Too Big to Know

For example, try reading (or, if you are of a certain age, rereading) David Halberstam’s award-winning classic The Best and the Brightest, published in 1972 as the United States was still dropping tens of thousands of tons of bombs on Vietnam.¹ Halberstam attempts to explain how the Kennedy White House, so full of superbly educated, dedicated men, could have failed so badly in Vietnam. The book’s world is populated by household names now known in few households: McGeorge Bundy, George Ball, Chester Bowles . . . the events it discusses are distant, recalled most often as an analogy to our worst current mistake. But Halberstam’s question remains deeply unsettling: How did the best and the brightest get us into the Hell of Vietnam? If these men, so well educated and so worldly, erred so badly, how can we trust the advice of lesser men?

Victor Navasky, in his 1972 review of the book, quoted Halberstam’s answer: “[T]hey had, for all their brilliance and hubris and sense of themselves, been unwilling to look and learn from the past.” Navasky, however, thought the problem was more institutional than individual. If the “good guys” in Halberstam’s account had been put into positions of power, perhaps they too would “have been swallowed by the war machine,” as Navasky put it.²

But the modern reader returning to a land that seems as distant as Oz is amazed that another problem isn’t front and center in any discussion of what went wrong in the White House. The best and the brightest are indeed brilliant and well-intentioned. They are hard working. They are patriots. But when you read the book today, you are immediately struck by their sameness. Male. East coast. White. Early middle-aged. Prep school followed by an Ivy League school. Not 100 percent homogenous, but close enough that if they were the board of directors of a corporation today, there would be lawsuits. Had the inner circle included experienced soldiers, old State Department field hands, or maybe even a Vietnamese or two, the White House policy-makers might not have gone so disastrously wrong about the problems they were facing.

Everyone is for diversity these days. Since the days of The Best and the Brightest, even our standard stock photos have changed. If a corporate brochure shows more than three employees working together, at least one of them will be a woman and at least one will be African American. While it is perfectly reasonable to criticize a department for not being diverse enough, it takes extraordinary circumstances to make the sentence “I think this group is too diverse” acceptable or even sensible. Diversity as a goal has become the default, in part as a matter of simple fairness, but in part because our culture has long accepted that a diversity of beliefs leads us to better, stronger, more grounded ideas.

Then we look out at the diversity of ideas on the Internet. At AboveTopSecret.com, the forums are filled with people debating whether an airplane really hit the Pentagon on 9/11. Search Google for who killed JFK and, as of this writing, the top results—WhoKilled-JFK. net—puts scare quotes around the phrase “The ‘conclusions’ of the Warren Commission.” If you want to go back not years or decades, search the Web for information about who wrote Shakespeare’s plays. After that, you can begin to look for the truly crazy stuff—the people who deny that there are atoms, that the universe is big, that germs cause disease, that two plus two equals four. And all that is just in the English-speaking world. There is so much radical disagreement that we have to rule out the worst of it as beliefs held by the insane—or by cultures that we uncomfortably give up on judging because otherwise we’d have to judge them as insane.

Noveck is all for creating networks of experts that look beyond the usual credentials. She gives an example: “If I want to figure out how to get benefits to veterans faster, maybe I should be talking to the people who are on the front lines of the organizations giving out those benefits.” ³ She adds, “Do we think about expertise as experiential or academic? Books or context-centered? The person driving a truck every day or the logistics expert from IT [the information technology department]? The answer is both, of course. Now the technology lets you find experienced people as easily as credentialed ones.” This opens up the range of opinions and ideas. Plus, Noveck says, it “taps the enthusiasm of someone who is totally self-taught.”

So, yay for diversity. But Noveck was in a high-pressure environment with a four-year window for getting things done. She is thus quite blunt about the point at which diversity becomes a problem, not part of the solution. For example, there are federal regulations governing the more than one thousand Federal Advisory Committees (FACs) to ensure that they include a diversity of opinions. Noveck is all for diversity when it makes sense, but as the process gets closer to making actual decisions, it is less and less helpful: If an administration needs help deciding what to do about, say, endangered species, “I don’t know that having creationists and evolutionists on the same committee is the best way to get advice.”

This is not a hypothetical for Noveck. When she arrived at the White House and quickly put up a public site to solicit proposals for an official open government policy, thousands of people contributed their ideas and discussed the ideas of others, almost always seriously and sometimes in great depth. Yet, the “tag cloud” told a different story. Tags are labels that participants can apply to comments so that they can be more easily sorted and found, and a tag cloud displays those labels in a font size that reflects their frequency. Among the largest tags: “UFO.” The “birthers” (people who believe that Barack Obama was not born in the United States) were also out in force, as evidenced by the use of the tag “usurper.” Their inclusion in a discussion of open government policy is an example of diversity taken to the point of diversion. Few of us would object to shutting the site to those who want to talk about something completely off-topic, but Noveck had a particular problem in this case. Because it’s a government site, removing irrelevant posts would certainly have been claimed by some to be an act of political censorship. So, she had to leave the irrelevant comments up. Fortunately, the community moderation features meant that users could flag posts as off-topic while still leaving them available.

First, there is a right degree of diversity. Too little diversity and you end up thinking it’s a great idea to invade Vietnam. Too much diversity and you have citizens haranguing you about Hawaiian birth certificates when you’re trying to come up with standardized formats for open government data.

Second, what counts as the right degree of diversity is highly context dependent. If you are holding a forum on the legitimacy of the Obama presidency, you want to make sure you have multiple sides represented. If you are organizing a rally to protest the legitimacy of the Obama presidency, you don’t want to invite to the planning meetings people who are dead set against you politically.

This is quite a general problem, not just one faced by policy-makers. When Ford was deciding whether to drop its Mercury line of cars, it didn’t invite in people who think that all cars should be immediately dumped into the ocean to form new reefs. But if Ford relied only on the same old folks with the same old perspectives and stock options, its decisions would likely be as realistic as those of JFK’s best and brightest. The same is true in every field from education to science, although there’s variability in the amount of diversity that’s helpful before it becomes merely disruptive.

1. Not all diversity is equal

Scott Page, in his book The Difference, presents evidence to support his bold claim that on or off the Net, “diversity trumps ability.”⁴ His explanation of this phenomenon is actually quite straightforward:

The best problem solvers tend to be similar; therefore, a collection of the best problem solvers performs little better than any one of them individually. A collection of random, but intelligent, problem solvers tends to be diverse. This diversity allows them to be collectively better.⁵

While “the homogeneous collection may just as well contain only a single person,”⁶ Page lists four conditions that have to hold if diversity is to be a better strategy. First, the problem has to be difficult enough that no single problem solver always comes up with the right answer; otherwise, you just need that one brilliant problem solver. Second, the individuals in the group have to be smart relative to the problem; if it’s a calculus problem, a diverse group of people who don’t know calculus is not going to outperform a single calculus expert. Third, the people in the group have to be able to provide incremental improvements on proposed solutions. Fourth, the group has to be large enough and has to be drawn from a large, genuinely diverse pool. Given these four conditions, it’s better to have a diverse group than a group that consists only of the very best minds—diversity trumps the best and the brightest.⁷

But what type of diversity? Getting a group of people who are diverse in their shoe sizes wouldn’t help much. Nor does it help to diversify solely by ethnic or racial identity, even though that’s how many businesses try to achieve diversity. This is the land mine that Supreme Court Justice Sonia Sotomayor stepped on (or was the hand-grenade lobbed at her, if you prefer) during her confirmation hearings. In various speeches between 1994 and 2003, Sotomayor said that “a wise Latina woman” might come to better conclusions than a white male judge. According to her detractors, she was asserting the superiority of racial or ethnic identity. According to her supporters, she was referring to “the richness of her experiences.”⁸ Diversity of experience might help to open one’s mind to unexpected ideas and to increase one’s sympathy for a wider range of people—traits presumably relevant to the sagacity of the nine-person committee to which Sotomayor was applying. But mere diversity of ethnicity is not: Imagine if Sotomayor were a Hispanic whose life experiences were identical to those of the other sitting judges. The racial, gender, and ethnic diversity on which businesses so often focus turns out to be irrelevant except when the individuals have experiential diversity because of those factors.

The Difference presents research showing that the sort of diversity that makes a group smarter than its smartest individuals is a diversity of perspectives and heuristics. Perspectives are the maps we give to ourselves to represent the lay of the land. For example, if the issue at hand were how to manage comments at an open government site, one perspective might organize the comments by topic, another might look at how they map across the political spectrum, yet another might cluster them by their emotional tone; each of these maps would lead to different approaches to managing the comments. Heuristics, on the other hand, are the tools we bring to bear on problems. For example, a heuristic might be that contentious discussions calm down when there are moderators, or that reputation systems (users giving a thumbs up or down to comments, perhaps) can marginalize the irrelevancies. According to The Difference’s analysis, a wise Latina woman will make the Court more diverse in useful ways only if her experiences have given her a different way of looking at the world (perspectives) and taught her different techniques (heuristics) for dealing with issues.

2. Have just enough in common

Mae Tyme is an anti-pornography feminist. Annie Sprinkle is a “prostitute/porn star turned artist/sexologist.”⁹ Unsurprisingly, they hold very different views about pornography. Those differences are absolute on this topic. Mae Tyme is the pseudonym of a lesbian separatist who believes that “all women that do pornography are either terribly misinformed, or have been enslaved.” Tyme would be kicked out of her community for talking to a “porn star,” according to Sprinkle.¹⁰ Yet, in 2000, Sprinkle and Tyme sat down for a conversation, which Sprinkle then posted on her Web site. It is a frank interchange. Sprinkle argues that porn can be “liberating” and goes so far as to recommend that people make their own porn in order to learn more about sex. Tyme says that pornographers engage in a form of child slavery, and even compares Sprinkle to a Holocaust denier. Despite the starkness of their differences, the conversation ends with each saying why the other is “a good teacher for me,” and hugging.¹¹

This seems to be evidence of the power of diversity: Two people completely opposed on an issue nevertheless have a respectful conversation in which each learns from the other. Tyme learns that female porn stars can find the profession “liberating and profitable”—and that the women are generally paid more than the men. Sprinkle learns that opposition to pornography does not have to stem from a fear of erotic experience. But mainly they both learn to respect each other despite their deep differences on the issue of pornography.

First, the difference between Sprinkle and Mae is overshadowed by what they have in common. Yes, they are diametrically opposed on the issue of pornography. But they are both committed feminists who have organized their lives around that core value. They also share a vocabulary: the words “porn,” “radical lesbian,” “dyke,” and “patriarchy” have precise enough meanings for them that they can use them without first defining them. Many other conversations would have been derailed by wrangling over these issues: “How dare you call me a porn star! I’m a sexually honest actor in the adult entertainment business!” In addition, from the conversation we can safely assume that she and Sprinkle are both women, speak English, are over forty years old, and are able to talk frankly about sex and sexuality. Take these away, and the conversation would have become far more difficult, or even impossible.

Second, the very last thing Mae Tyme says is: “A conversation like this is possible when each of us has freedom of expression and no one is required to change. I don’t expect you to become anti-porn, and you don’t expect me to become pro-porn.” Had the discussants insisted that there be a winner and a loser—or if they were on a committee charged with setting a policy for a business or the government—the conversation would have been far more difficult. Tyme and Sprinkle entered the conversation with an implicit understanding that the goal was simply to explore an alternative view held by someone they already like and respect.

Unfortunately, the sorts of conversations we often have in mind when we think about civility are precisely ones in which people who strongly disagree need to come to agreement about some practical matter of policy. But diversity works best when there are shared goals, as Scott Page points out. That’s why Beth Noveck doesn’t want to invite climate change deniers into the discussion of practical steps to slow down climate change. It’s not because she’s close-minded. It’s because the goal of that discussion is to slow the warming of the earth. Among people who share that goal, there is plenty of room for diversity of perspectives and heuristics. A diverse group of people who share a goal are likely to be more effective than a homogeneous group of people. Communities of knowers need walls around them. Those walls used to be like those of a fortress. These days, they tend to be usefully semipermeable. But they are walls nonetheless, and serve the good purpose that walls of every sort serve: permitting a group with enough in common to get something done by keeping out disruptive diversity.

3. Mix well by hand

A cookbook with a soufflé recipe that tells you to add “just enough milk” would be worse than useless. The same is true of a book that tells you to add “just enough diversity.” How much is enough? We know that it’s considerably less than we thought, but the problem is that there is no set amount of diversity that is just enough.

That’s why The WELL has moderators. Founded in 1975 by the generational icon Stewart Brand, with Larry Brilliant, The WELL has been one of the longest-running conversations on the Net. Its origins are in the hippie culture of which Brand is an avatar—the name stands for The Whole Earth ’Lectronic Link, a reference to Brand’s Whole Earth Catalog—but the 4,000 current members seem to reflect more of an earnest coffee-shop culture than the shirtless non-linearity of Haight-Ashbury. Jon Lebkowsky, who has been on The WELL since 1987, says that the site’s success was not accidental. “They were successful in building the community by seeding it originally with people who were great conversationalists,” waiving the fees for the people they wanted involved. Lebkowsky adds, “They also invited the Grateful Dead crowd, and a lot of journalists.”¹²

After the seeding comes the gardening. Every “conference”—a topic of conversation, some of which have been running since The WELL’s inception—has two “hosts,” or moderators. Moderators are allowed to participate, but they are there primarily to keep the conversation just diverse enough. If someone becomes uncivil beyond the norms for the group, the moderator may step in. On occasion, people are banned from the discussion for a cooling-off period—the conversational equivalent of a time-out. If the conversation steers off course, the moderator may remind people what they are there to discuss. Howard Rheingold, one of the founding parents of online discussion and a denizen of The WELL since 1985, urges community forums to have moderators. Even the mere presence of moderators—even if they never moderate a single posting—is enough to keep out the trolls, he says.¹³

Moderation does not have to occur through designated moderators. Sometimes there is simply too much traffic to make that feasible. Community moderation frequently does the trick—as at Beth Noveck’s OpenGov site, where the group of people actually interested in open government policy moderated the “birthers” into their own corner. But moderation almost always requires dedicated people for the same reason that the right amount of diversity cannot be specified beyond “just enough.” Human conversations reflect every dimension of human individuality, sociality, and culture. There isn’t a single rule that we wouldn’t want broken at some point. No foul language in a parenting discussion? Okay, but how about when the discussion turns to what to do about a potty-mouthed child and it becomes important to know exactly what words are being used? No hate speech in a political discussion? Okay, but how about when the discussion is about the language being used at some more hateful discussion site? English-only? Fine, until someone asks for help translating the angry Spanish remarks of a customer who refused to speak English. Stay on topic in a scientific discussion? Great, until a desperate parent apologetically interrupts with a question about a sick child. We need moderators because conversations cannot be entirely rule-based. How much diversity is “just enough” depends on the people, the topic, the purpose, the social bonds—on everything.

4. Fork it over

That’s what greets you when you browse over to the Los Angeles Times’s “wikitorial” experiment. Back in mid-2005, when the success of Wikipedia was whipping up Wiki Fever, it seemed that even the most contentious issue could be resolved if people were allowed to work together on a webpage devoted to it. So, when the LA Times ran an editorial called “War and Consequences” asking the Bush administration to clarify its plan for withdrawing troops from Iraq, it followed up with an invitation:

How do you like the above editorial? A lot? Thanks! Not so much? Do you see fatuous reasoning, a selective reading of the facts, a lack of poetry? Well, what are you going to do about it? You could send us an e-mail (or even write us a letter, if you can find a stamp). But today you have a new option: Rewrite the editorial yourself, using a Web page known as a “wiki,” at latimes.com/wiki.¹⁵

Four days later, the LA Times ran the following wiki-bituary:

The Los Angeles Times has canceled a novel Internet feature that allowed readers to rewrite an editorial on the newspaper’s website, after some users sabotaged the site with foul language and pornographic images.

The newspaper launched the experimental “wikitorial” Friday and killed it early Sunday after an unknown user or users posted explicit photos.¹⁶

Within two days, the original editorial had been edited 150 times. At one point it was turned into an editorial critical of the role that the LA Times had played in the run-up to the war. Comparisons to the Philippine-American War were inserted by some people and removed by others.¹⁷ And then, of course, there were the disgusting images repeatedly posted by vandals.

Jeff Jarvis, an important voice for openness in the debate about the future of journalism, blogged that “[a] wikitorial is bound to turn into a tug-of-war” and suggested that an alternative wiki page be set up for those who disagreed with the editorial. The founder of Wikipedia, Jimmy Wales, responded that he had already done so, creating a “counterpoint” wiki on the Los Angeles Times site for those who differed from the newspaper’s view.¹⁸ “I’m not sure the LA Times wants me setting policy for their site,” wrote Wales, “but it is a wiki after all, and what was there made no sense.”¹⁹

No sense at all. Wikis try to get everyone on the same page, quite literally. But when a diversity of passionate opinion is inevitable—no editorial has ever been powerful enough to dispel all contrary ideas—a wiki is exactly the wrong idea. A better idea is to enable the discussion to fork. Forking is a familiar tactic. For example, when on a mailing list a few people start flinging emails back and forth on a topic of marginal interest, someone might sensibly suggest that they take it off the group and carry the conversation through private email. At The WELL, under those circumstances, a moderator will suggest creating a separate topic thread. In the real world, discussions fork when people break off from a group to go talk among themselves.

The Internet is perfect for forking. It’s got infinite space where the divergent conversation can continue out of earshot of everyone except those who choose to hear it: If it’s a mailing list, you don’t get the forked discussion’s emails unless you ask to participate. Forking enables a group to find its own level of diversity.

If the Net is creating more echo chambers, the biggest loser will be democracy, for the citizenry will be polarized and thus be less able to come to agreement, and to compromise when it cannot. This is perhaps the greatest concern expressed by Cass Sunstein, a constitutional scholar and currently the administrator of the White House Office of Information and Regulatory Affairs. Sunstein, who is the most-cited living legal scholar in the United States,²⁰ has written a couple of books on the topic. In Republic.com, published in 2001, he argues that when people get to choose what they see, they will tend toward that which is familiar, comfortable, and reinforcing of their existing beliefs, a tendency others call “homophily.”²¹ Sunstein shows the distressing power of homophily by pointing out that “[i]f you take the ten most highly rated television programs for whites, and then take the ten most highly rated programs for African-Americans, you will find little overlap between them. Indeed, seven of the ten programs most highly rated by African-Americans rank as the very least popular programs for whites.”²²

“Similar divisions can be found on the Internet,” he adds.²³ He lists sites that are explicitly designed for African Americans, for young women, for young men, and so on. He also cites research that he and a colleague did that found that of sixty randomly chosen political sites, only 15 percent put in links to sites of their opponents. “[M]any people are mostly hearing more and louder echoes of their own voices,”²⁴ because the Internet so increases the range of choices that citizens can find narrowly focused groups that precisely mirror their point of view.

The news is even worse than that, Sunstein fears. Studies have shown that when people speak only with those with whom they agree, they not only become more convinced of their own views, they tend to adopt more extreme versions of those views.²⁵ This group polarization happens for two reasons, Sunstein says. First, the members of the group have a smaller pool of views from which to drink. Second, because people “want to be perceived favorably by other group members,” they will often adjust their views toward the dominant position. “In countless studies, exactly this pattern is observed.”²⁶

The picture Sunstein paints is scary for those who care about democracy and disappointing to those who had hoped that the Internet would move us toward the traditional ideal of a knowledgeable person: open-minded, fact-oriented, and eager to explore other perspectives. Sunstein’s studies of group polarization specifically looked at offline interactions. So, we need to know: Is the Net in fact closing our minds and moving us toward more extreme views?

Sunstein is convinced: “Group polarization is unquestionably occurring on the Internet.”²⁷ His evidence is that “it seems plain that the Internet is serving, for many, as a breeding ground for extremism,”²⁸ pointing to “cybercascades” in which a belief rapidly gains many believers because it is being passed around the Net as true. Plus, “[a] number of studies have shown group polarization in Internet-like settings.” ²⁹ But, how big a problem is it on the Net? As Sunstein acknowledges in an edition of Republic.com published a year later, “To know whether this is a serious problem we need much more information.”³⁰ Does it happen a lot? A little? All the time? Compared to what? How often? How much? Does the Internet’s diversity of sources ever depolarize some groups? If so, why those and not others? Perhaps, as Clay Shirky has suggested, Sunstein has it “exactly backwards”: Perhaps “political discourse is coarsening” not because people are walling themselves into echo chambers but “precisely because people are constantly exposed to other points of view.”³¹

Indeed, some recent evidence suggests the polarization may not be as extreme as Sunstein thinks.³² Economists Matthew Gentzkow and Jesse Shapiro published a paper in 2010 that looked not at which sites link to which but what sites individual users actually visit as they spend time on the Net.³³ This study’s results seem to be the opposite of what Sunstein’s “group polarization” idea would lead us to expect: “Visitors of extreme conservative sites such as rushlim baugh.com and glennbeck.com are more likely than a typical online news reader to have visited nytimes.com. Visitors of extreme liberal sites such as thinkprogress.org and moveon.org are more likely than a typical online news reader to have visited foxnews.com.”³⁴ That is, those visiting the most obvious examples of partisan echo chambers are also more likely than most people to visit sites on the other side of the political divide.

So, is the Net reducing our shared experience, leading to group polarization, and thus hurting democracy by making us narrower knowers than ever? The Gentzkow-Shapiro study suggests not, but it’s just one study, and it is subject to dispute. For example, Ethan Zuckerman, my colleague at the Berkman Center, took a careful look at it and drew exactly the opposite conclusions.³⁵ He points out that the study finds that Net users are more insular than users of just about all the old media. Indeed, if we were simply to look around the Net, using our own experience as a guide—the opposite of a careful methodology, granted—many of us would, like Cass Sunstein, conclude that people do seem to be more polarized and more uncivil than ever. If you want to attract attention on the Internet, talking in extremes seems to be an effective tactic.

We are not yet close to having a solid answer to Sunstein’s question. Yet, it’s worth noting that it always seems to be “those other folks” who are being made stupid by the Net. Most of us feel, as we’re Googling around, that the Net is making us smarter—better informed (with more answers at our literal fingertips), better able to explore a topic, better able to find the points of view that explain and contextualize that which we don’t yet understand.

Not Nicholas Carr. He thinks the Net is making all of us stupider, including himself, but more or less for the opposite reason that Sunstein worries about. Carr notes at the beginning of his wonderfully titled book The Shallows that he realized in 2007 that his own cognitive processes were changing because of the Net, and not for the better. “I missed my old brain,” he writes.³⁶ For Carr, the cause is not the presence of echo chambers but their rough opposite: The linking, blinking, twittering diversity of the Net is making us dumb. The Web is reshaping our physical brains, Carr contends, “weakening our capacities for the kind of ‘deep processing’ that underpins ‘mindful knowledge acquisition, inductive analysis, critical thinking, imagination, and reflection.’”³⁷ He cites studies of the brain and of behavior to prove that the Internet is getting us not only to think differently but to think worse.

Carr’s picture accords with what many of us have sensed: These days we seem to be more easily distracted, we have less patience for long books, we want to jump over the boring parts to get to the “meat,” we have difficulty remembering how we got to where we are on the Web. At the same time, the studies Carr cites do not accord with the sense many of us have that we are now smarter than we were because the only limit on how quickly we can get answers is our typing speed, and because our curiosity has to travel only the length of a finger flick to be satisfied, and then be aroused again.

We all know that some of the places where we are smartest work only because they have properties of echo chambers: The clamor of disagreeing voices is muffled or silenced. Knowledge has always needed communities to flourish. Communities need walls so that they can let in the right amount of diversity, even if too frequently they err on the side of homogeneity. But now the Net has made community walls semipermeable. The transparency of the Net lets outsiders look in and insiders look out. And you may be exchanging ideas in a community that Cass Sunstein would call an echo chamber, but you got there by passing through the daily chaotic roil of ideas on the Net. Our old echo chambers were like quiet libraries in quiet communities. Our new echo chambers—knowledge communities—are on the busiest street in the world and there are no windows thick enough to cut out all the noise.

So, is the Net making us stupider or smarter? The Net is new, the research is relatively scant, and the Net is rapidly evolving. The answers may well—in fact, probably do—vary by the usual variables in such studies: economic level, education, gender, politics, interests, geography, culture, and so on. The concept of “echo chambers” is itself slippery. And then there is the difficulty of measuring any quality as culturally determined as “smartness.” As Carr writes, “The Net is making us smarter . . . only if we define intelligence by the Net’s own standards.” ³⁸ The answer to the question “Is the Net making us smarter or stupider?” is going to be settled not by thinking through the problem but by living through it.

Yet, there is a sense in which it does not matter. Whether or not the Web tends to make us more insular, we know that human beings have a tendency toward homophily; we prefer to be with people who are like us. All the participants in this debate agree that excessive homophily is a bad thing. All the participants agree that we should be bending our efforts to work against our homophilitic tendencies. And no participants—not Cass Sunstein, not Nicholas Carr—are suggesting that we roll the Net back up and throw it away as a bad idea. So, why so many years of debate and with such passion?

Al Gore published The Assault on Reason³⁹ in 2007 in the middle of George Bush’s second term,⁴⁰ so it’s understandable that he felt some despair. “Why do reason, logic, and truth seem to play a sharply diminished role in the way America now makes important decisions?” he asks on the first page. After many chapters convincingly making the case that governance has become unmoored from fact and argument, Gore talks about the Internet as a “source of great hope for the future vitality of democracy.”⁴¹

[T]he Internet is perhaps the greatest source of hope for reestablishing an open communications environment in which the conversation of democracy can flourish. . . . The ideas that individuals contribute are dealt with, in the main, according to the rules of a meritocracy of ideas. It is the most interactive medium in history and the one with the greatest potential for connecting individuals to one another and to a universe of knowledge.⁴²

This idea has been with us for a long time. Socrates thought so. The picture of reasonable people sitting together, talking over their differences in a respectful, honest way is the image on the Enlightenment’s own Hallmark Card. Jürgen Habermas, the influential German philosopher, marks the beginning of the “public sphere” with the rise of public spaces such as coffee shops where those conversations could happen. Al Gore is hopeful because he sees the Internet as an expansion of this reasonable public sphere. Sunstein is worried because he thinks we’re retreating into semi-private spheres. But we have all—almost all—thought that the way forward through our manifest differences has been to be open to contrary ideas, and to talk about them reasonably.

There’s a set of presuppositions behind this belief about the power of reason, the purpose of conversation, and the relation of knowledge and the world. Even if we leave those presuppositions alone, we don’t have to spend much time on the Net to come to the sad conclusion that simply as a matter of fact, we’re not ever going to learn to talk together reasonably and come to single conclusions. We are going to disagree about everything. That’s the fact of diversity with which we now have to deal.

What do we do about it? When it comes to climate change, Al Gore’s general strategy has been to say that those who disagree with him are not within the community of Reasonable People: The deniers are wrong on the facts, they don’t believe in science, and thus they have no place in the Coffee Shop of Reason. Evolutionary scientists often treat creationists the same way. And the other side uses exactly the same tactic: Al Gore is a close-minded hysteric with whom there is no arguing; evolutionists are Godless rationalists who won’t acknowledge the reality of the Divine so there’s not enough common ground to even begin a discussion. The only place we have the sort of rational discussions Gore, Sunstein, and Socrates value so highly is within an echo chamber—a room in which people agree thoroughly enough that they can disagree reasonably.

For example, in May 2010, Republicans in Congress launched AmericaSpeakingOut.com where people could post ideas that the Republican Party congressional leadership said they would read. The Republicans hailed this as a “revolutionary” democratization of the governing process. Then they watched in horror as, within the first couple of days, the ideas posted included the repealing of Section II of the 1964 Civil Rights Act because it is (as the suggester “explained”) “UNCONSTITUTIONAL, PROGRESSIVE and HITLER.”⁴³ Another suggested putting all American Muslims under surveillance. Another suggested a tax hike. The Republicans allowed the Muslim surveillance and tax hike suggestions to remain on the site, but asserted that they are not going to consider acting on those ideas. “The key is to remember that we are focusing on . . . settled principles,” said Representative Peter Roskam of Illinois. Those settled principles formed the outer edges of the permissible discussion. Stray beyond them and your ideas would be ignored or, in extreme cases, removed. This means America SpeakingOut.com is an echo chamber. But that’s a requirement if Republicans are going to make progress discussing the issues in the ways that define them as Republicans.

It’s important to be clear about this. We still need as much difference and diversity within the conversation as we can manage. We still need to continuously learn how to manage to include more diversity. We need to be on guard against the psychological tricks echo chambers play, convincing us that our beliefs are “obviously” true and nudging us toward more extreme versions of them. But it’s also fine for the Republicans to have an online “coffee shop” where they can discuss their ideas. That conversation needs far more agreement than diversity.

The Internet is showing us that our old ideal of a Coffee Shop of Reason exists only within a city with millions of other coffee shops that look wrong, wrong-headed, or totally unreasonable. But that fragmentation is exactly what the Age of Reason thought we could overcome. We now have pretty good evidence that we cannot. That evidence is the Net itself.

For a couple of thousand years, Western philosophers have debated whether human reason is sufficient for understanding our world. But the critics of reason generally have been lonely. In the past fifty years, these voices have become a chorus that in some quarters has become the loudest singers around. They got labeled “postmodernists,” which has stuck no matter how much they’ve objected.

When I was a graduate student in philosophy, the postmodernists hadn’t yet had much impact on North American soil. I wrote my dissertation on Martin Heidegger, a contemporary German philosopher who was considered by mainstream philosophers to be purposefully obscure and prone to making “everything you know is wrong” statements simply to appear deeper than everyone else. But I, and the academics in my circle, thought the obscurity of Heidegger’s writings was due to the profundity of their challenge to the fundamental assumptions of Western philosophy. Then, in 1978, postmodernism became an unexpectedly central topic at the annual Heidegger Circle meeting. Much of the talk, especially in the hallways, centered on writings by Jacques Derrida, who maintained that Heidegger’s radical philosophical project had not gone far enough. As a new member getting to hang out with the scholars whose work had guided me, I was struck by the fact that the sorts of things they were saying about Derrida were precisely the sorts of things non-Heideggerians said about Heidegger: He was out to shock, incoherent, purposefully vague, an intellectual charlatan. We were, in short, having a classic echo chamber moment, refusing to take seriously claims that challenged our own. The irony is, of course, that the rest of the world would consider Derrida and Heidegger to be overwhelmingly alike in their ideas.

Over the years, I have struggled with the works of this new generation of thinkers. Postmodernist writings tend to be remarkably dense, either—depending on who you talk with—because they are attempting to undo deep, basic assumptions embedded in language itself or because they are using a fog of language to hide the emptiness of their ideas. Obviously, no brief introduction can claim to be adequate, especially since there are so many differences among them. Fortunately, we need only a few crucial ideas from them to help us understand the world of difference the Net exposes to its every visitor.

Interpretations are social. Interpretation always occurs within a culture, a language, a history, and a human project we care about. The tree is lumber to the woodcutter, a place to climb to the child, and an object of worship to the Druid. This inevitably adds human elements of uncertainty and incompleteness.

There is no privileged position. There are always many ways to interpret anything, and none can claim to be the single best way out of its context. Some postmodernists talk about this in terms of denying that there are “privileged” positions, intentionally invoking not only the Einsteinian sense (all motion is relative) but also, pointedly, the socioeconomic sense (the elite should not get to marginalize the ideas of the rest).

Interpretations occur in discourses. You can’t make sense of something outside of a context. Even something as simple as a car’s turn signal can only be understood within a context that includes cars, the basics of physics, the unpredictable intentions of other drivers, the restrictions of law, and the way left and right travels with one’s body. Ludwig Wittgenstein talked about this in terms of “language games,” by which he meant not something you do for fun but, rather, the way our words and actions are guided by implicit rules and expectations. Postmodernists have many different words for these contexts, but we’ll use the term “discourses.”

Within a discourse, some interpretations are privileged. If you are within the discourse of science, fact-based evidence carries special weight, and emotions do not. On the other hand, if you propose marriage by compiling tables of data as if you were within a scientific discourse, you either are making a bad joke or are seriously disturbed. Discourses are themselves social constructions—they are ways people within a culture put ideas together. They are not themselves part of nature, and they change throughout history.

There is much, much more to postmodernism. But these five ideas are crucial and, in various forms, have been resisted and debated for decades now. That was, of course, before the Internet showed us that the postmodernists were right. Freed of the privileged position accorded them by the limitations of paper-based media, the old, centralized authorities are losing their purchase. On the Net we see just about every possible interpretation. When these interpretations contend across discourses—or across cultures, or socioeconomic groups, or any other group that has its own norms and values—there is little possibility of resolving the differences and sometimes even of understanding them.

And yet we cannot afford to let the best and the brightest stick to their own discourse, because when we do, hundreds of thousands of people can die in a needless war. Put people on a network and they might form the sort of echo chamber that Cass Sunstein worries about, making themselves more certain, more extreme, and more dangerous as they reconfirm their old opinions with white papers and backslaps. For those who have no interest in intellectual rigor, or who lack curiosity (which, by the way, characterizes each of us for at least part of every day), the Net may well be an environment that degrades knowledge. We need to be concerned about all this. And we need much more research to ascertain the actual risk and actual damage. But the network also offers the possibility of connecting across boundaries, forming expert networks that are smarter than their smartest participants. The network can make us smarter if we want to be smarter.

But we can’t leave it there. The question of difference drives us to hold paradoxical positions—the Net is an echo chamber, the Net distracts us with all its diversity—for the same reason that we have been so resistant to postmodernist ideas. We worry, understandably, that without a privileged position we will be lost in a swirl of contradictory ideas.

So, let’s take a look at how networked knowledge puts ideas together, especially as compared with the old, bookish ways by which we used to work our way toward the truth. Do the differences and distractions on the Net result in an inevitable frittering away of ideas? Is the hyperlinking of ideas an admission of failure, a new way forward, or a bit of both?

Scoping Diversity

1. Not all diversity is equal

2. Have just enough in common

3. Mix well by hand

4. Fork it over

Into the Echo Chamber

Unsettled Discourses