Information Cocoons

Thomas Nagel

  • by Cass Sunstein
    Princeton, 224 pp, £12.95, April 2001, ISBN 0 691 07025 3

One of the most remarkable effects of the Internet is that it permits unlimited specialisation of contacts and information. If you’re looking for an out-of-print book on an esoteric subject, you can find out instantly where there are copies of it in second-hand bookstores from Iceland to Australia, compare prices and conditions, and order it in a few seconds. You can read what people all over the world have to say on any topic that interests you, join in discussion with others who share that interest, and communicate your own ideas to the scattered community of specialists even in an arbitrarily narrow field – a community that could hardly exist without this possibility of electronic identification and expression. The availability of specialised networks applies to any type of interest whatever, and they are proliferating rapidly.

Cass Sunstein, a leading American Constitutional lawyer, is duly appreciative of this expansion of possibilities, but he wants to raise the alarm about another of its consequences, like those dire warnings of possible side-effects that come with every patent cold remedy. In this short, journalistic broadside he argues that the power the Internet gives each one of us to control deliberately what information and opinions we are exposed to – to tailor a communicative world to our prior interests and convictions – is a threat to democracy. The threat is that the choices made by individuals will add up collectively to a fragmentation of society so pervasive that the public sphere will cease to exist, and we will be left with multiple like-minded subgroups whose members talk only to their fellow members and never to those with whom they disagree, and who are exposed only to the kinds of opinions they want to hear.

If this happened, it would be a threat to democracy because on the deliberative, republican conception which Sunstein favours, democracy does not work merely by summing up the disparate private preferences of all the members of a society. Rather, it is supposed to force them to arrive at collective decisions by public debate, carried out partly through their elected representatives, and driven by the need to persuade one another or at least to arrive at mutually acceptable compromises. Sunstein fears that without the public space provided by ‘general interest intermediaries’ such as large-circulation newspapers, magazines and network television, there will no longer be any place where these debates can be had, in full view of the voters. There will be nothing that most people can’t help being exposed to – nothing that simply comes at them without their choosing it.

Sunstein frames the issue as an opposition between two conceptions of the value of freedom of expression: one based on consumer sovereignty, which is individualistic, and the other based on political sovereignty, which is public. Consumer sovereignty protects every individual’s right to say and hear what he wants, but gives no one the right to be heard. Political sovereignty, under the ideal of deliberative democracy, requires some rights of common access to the public forum, even if it means imposing on other people’s attention by haranguing them with speeches in the public square or handing out leaflets on the sidewalk. The first can clearly justify prohibitions against censorship, but the second is needed to justify the less familiar legal protection under the US Constitution of streets, parks and even shopping centres as venues for free expression. And Sunstein believes that even when we are thinking about the circumstances in which censorship is and is not justified, political sovereignty should rule: the protection of political speech as an essential condition of democracy should be regarded as the core of freedom of expression. Though it is tendentious to describe the individualistic value as ‘consumer’ sovereignty, there is an important issue of principle here, to which I shall return.

But what about the fear of fragmentation? Do the expanded possibilities for selection afforded by the Internet mean that the citizens of modern democracies will talk to each other less than they do now – will abandon even the limited political debate that now engages them? One of the immovable facts of life is that time is not expandable. If we allocate more of our attention to particular and specialised concerns, we will have less of it for what is common. I think Sunstein is being alarmist, however, about the effects of specialisation and individual control.

He himself points out that before the Internet, through other highly differentiated media such as niche magazines and talk radio, and through voluntary associations, people have always sought out the like-minded and have tended to protect themselves from unwelcome opinions. And he cites grave psychological studies demonstrating with statistical precision the obvious fact that discussion among people of opposed views tends to encourage movement of opinion towards the centre, whereas discussion limited to people on one side of an issue tends to reinforce their distance from the other side. This is called ‘group polarisation’, and it is evident in Internet discussion groups patronised by racists, gun fanatics, purveyors of bomb recipes and so forth. Certainly, the Internet has helped these people to find each other, just as it has helped collectors of antique toy soldiers and rare comic books.

Yet the growth of such possibilities doesn’t imply that specialisation will eat up all our attention. It makes specialisation easier and more precise, but there is no reason to believe it will displace the continuing desire to be aware of what most people are aware of, and to know something about the range of things most people think. Sunstein seems to believe that exposure to unsought information or divergent opinions is for most people like advertising: they can’t avoid it, as the price of getting what they are really after; they read newspapers or magazines or watch television news for their own narrow purposes, and are then shown other things that they would not have included if they had the choice. Most people are not such control freaks, however: they have a general curiosity and desire that not everything in their lives be planned.

More serious than Sunstein’s fear of fragmentation is the problem of how the common knowledge and salient range of opinions that can occupy an inevitably limited public space are to be selected. The more information that is available – the more cultural contributions and discoveries of every kind that are produced – the more competition there is for the limited room at the top where most people focus that portion of their attention which they reserve for the public domain. With the material multiplying ever faster, and everything also being preserved from the past and made instantly available (the history of film, for example), the problem of selection becomes overwhelming. What will our grandchildren have to learn to become part of a single civilised community? What institutions can we ourselves rely on to choose what will fall under our eyes without our searching it out?

Four hundred years ago, access was extremely difficult, but selection was not. An educated person could read everything that he could get his hands on. Today the flood of publications, films and television alone is paralysing, and the Internet is multiplying the available material by orders of magnitude. The power of intermediaries that command the allegiance of a significant public is enormous. In any modern democracy the selective decisions about coverage made by a handful of leading newspapers have a huge impact on what educated citizens learn (think only of the importance for a book of whether it is reviewed in their pages). Television news is in a similar position, and this fierce competition for salience can only get worse over time. We want to be informed of what most people are informed of, but that requires convergence on some source that acts as a co-ordinator, which, in turn, inevitably carries the risk, indeed the certainty, of arbitrariness.

So far, in spite of huge savings in production and distribution costs, electronic newspapers and magazines have not replaced print. People aren’t willing to pay for them, and they don’t yet attract enough advertising to be profitable. But even if the print media eventually go under, the demand for something that large numbers of other people see, simply because it creates the public consciousness, will not go away. Most people do not want to segregate themselves completely from unwelcome or unexpected facts and opinions; they will try to occupy a public space, and consumer demand will create one, or several, because people want to be where other people are, at least some of the time. This is due in part to the demands of the political system itself.

That is my guess. But Sunstein believes that the dominant effect of the Internet will be towards fragmentation and group polarisation, and that steps need to be taken to counter this tendency. Or perhaps he just wants us to worry about the possibility; his warnings are so often qualified that it is hard to tell. And although the book includes a general defence of the claim that individual expression may be regulated to serve deliberative democratic values, what he is willing to advocate is fairly innocuous, and of doubtful effect.

US law has struggled with the question of how far to extend rights of speakers’ access to the public forum. The attempt to force newspapers to carry diverse expressions of opinion has been firmly rejected by the Supreme Court. In the 1970s, responding to the scarcity of television channels at the time, the Court authorised Government to impose ‘must carry’ rules on broadcasters, requiring attention to public issues and opportunities for response from opponents of views that had been broadcast – a kind of fairness doctrine. This turned out to have a stultifying rather than a diversifying effect, since networks responded by avoiding anything controversial for fear of having to air the other side. In any case, scarcity is no longer a problem in the age of cable TV, and ‘must carry’ rules have been weakened, except for the requirement that cable providers must allot a certain number of channels for local and educational stations.

Still, Sunstein suggests that it would be good if Internet websites representing one political outlook automatically provided links to sites on the opposite side – gay rights groups to religious fundamentalists, and vice versa. ‘Ideally,’ he says, ‘such links would be provided voluntarily. It might also be worthwhile to consider legislation designed to ensure more in the way of links and hyperlinks, on a viewpoint-neutral basis.’ (Hyperlinks are those instant connections from one document or site to another that is mentioned in it.) ‘It might be worthwhile to consider’ is pretty weak, and I doubt that any actual legislation in this direction would pass Constitutional muster even by the regulation-friendly standards Sunstein derives from deliberative democracy.

He also suggests subsidies for websites devoted to democratic debate – sites called ‘’, for instance, or ‘’, where we could all find people to argue with about our favourite topics, and not just views we already agree with. No doubt there will be developments of this worthy and boring kind as part of the general experimentation inspired by the new order. But the impulse to exercise centralised control here, an impulse Sunstein seems to suppress with difficulty, is truly misplaced. I’m not talking about censorship, which is an issue internationally but not a serious question in the US – except for the design of filters that permit parents to determine what their children may see. I am talking about Sunstein’s broad distrust of ‘consumer sovereignty’ and his belief that here, as in certain other areas, the free run of individual choice will produce large collective harms, and not only through the effects on democracy of group polarisation.

That is certainly a danger we must always keep in mind – the danger of a destructive prisoners’ dilemma such as environmental pollution, population growth or the arms race, where the result of many individuals rationally pursuing their own interests ends up making all of them worse off. But Sunstein wants to convict even the general proliferation of consumer choices of this fault, on the ground that individuals have to scramble to keep up with the rapid obsolescence of computers, television sets, audio systems and so forth, simply in order to have what everyone else has – but they end up no happier than they were several years before with the lower-tech version, and they may even be rendered miserable by what is known as the ‘consumer treadmill’.

This is another case where deferentially cited psychological studies ‘show’ something utterly useless. Of course people aren’t made happier by larger TV screens, faster computers, smarter automobiles, easy jet travel or instant Internet access to vast quantities of information. What does happiness have to do with it? Provided they have enough to live on and are not horribly persecuted, people will fall into a standard range of happiness or unhappiness, depending on what they make of their lives in the context of the choices available to them. The value of technological innovations is not that they make us happier, but that they allow us to do and experience more, that they expand and enrich our possibilities. They also expand the opportunities for anxiety, failure, envy and the knowledge of what we are missing, but so what? It is better to be Socrates dissatisfied than a pig satisfied. Happiness is wonderful, if we can attain it, but it is not the only good. Living longer probably doesn’t make you happier either, but it’s better than dying young.

When it comes to freedom of expression, and freedom in the choice of what expression by others one wishes to be exposed to, the collective values of deliberative democracy and membership of a common world are of great importance, as Sunstein says. We should not be shut up in private informational cocoons. But most people want a mixture of public and private existence; the real problem with the electronic information explosion, a problem that antedates it, is how the demand for a public forum will be met, and who will make the more and more draconian choices of what gets into it. I believe it would be a mistake to give the state a significant role in this process.

The main theoretical objection to Sunstein’s approach is that it underrates the pure importance of individual freedom – which is vastly enhanced by the Internet. Freedom of expression and freedom to hear or read what one wants are aspects of the individual’s sovereignty over himself that is at the core of modern liberalism – something the collective will has no right of control over. This is not derived from republican values. It is part of the ideal of limited government, according to which we do not hand over plenary powers to the state as a condition for being able to enjoy the benefits of a peaceful, just and orderly collective life. The preservation of liberty in the use of those ever-proliferating goods and opportunities that the state makes possible, through the legal regime that underpins technological progress and protects its products, is just as important as the encouragement of a public sphere for democratic deliberation. Sunstein is, I hope, too pessimistic in thinking the first of these values poses a serious threat to the second.