“Computers and democracy” is a big, vague, hot-air balloon of a topic. I'll try to bring it down to earth; I'm not sure whether I'll succeed. Americans' attitudes toward democracy and computers are curiously similar: we depend on both, believe in both, pride ourselves on both, but we aren't particularly interested in the details of either. The first role of any well-functioning democratic government or desktop computer is to annoy us as little as possible. Traditionally, in this country we have not admired people who are too eager to be part of government; we are not too crazy about computer experts either.

Computers are likely to affect American democracy mainly by influencing American culture—the atmosphere in which democracy lives. Networked computers at their best are like the feeling of spring in the air; among human beings, the feeling doesn't make things happen, but it makes people more eager to make them happen. It puts us in the mood for action. Computers can do that too. They can put us in the mood for change. They can give us an excuse to get good things done.

But good things don't happen automatically. Many of us approach technology deferentially. We are content to sit back, see what happens, let matters take their course. Many are content for software and computers to be left to (p.259) technical experts. Some of us, for that matter, feel the same way about government: that mere ordinary people aren't equipped to make public policy.

But computers and the Internet have a shaping effect on American culture, American language, and the texture of daily life. Their influence will increase. Their gravitational field will continue to grow.

To design software and help guide technology evolution takes wisdom, experience, and common sense—qualities that are no easier to find among technical experts than anywhere else. Americans ought to take technology in hand, make computers their business the same way they make government their business, not because they are ideally qualified to call the shots but because they are better qualified than anyone else. Computers could be a great thing for American culture, and therefore for American democracy and life in general. They could also bring out the worst in us, leave us richer in money and worse off in spirit than we have ever been before. The outcome will depend on whether we sit back and let the technical experts run the show or assert our right and our duty to shape the evolution of this astonishingly, unsettlingly important technology.

First let's consider the mechanics of democracy: what are the probable consequences of using computers as voting machines, and how would we go about doing so?

For some time people have been considering using computers as voting machines. Proposals come in many formats, but let's focus on a simple, radical concrete plan. Let's say that you could vote by using any Net-connected computer during official voting hours, at home or in any other place. Let's forget about the details: we would have to make sure that only qualified people vote, that no one votes more than once, that voters see only the ballot for their town. We would have to make sure that accumulating vote totals weren't messed with or snooped on. We would have to make sure that Net-connected computers were no less accessible than today's voting machines, ballot boxes, and so forth. We would have to decide what kind of software to use and how the interface should work. (Presumably each state would make its own decisions.)

We would also have to solve other problems that computer voting wouldn't create but would bring to the fore. For example, if we were allowed to vote using computers in our homes, many of us would spend a good part of every election day within a few steps of a voting machine. Inevitably, we would start wondering whether we could change our votes and then change them back. We shall need a policy for computer voting, as well as a way to carry it out.

But let's assume that we have solved all these problems. Some are hard, but we could probably manage them all within a year or two if we worked at it. The real question is: Is computer voting a good idea?

(p.260) At first it sounds rotten; at least it did to me, for several reasons. The most important thing about the process of an election is not that the final count should be completely accurate (with every last vote recorded and totaled correctly) but that it should be, as far as possible, transparent. Tampering with the results should be difficult and should seem difficult. We ought to be able to picture clearly in our minds the procedure that herds millions of scattered, scampering votes into a majestic grand total. If we can picture it, we can believe in it. And today, on the whole, we can picture it; our elections score fairly high in transparency. A paper card or ballot is a concrete thing. So is a voting machine. Personally, I've always liked the solid little crunch that a voting-machine lever makes when you swing it down, and the ringing thrunk when you open the curtain and register your votes. To me these are the sounds of democracy. I suspect other people feel the same way. It's easy to picture the process that connects our tangible voting acts to the final total. Maybe our picture is incomplete, even wrong; and of course computers are already heavily implicated today in vote counting. But at least the voting act itself is easy to understand, and the mental picture exists and is comforting.

Computer voting might be far more accurate but it would be less transparent, and I'm not sure that's a good trade. Every programmer knows not to fix what isn't broken. Every programmer also knows that software is hard to get right and prone to break—and everyone else knows it too. I think we could build voting software that is more reliable overall than the system we use today, even if it were prone to error and fraud. And I think we could sell it to the American public. But our elections would become less transparent; a vote would amount to a handful of bits in a complex series of software structures, and few people can picture software at work. Inevitably, we would pay a price in public confidence: we would have less in the bank when elections were disputed, as they are bound to be occasionally, no matter how we do our voting. The real story of the 2000 Bush-Gore presidential election was the crisis that did not happen. If we change our way of voting, our first responsibility will be to do no worse than we are doing today.

At least one other problem with computer voting comes to mind. Voting shouldn't be difficult, but it shouldn't be too easy either. Making people leave their homes, go to a public place, and maybe even wait on line is hardly an unbearable imposition. But it does quietly underline the importance and dignity of the election. It seems right and fitting that voting be done in a public place—usually with flags on hand, with your neighbors standing around, with the feel of a special occasion. Voting is a little (just a little) like getting married. Other things being equal, the nation has an interest in its citizens marrying and voting. Once people have made up their minds to go through with either (p.261) process, we don't want to talk them out of it. But we don't want to make it too easy either; we want to make it clear that these are not casual acts—deliberation is called for. In theory you could get married in front of a PC, too; you could go to the wedding Web site of your choice and download appropriate music and the necessary forms. But this is a bad idea.

Nonetheless, computer weddings almost certainly will happen, and so will computer voting. When we do make the switch, we ought to stipulate something like this: the company or agency that builds the software must be able to explain how it works to any voter in five minutes, not in comprehensive technical detail but in a fundamentally accurate way. It's a difficult requirement, and admittedly subjective. But it would put pressure in the right direction.

And despite its disadvantages and even dangers, computer voting might even be good for American democracy. It might turn the public's lack of interest in politics into a force to improve politics. By making it easy to get the minimum information you need to vote at the moment you need it and not a moment sooner, computer voting could help make political campaigns as we know them disappear.

Why? Because, first, the social dynamics of computer-based elections would probably allow (and even encourage) voters to spend a few minutes casting their ballots. To take their time. Even when a person is voting via computer in a public place, computer habits are different from voting-machine or ballotbox habits. People expect to sit down and think when they use a computer.

So we might allow each candidate to put a short statement on file, maybe five hundred words; you could read as many of these statements as you felt like before casting your vote. You'd click the “show me the statement” button, or something like that, next to the name you were considering. Our election laws might have to be changed to allow this system, but they might be worth changing.

What's potentially useful here is not the mere existence of these statements—there are plenty such statements floating around today—but the way they would become available at the right moment to be worth reading, in a setting where reading them seemed like a reasonable thing to do. Voters who dump campaign mailings in the trash might nonetheless read these. Voters who ignore the leaflets thrust in their faces by campaign workers hovering hopefully outside the polling-place perimeters might read these statements. There's a big difference between information you ask for and information that is flung at you the way popcorn is thrown at elephants by zoo visitors. When you ask for information your attitude is different, and the information itself can be presented differently. Most Americans aren't sufficiently interested in politics to ask for information—but we are a practical people, so we say, and (p.262) voters who would never ordinarily solicit this kind of information might conceivably ask for it if they were face to face with the ballot and the asking were easy.

This last-minute, point-of-sale message would inevitably carry weight and—more important—give voters an excuse to do what most of them would like to do anyway: ignore the campaign entirely. Block it out, with the thought in mind that candidates' statements will be available when they are needed. The whole plan conflicts with the idea some people have that voters solemnly ponder their election-day choices for months beforehand, but I doubt whether this view of American voters is realistic. In any case, we could gradually increase the amount of information available at voting time—starting with the five hundred–word statement, which would no doubt turn into a video snatch, we could add a two thousand–worder for voters who ask for more, then a five thousand–word statement, and so on. Meanwhile, the rest of the campaign might become increasingly marginal.

The difference between information you ask for and information that is thrown at you affects all of American politics. We turn our politicians into court jesters and professional party crashers, begging for small handouts, mugging and cavorting for attention. And then, naturally, we despise them. Putting a bit of information in the right place, at the right time, in the right way, won't in itself cause our existing campaign system to collapse. But it might give it an excuse to collapse.

In fact, computer voting is probably our best shot at giving it an excuse to collapse. I don't think we will ever have publicly financed campaigns in this country, to a greater extent than we have them (at the presidential level) today, and I don't think we should. People don't enjoy paying for things they're used to getting free and didn't want in the first place. Given the state of American election campaigns, forcing the public to pay for them would be like forcing us to pay for junk mail or obscene phone calls. What candidates desperately need from us is not our money but our attention. They want money only to buy attention. Donating a little attention free of charge would be the perfect fix—and if we could do it strictly on our own terms, we probably would. Neither the campaigners nor the campaignees like today's system, and it will fall apart if we give it a chance.

Here is a general point about the probable evolution of technology. The value of a piece of information depends on when you get it. Information that is worthless at the wrong moment can be invaluable at the right one. Using computers to deliver information at the right time in the right way is likely to be a big topic of technology and culture research over the next decade. For example, we've been working for several years on a research software system, (p.263) now also a commercial project, called “lifestreams,” in which users can maintain automatically a comprehensive, time-ordered journal of their electronic lives. Every document, image, email, Web bookmark, news update, or other piece of information they receive or create is dropped at the end of a constantly growing, time-ordered stream. At the head of the stream users see the latest stuff; as new documents arrive, old ones move back. So the stream flows, the way time flows and life flows. The stream has a future too. Users store their plans and appointments in the future. These flow toward the present. When they reach “now,” they hop over the “now line” and shuffle off into the past.

But we can also use the stream's future as a staging ground for information that we would like to deliver exactly when it's needed. If you're traveling (for example) and will need information about your plane ticket around three in the afternoon, about ground transportation two hours later, and about your hotel, various phone numbers, and background information for a meeting the next morning, you could lay out all these documents in sequence in the stream's future (like clothes a mythical valet might lay out on a theoretical bed); each piece of information would cruise forward toward “now,” and present itself when its big moment arrived.

So changing campaign politics could be one facet of a larger problem of technology and culture. The “information age” is all right as far as it goes (it's been under way for at least a century if it exists at all); but what we really need is an “information at the right time” age.

What does democracy need? Citizens who are well informed and thoughtful, who feel responsible to the community. Computers and democracy intersect as soon as we talk about information: computers and networks supply us with some of our information and, more important, they handle nearly all of it. Almost all the words that are published and nearly all the prepared words that are spoken pass through computer word processors, over networks, and out of printers.

George Orwell wrote in his famous essay on politics and the English language: “If thought corrupts language, language can also corrupt thought.” It's no accident that the twentieth century's most notorious tyrannies expressed themselves in opaque, muddy, swollen language. Crisp, clear, clean information—sparkling, if possible—is a necessity of democratic life. A democratic society had better be an articulate society.

Are computers good or bad for prose? In a sense, the word processor and the cheap printer are the best things that ever happened to writing. To write well is to revise obsessively. Word processors are perfect tools for trying out the look, sound, and feel of a sentence before you commit yourself. (Writers don't like to commit themselves to anything as important as a sentence until (p.264) they are sure of it.) Computer printers let you put your draft on paper, change it, print it again, change it again, and on and on, with the neurotic persistence of a child working a yo-yo, until your editor goes crazy.

Email brought the personal letter back from near death, admittedly in a strange new shape. During the 1970s and 1980s, phone calls grew insidiously less expensive, and then phones themselves became cheaper, smaller, wireless—it was a nightmare out of Hitchcock. It seemed clear that before long we would be crunching cellphones underfoot like cockroaches. Meanwhile, the U.S. mails were nothing to write home about. So by around 1990 the personal letter was gasping for breath. The rise of email has put it back in play.

And yet computers can also be bad for writing. Email resurrected the personal letter. It might also be fostering a culture in which our prose is so feeble we have to load it up with little smiley and frowney faces because otherwise no one can tell whether we're kidding. My sons are often told to write school papers based on research using the Web. The great thing about this form of research is that, for the first time in history, you can write a research paper in less time than it takes to read one. Furthermore there is no quality control on the Web; there are high-quality sites alongside junk sites, where junk information is expressed in junk prose.

Once again the technology is neither good nor bad in itself; what matters is what we make of it. But if I had to guess, I'd predict that the new technology of writing could yield something remarkable for American culture by giving us an opening, a good excuse, to do what we want to do anyway.

Today we have by far the best writing tools that ever existed. We are only a generation or two removed from the greatest outpouring of brilliant writing in American history, an extraordinary eruption that lit up the skies from the 1920s into the 1960s or early 1970s. There are always a few brilliant writers around; you can find them today, guaranteed. But the interesting thing about the great age of twentieth-century American prose is how the formidable parts added up to a remarkable whole, how dozens of sharply defined and distinctive individual voices added up to a national style that caught up the whole country and carried it forward—these authors wrote poems and art stories that were also best-sellers and popular magazine pieces.

Such writers of vivid, crew-cut English as Sherwood Anderson, Gertrude Stein, and Ernest Hemingway led naturally to the great essays and journalism of E. B. White and Joseph Mitchell, A. J. Liebling and Ernie Pyle, and on to a new crop of novelists and essayists who continued in their own voices the same vibrant, funny, often belligerent national narrative: Norman Mailer, Irwin Shaw, Saul Bellow, John Updike, Philip Roth, Eudora Welty, Tom Wolfe, Cynthia Ozick, James Salter. I could go on. This remarkable outpouring ran (p.265) out of steam in the 1960s or maybe the early 1970s. But that wasn't so long ago—the engine is still warm, and it could easily restart if something were to kick it back to life. Maybe our unprecedented new writing tools will do that, together with email, which makes it easy for writers and editors to work together.

Computers wouldn't cause this big cultural event; all the great writing and editing technology in the world never produced a single decent sentence. But it might give us an opening or occasion or excuse to get back on track and do great things.

Americans have many things on their minds; we are not a monomaniac nation. But it seems to me likely that more or less continuously since the 1930s and occasionally during earlier periods one “big theme” or another has preoccupied us. The big theme is the spare topic in every lagging national conversation. It's the pattern on the national wallpaper. During the 1930s the Depression was the big theme. Then World War II and its immediate aftermath took over, and then for the forty-odd years between the Berlin airlift and the collapse of the Soviet Union it was the Cold War. Today, our big theme has to do with computers and the Internet.

Identifying a single big national theme is a thought experiment, an exercise in deliberate oversimplification. But such exercises serve a purpose. If you are trying to draw a human figure, for example, you will have many complex details to deal with. But you'll also have the problem of finding the one characteristic line of action or tension or flow that makes the drawing come to life. That one line is not the figure or anything close to the figure; it's a radical abstraction. But it's a necessary abstraction. It does not suppress the details; it makes them emerge more clearly.

Granted it seems crazy to propose an analogy between the Cold War and the desktop computer as national big themes—they come from far-apart categories of existence. And I'll admit that such a claim can only be arbitrary and subjective. But I can't shake the strange feeling, and it is strange, that the role the Cold War played in my own childhood is being played in my children's lives by the Internet and computers. When I was a boy President Kennedy, the intrepid Cold Warrior extraordinary, was the official hero of all boys—not because of the Cold War per se but because of PT-109, a story of wartime heroism that every child knew. Of course we admired other people too: war heroes, sports heroes, Hollywood heroes, television and pop-music stars, and all presidents of the United States ex officio. Probably we admired presidents most. In around 2000 we were told that Bill Gates was the man children admired most. It seems impossible—but it fits.

Any analogy between the Cold War and computers can only be a halfanalogy, (p.266) similar to a half-rhyme or an unresolved dissonance. The halfanalogy is inherently awkward, and it's not an officially sanctioned rhetorical device, but nonetheless it can be useful.

So consider this half-analogy between the Cold War then and computers now. The most important function of a big theme is to answer questions automatically. The big theme is an answering machine with the prerecorded message built in. If it helps our position in the Cold War, it's good. If it has to do with computers or the Internet, it's good. This is a valuable service at times. It keeps us from going over the same ground repeatedly. It allows us to put a national consensus to work and get things done.

But it can also be dangerous. The Vietnam War, for example, was a watershed catastrophe in American life. We fought in Vietnam for many reasons, but ultimately the Cold War made us do it—our global struggle with communism and the Soviet Union. Whether you believe that it was noble or evil for us to have gone into Vietnam, it's hard to doubt that we got involved in the war without thinking carefully about what we were doing. During its early years, as we edged out calmly into those huge rolling breakers step by step—breakers that eventually swept away so much that was good and valuable in American life, above all our endless confidence in ourselves and our future—there was relatively little discussion or argument about Vietnam. Not much discussion of our goals; not much discussion about whether our methods were in line with our goals. The arguments came later. In a sense, they came too late.

Ideology is a sedative. It makes you accept propositions without thinking, gives you the answers without making you solve the problems. It is dangerous to operate a nation when you're under the influence. Today computers are the big theme, and they are a new sort of ideology: an ideology without ideas. Computers are an ideology the way Bill Gates is a hero. American culture is in sad shape.

Today's prerecorded message is: If it has to do with computers or the Internet, it's good. Karl Marx made the famous pronouncement that history repeats itself, the first time as tragedy, the second as farce. (This might be his most important accomplishment.) Probably there is no national tragedy lurking in our fixation on computers. So far there has only been farce: foolish, expensive mistakes.

Take the collapse of the “dot-coms.” People poured money into companies that had no logical reason for existing on the theory that commercial reality could be suspended like alternate-side-of-the-street parking rules as long as a business plan centered on the Web. The big theme made people turn off their brains.

Or consider the Clinton administration's aggressive promotion of the Internet (p.267) in the schools. Most Republicans supported the policy in principle even if they opposed the tax rise that accompanied it. Our schools were evidently failing to teach reading, writing, history, and arithmetic, but few people claimed that our students suffered from information underload, or excessively long attention spans, or too few in-school distractions. I don't claim there was nothing to be said in favor of putting the Internet in schools, but there were definitely things to be said against it, on the grounds that at best it was a distraction and at worst an attractive nuisance. But almost no one said them, which should remind us that we have two political parties in this country in order for them to disagree, not congratulate each other on bipartisanship. Especially on domestic policy, political parties are supposed to disagree. If they don't, there's probably something wrong with them.

Computers didn't ask to be appointed our national big theme. The situation developed by accident and is nobody's fault. Nonetheless: as big themes go, this is a bad one. It makes us painfully credulous. We need a loyal opposition, and we don't have one. We are setting ourselves up for bigger business failures, and possibly in the long run for an anti-technology backlash. More important, “computers and the Internet” makes an unworthy big theme for a great nation, because it is morally, spiritually, and intellectually vacuous.

Computation is not at all vacuous in itself. Developing the science and art of computation was one of the twentieth century's greatest achievements. But we choose to treat the subject as if it were vacuous. More than a generation after most universities established computer science departments, only a handful of people in and out of universities have any idea what computer science is. The public generally doesn't care, and computer scientists generally don't care whether the public cares or not.

A nation does not consciously choose its themes, and it would be foolish to suggest what the next theme should be. But we can make predictions if not suggestions. Big national themes are likely to grow out of earlier themes. In a certain limited sense, modern computer technology did indeed grow out of the Cold War. American military research was the driving force behind the original Arpanet of the late 1960s, which became the Internet in the 1980s; military research funding was a factor in the emergence of nearly every aspect of modern computing. Of course, there were many other contributing factors, too—freedom of thought, free teaching and research, free markets, and the nation's tendency to attract many of the world's smartest people to its corporations and universities. Ultimately these are all facets of American democracy.

No one can predict future themes with any confidence, but if we had to guess, we should probably expect the next to grow out of the current one—unless some new, unforeseen national emergency arises (for instance, a resurgent (p.268) Russian empire). But we know that the next theme will be something concrete. People have proposed community, civility, and spirituality as important topics for the national agenda, but these wouldn't even make good titles for directed studies courses.

One possible new theme might have to do with the emergence of new Internet-based cultural institutions. During my lifetime, thousands of major new businesses and corporations have been created in this country, but not a single important new university. No important newspapers have been born, although many have died; some first-rate magazines have been created, but only one or two have real cultural weight. Lots of publishing houses have bit the dust. I could go on: it's been an extraordinarily creative era for business institutions and a dead period for cultural institutions.

Creating new cultural institutions is important because successful existing institutions have an obligation to continue succeeding. They are conservative by nature. Yale's most important obligation is to go on being Yale. If technology means that a university could become a radically different kind of institution, Yale is the wrong place for it. Radical experiments make sense at institutions that have little to lose. That's why new institutions are indispensable to American culture; they are the right settings for the risky experiments that can change society.

Of course, technology will change existing successful institutions too—it will change Yale, for example, although the changes had better be in character or they won't have a chance and won't deserve one. Universities like Yale are working hard to deliver courses online. This is an interesting, important project. We might also consider the related but different project of putting the university itself online, building an electronic campus.

The electronic campus would be a meeting ground, like any campus, but it would be laid out in time instead of space. Think of it in terms of the lifestream structure I mentioned before. In this version, the electronic campus would be a stream—the ongoing, emerging narrative history of the university. It would be made up of lots of separate histories shuffled together, like decks of cards shuffled into one stack. The art gallery would have a history: a series of announcements, lectures, acquisitions, new hires, and so forth. Each event in this history would be a separate document posted on a growing, time-ordered stream. The separate documents might be skimpy or detailed; chances are they would start skimpy and become detailed. Each course in the college would have a history—a series of assignments, handouts, exams, and so on. The football team would have a history; each college would have a history; the physics department would have a history; the dining halls would have a history (the story of what they served yesterday, last week, last year, last century); (p.269) all these documentary stories shuffled together would be the backbone of the electronic campus.

Each member of the community would have a personal history too: your email, notes, and photographs, the drafts of your papers, your address book, your diploma. When you arrived as a freshman you would be set up with your own personal view of the university stream. Your view would have your private documents shuffled in. No one but you would be able to see them; you would see your own documents and as many public documents and conversations as you cared to. Each person would see a different stream, but all these streams would overlap. To put it another way, everyone would see a private foreground—his own story—set against a public backdrop, the university's story. The stream would have a future where the university's plans and your personal plans would be stored. Each document in the future would roll steadily forward toward the present. And the stream would have a past, which in this case goes back to 1701. All the documents, drawings, photos, films, and so forth that document Yale's history would be filed in the stream in proper time order. The future would flow into the present, which would flow into the past. It would all be one stream, constantly moving and growing. You could pick a topic or category and follow it forward or back as far as you wished; or you could sit back and watch the stream move. When you graduated, you would keep your stream account. You could tune into the university whenever you liked. Your history and the history of your class would continue to accumulate on the stream.

There's more to be said, but that's one version of an electronic campus in brief. How does the electronic campus compare to the real one? Its disadvantages are obvious. On the real campus people can talk face to face, play football and Beethoven and look one another in the eye. But the electronic campus has certain advantages too. You can be part of it no matter where you are. Everything takes place in historical context: you can rewind the electronic campus to last term or last decade or last century. Everyone can watch the passing scene from a front-row seat.

Streams like this have been established at a few small institutions, for example at the New Haven technology company that built the commercial version of the software, Mirror Worlds Technologies. One of the stream's big advantages at Mirror Worlds is that when you are tuned in to the stream, you feel that you're right at the center of the action even if you are nowhere near the company's offices. One of the stream's big disadvantages is exactly the same: you feel you're right at the center of the action, but you aren't. You've lost face-to-faceness, and that's important. You gain something in return; it can be a reasonable trade so long as you don't kid yourself about the cost.

(p.270) Before long Yale will probably be in the strong position of having both kinds of campus, the physical one and some kind of electronic one. Similar techniques make it easy to imagine that we could create new universities whose faculty and students are spread throughout the world. These new institutions would exist mainly online. Their members would meet occasionally, but they would be unlikely to have an actual campus. Do the potential gains make up for the big losses? I don't know. But I'm sure we'll have a chance to find out. This experiment will be tried. We're seeing some tentative first steps already.

Everyone knows that it's almost impossible to start a newspaper nowadays. Of course it might be easy to start an electronic one, but on the whole, Web news services have been disappointing. They're good at providing focused, specialized news but not at creating a document that people turn to for the pleasure of browsing and reading. We know one thing about electronic newspapers: when a successful one is created, it won't look anything like an online version of the Times or the Wall Street Journal. The organization and design of today's newspapers reflect the character of printing on paper. A successful online newspaper will come in a new shape and use new organizing principles that don't exist on paper. It might be a stream, for example, where you could look at the future to find out what's coming up, and the present to find out what's happening now, and as far back into the past as you'd like. Or it might take a completely different shape. In any case, its shape will be new, not a mere translation of paper into a new medium.

Suppose electronic newspapers were structured as streams. You could take the equivalent of the New York Times stream and the Wall Street Journal stream, mingle them together, and read both newspapers simultaneously. You could follow a story backward step by step to the beginning. You could make everything disappear but the graphs and pictures. You could read a newspaper once a day, once a week, or once a month and it wouldn't make any difference; whenever you tuned in, you could take up exactly where you left off. The stream is an easy structure to navigate—you can go forward or back, look for this or that—and so you could listen to a newspaper while you drove to work; your car computer could read out the newspaper stream. Whether these possibilities are practical remains to be seen. The point is that we can experiment. Newspapers are important to democracy. Nowadays they aren't doing terribly well. Computers will give them a chance to change.

Some of our new institutions won't have any precise analogues in today's cultural universe. Take the following obvious stream-based service, or institution, or whatever you want to call it: New York City is the center of the publishing industry, but it's radically underserved in book reviews. There are only two weekly book reviews for the general public; hundreds of new books (p.271) are barely mentioned or ignored altogether. In fact, all sorts of New York cultural events receive cursory or no notice. I know an artist who worked for years on a big New York gallery show. When it finally opened not long ago, everyone waited anxiously for the reviews. When a review appeared at last, she and her show were blown away by a single paragraph in the New York Times. The art world shouldn't work like that. But in this allegedly information-rich world, the arts are starved for attention.

Before long, someone will found a New York culture stream: it will publish news and reviews but will also be a ticket and shopping and calendar and archive service. When you tuned in this culture stream, you would see latebreaking news and reviews. You could watch everything or focus on books, music, art, Lower Manhattan, whatever you chose. If you read a review for a show you wanted to see, a seating plan for the theater would be on the stream, so you could buy your tickets by clicking the right boxes. The tickets would be delivered to your stream, and a note would be dropped into the stream's future reminding you to go see the play when the date rolled around.

We're already seeing tentative movement in this direction; we'll see a lot more before long. These culture streams and related electronic institutions have no analogues in the pre-Web world, but they could easily become more influential than any newspaper, television, or radio outlet that exists today.

Computer technology in itself accomplishes nothing. If we're going to have an American cultural renaissance, computers won't supply the ideas or the will, let alone the people or the money. Computers won't change our election campaigns or improve the quality of our information supply or create new institutions. But they might give us openings, if we choose to go through them.

Information is no good in itself. It becomes valuable when it shows up at the right place, in the right way, at the right time. The technology world hasn't gotten around to taking this problem seriously yet, but it will. We can make transparency a goal of software design if we want to. We can require that every significant piece of software be able to explain not only how it works but how it's built. Nowadays few people care. As software becomes increasingly central to the workings of society, more people will care; at least I hope they will.

The quality of our language is central to the quality of our democracy. By making it easier to generate written words, computers could be the best or the worst thing that ever happened to our capacity to express ourselves. The outcome depends on whether we believe that good writing is important. Today, on the whole, we don't. But computers could give us an excuse to change our minds.

We each have a fixed number of hours to dispose of, and we spend some of them dealing with information. Other things being equal, the more bytes of information we look at, the less time we invest in each. It's good to know a lot, (p.272) have seen a lot, been exposed to a lot. It's also good to study important things in depth. I think we need to recalibrate the balance between depth and breadth in American education and perhaps in society at large. Thanks in part to the Web and to other electronic information sources, our balance is temporarily out of whack. I sometimes think about early American households, where books were rare; you might find a Bible, perhaps Christian or patriotic pamphlets, often the works of Shakespeare. It wasn't an atmosphere that encouraged studiousness, but when a child was studious—when he was Abraham Lincoln, for example—he put all his energy into a relative handful of bytes. But he studied them well and thought about them deeply. His thinking was deep instead of broad. Would his education have turned out better if he'd had the Web to surf? I doubt it. Deep reading, close reading, careful and meticulous reading, is the basis of education. We can't let the Web or anything else make us forget how to do it.

Finally, American society in recent generations has let one big basic theme dominate our national discussion. Today we are focusing on computers and the Net. We can do better. It's time for a new theme. Over the years we've come to accept responsibility for the air we breathe, the water we drink, the schools we operate, the culture at large. We could accept responsibility also for the computers we live with. We could reject bad software, send it back to the factory; we could think carefully about how software ought to behave and what it should do, and about where the field as a whole ought to go. If we do that, if we choose technological democracy, then we can realize the enormous possibilities of computers and the Net.

If we don't, if we sit back and leave computers to technologists and the industry, then society as a whole could easily come to resemble a piece of commercial software. It could grow steadily more complicated, with more and more fancy features that fewer and fewer people understood; with an underlying structure that was opaque and mysterious. The gap between ordinary citizens and expert users would grow larger and more ominous. Gradually the spirit of civil society might come to resemble the spirit of the Microsoft paperclip—an onscreen cartoon character (for those of you who never met it) that Microsoft uses to express its cheerful, charming, easygoing, good-natured contempt for us poor, childlike fools who use its software.

But things don't have to be that way. Computers are one of the most powerful tools we have ever invented. The choices on this ballot proposition are clear. What to do with this amazing tool? (A) We use it. (B) We sit back and let it use us.

I vote for (A).