EyeTrack ’07, the largest effort yet to decipher how readers look at newspapers, both in print and online. Five hundred and eighty-two regular readers of newspapers in four cities were selected for the study. Those who got their news primarily from the Web were placed in one group, the holdouts from print in another. Cameras tracked their retinas as they went about what for them was still a common routine.

The results, heartening to most but perhaps disappointing to those who subscribe to the doomsday theory of online reading, described something markedly different from the scenario recommended by Nielsen and warned against by Wolf and Carr. Web readers were more selective in the stories they chose, but once they found what they wanted, they read a substantially higher percentage of text than their print counterparts—a result that was true across all story lengths. Rather than running from words, Web users tended to be more textually based, and typically entered a story through a headline rather than a photo.

In fact, all of the differences between the actions of print and online readers in EyeTrack could be far more easily attributed to the navigational structure of a news Web site than to the mysterious force of a new medium. The Web readers weren’t frantic or scattered. Their heart rates, from all outward appearances, remained at a reasonable level. They were more or less just engaged members of a community, wearing cameras on their heads and turning pages with a button.

The study proved the obvious but still anxiously held point that the Web is capable of delivering stories of any length and complexity. It also proved that people are still interested in long-form content—even people who choose to read their news online. What the study was incapable of telling, however, is how a newspaper fits into the Web as a whole. Although participants were instructed to read as they would normally, the one trillion Web pages outside of their newspapers’ domain were off limits. No one stopped to check e-mail, cross-reference a favorite blog, or Google an interesting prompt or whim. Their banking had to be taken care of at another time. So, were the participants reading the newspaper on the Internet or simply on a screen?

The extent to which newspapers will be able to retain their most important traditional assets as they move online depends, in large part, on the answer to that question. Put another way: Does the Web drive people to a different type of behavior, or does it simply allow less fettered access to the desires they already had? If the former, the rules for attracting an audience would be completely different and any of the old ways of doing so that overlapped would be a matter of luck. It’s easy these days to get caught up in the fear that everything of import will be lost in the transition from print to online, but it seems extreme to think that a new medium would force everyone to start from scratch. Even Nielsen, who contends that the Web does indeed drive user behaviors, bends his own rules by posting long and sometimes complexly worded arguments on his own Web site. He does this because he knows that his readers have a level of education and interest in his topics that will compel them to stay on the page long enough to hear him out. He senses the Web’s power but doesn’t let it exclude ideas he wants to communicate.

The Web may well have influence on human behavior. But it seems more helpful, for our purposes, to think of the equation as one of human behavior influencing the Internet—in which case, there is no reason that the impulses that compelled people to read the print newspaper couldn’t exist in a similar way on the Internet. And those who fear the way people read (or behave or philosophize) online would do better to look to our education system and parenting, for example, to solve cultural inadequacies that the Internet, rather than encouraging, has simply made more apparent.

Has America ever needed a media defender more than now? Help us by joining CJR today.

Michael Canyon Meyer is a freelance journalist and former CJR staff writer. Follow him on Twitter at @mcm_nm.