Ten Short Arguments About Cyberbalkanization
I’ve got some interest in questions about civility in political discourse
, especially on the internet, so I read with great interest a recent post by Jack Balkin of Balkinization
concerning the alleged problem of “cyberbalkanization.” Below I quote some relevant passages from Balkin’s essay and comment on them. Balkin’s essay is sensible and I find myself inclined to agree with much, but not all, of what he writes. But without further ado:
[Passages tagged with numbers in square brackets are quotes from Balkin’s essay; my comments following each passage are numbered like so: [n/c] where ‘n’ is the number of the passage I’ve taken from Balkin.]
 “The development of the blogosphere mitigates, to a considerable degree, two key concerns about freedom of speech on the Internet. University of Chicago legal scholar Cass Sunstein made both of these points eloquently in his book Republic.com. The first concern was that the public sphere would become fragmented because there were so many speakers, no common sources that everyone was exposed to, and new filtering technologies allowed people to filter out the speech they did not like and only read the topics and opinions that interested them. The second concern was that people would become increasingly extreme in their views because there is no Internet equivalent to the fairness doctrine. Liberals would listen only to liberals, conservatives would listen only to conservatives, and the resulting ideological division would produce ideological polarization with increasingly extreme positions, further fracturing the public sphere and preventing democratic deliberation. For this reason, Sunstein at one point suggested requiring people with websites to include links to people with contrary views, or, if that posed constitutional difficulties (it would) at the least giving tax or other incentives for people to add links to others. Sunstein imagined a sort of Fairness Doctrine in Cyberspace. When it was pointed out that Cass didn't have any such links on his own site, he promptly placed a link to Richard Epstein and Catharine Mackinnon on his home page.
In hindsight, both of Sunstein's concerns about freedom of speech seem overstated and his proposed remedy seems not only ineffectual but beside the point because it misunderstood how the Internet differs from traditional mass media. The development of the blogosphere helps us see why this is so.”
[1/c] Two points:
First, I had never heard of Professor Sunstein or his book when I wrote posts on this subject. Which is not to say that the ideas were mine. They were all floating around in one form or another, and although I didn’t make that clear, that was because I took it for granted and thought that the ideas were fairly common and “in the air.” I certainly hope I didn’t suggest otherwise.
Second, let me just say that the idea of forcing people to include links on their sites strikes me as appalling. We’d have to be on the verge of cyber (or actual) civil war before I’d even consider anything like that.
 “Sunstein assumed that speakers on the Internet would in some respects be like radio and television broadcasters who could simply deny access to viewpoints they did not agree with. That is why he wanted to transpose the Fairness Doctrine into cyberspace. That is why he put links to Epstein and MacKinnon on his own website. He was working with the paradigm of broadcast television, a unidirectional non-interactive and non-participatory mass medium in which it is relatively easy to exclude speakers.”
[2/c] Good point, though it may be worth noting that it’s still easy to exclude speakers. One way is to edit your comments. Sure, it makes you a weenie, but some blogs do it. How many? I don’t know for sure (but there are quite a few weenies in the world…) I’d actually guess that few blogs edit comments. But—and here’s the first instance of the main point of this response—it’s an empirical question. Another way to balkanize is to form a community of commentors that is openly hostile to those who disagree with them. Some of us are not easily intimidated, especially by mere words, especially by mere words over the internet. But many people find such hostility so unpleasant that they shy away from such sites. This would have at least some effects of the relevant kind.
It seems like Balkin has got to be right about this part, though: blogs are almost always more permeable to opposing opinions than broadcast television. Is this enough to prevent balkanization? I don’t know. If I had to guess, I’d guess yes; but real data would be required to do anything other than guess about this issue.
 “But most bloggers who write about political subjects cannot avoid addressing (and, more importantly, linking to) arguments made by people with different views. The reason is that much of the blogosphere is devoted to criticizing what other people have to say. It's hard to argue with what the folks at National Review Online or Salon are saying unless you go read their articles, and, in writing a post about them, you will almost always either quote or link to the article, or both. Ditto for people who criticize Glenn Reynolds, Andrew Sullivan, or Kos, or Atrios. If you don't like what Glenn said about Iraq, you quote a bit of his posting, link to it, and then make fun of him. These links are the most important way that people travel on the Web from one view to its opposite. (And linking also produces a good check on criticism because you can actually go and read what the person being criticized has said.).”
[3/c] True, but possibly irrelevant. When people worry about cyberbalkanization, I don’t think they are worrying that people will develop cyber-enclaves (e-nclaves?) in which they will be completely and totally shut off from people who disagree with them. Even Rush Limbaugh quotes his enemies. Yes, there’s lots of linking and quoting. Is this enough to prevent cyberbalkanization? Again, empirical question. Balkin’s points are important but, I believe, inconclusive. If we are really interested in this, we’ll have to collect some actual data.
But remember: I think that balkanization is a phenomenon of secondary importance. I’m primarily interested in the (potential) decay of public political discourse. Here Balkan notes that the primarily critical nature of blogospheric political discourse fights against balkanization. I have some doubts about that, but even if I didn’t I’d be worried that the primarily critical atmosphere of the blogosphere would do more discursive harm than good.
But again: at this point, we’re mostly guessing.
 “In addition, most bloggers have blogrolls which include a wide variety of different sources with very different ideological views. If you check my blogroll, you will see that it contains both lefties and righties, and among the righties, a fair dose of libertarians like my favorite freedom loving gang at the Volokh Conspiracy. Because I am a lefty, I probably have more lefties than righties on that blogroll, but what's important is not whether there's a perfectly proportional distribution but whether there's a substantial variety of different views. There is, and I would wager that my blogroll is not at all unusual in that respect. The customs of the blogosphere produce this pluralism.”
[4/c] I’d wager otherwise, but, again: empirical question. And I get the feeling that Balkan does not constitute a representative sample. Aside from the fact that he’s interested in this issue at all, he seems more reasonable than the average blogger to me.
Anyway: I have righties on my blogroll, too. But, to my great shame, I check those sites out far less than I check out the liberal sites. And I basically never check out sites on either extreme of the spectrum at all. That proves little, since without the web I’d pay little attention to the extremes anyway; so I doubt that the web makes that part worse. But, at any rate, I doubt that it’s true that “what’s really important…[is] whether there’s a substantial variety of different views” [on the blogroll]. What really matters is whether people are visiting the sites that express views they disagree with and staying around long enough to read anything.
Or, rather: what really matters is whether they are getting more or less exposure to opposing viewpoints than they would get without the web (or than they would have gotten 20 years ago). Or, rather: what really matters is whether, despite their exposure to opposing opinions, people are becoming more ideologically balkanized because it's so easy for them to find communities of thought in which their opinions are confirmed without question.
 “The other fear often expressed is that Internet speech will become more extreme. There is a lot of extreme speech on the Internet. And there is a lot of personal invective, too. The Internet is not a debating society held in the Senior Common Room. It is often quite raucous and unpleasant. But the reason for this is *not* the group polarization mechanism Sunstein is concerned with-- the notion that people of different views aren't talking to each other so they gravitate to increasingly extreme positions. The reason why Internet speech is often sharp and unpleasant comes from the fact that people are talking to each other but are *distanced* from each other. It's very different saying something nasty to someone in a blog posting and saying the same thing to their face. (It's even easier to be nasty when one is anonymous, but even non-anonymous postings on the Internet give people greater license to vent than in-person interactions.).”
[5/c] This may be true. If so, then it is significant to the extent that we are primarily concerned with relatively more theoretical questions about balkanization. It is less significant to the extent that we are primarily interested in relatively more general and practical questions about the effects of the blogosphere on the civility of political discourse. Since I’m more interested in the latter question, it matters less to me whether it is the blogosphere’s impersonality or its (putative) balkanization that is more responsible for this lack of civility.
Which is not to say that the more theoretical question is of no interest, of course. For one thing, if anonymity is to blame, this may provide an argument (the only potentially cogent one I've encountered so far) against anonymous blogging. Andrew Sullivan raised (different) objections to anonymous blogging in his radio appearance with Atrios. Sullivan’s point was pretty clearly not a very well-thought-out one, and seemed to be primarily just a cheap shot at Atrios. But that doesn’t mean he might not be right about it; the point is at least worthy of some thought.
 “Even if Internet speech has its share of heated and unpleasant exchanges, the blogosphere has also shown, I think, that fears of group polarization produced by the Internet are overstated. It's important to distinguish distribution of viewpoints from polarization of viewpoints. The Internet allows for a much wider distribution of ideas to be expressed than in the traditional unidirectional mass media, but that is not the same as increasing group polarization. Indeed, wider distribution along multiple dimensions is the opposite of polarization, which is an increasingly tight bimodal distribution along a single dimension.”
[6/c] True and interesting points. But like many true and interesting points, these don’t show us that the alleged problem isn’t a problem, they allow us to understand the alleged problem more clearly. What Balkan shows here is that the terms ‘polarized’ and ‘balkanized’ suggest different phenomena in the context at hand and taken relatively more literally. (Both terms are being used at least somewhat metaphorically in this context, though.) We become polarized when we separate into two groups and move apart along one dimension. I suppose we become balkanized when we cluster in to enclaves, and there is a strong suggestion that the enclaves communicate little, and are at least somewhat hostile toward each other. But, again, if we are worried about detrimental effects on discourse, we can’t be comforted much by being told “don’t worry, we’re not getting polarized—we’re getting balkanized!” Separating ourselves off into ideologically cleansed cyber enclaves might not be as bad as becoming polarized (though it might, for all I know, be worse), but I don’t think it would be good. Again, we don’t know that it’s happening, but we don’t know that it isn’t, either.
 “We should also distinguish extremism among relatively small groups (like neo-Nazis) from society-wide group polarization. The Internet does allow like-minded people with extreme views to find each other. But that is not the same thing as group polarization in the Internet as a whole.”
[7/c] Good point.
 “If the concern is that *a small group of people* with extreme views will be able to meet others of similar views on the Internet and that their views will become even more extreme in the process, that may well occur. In that case, however, what you are really worried about is that people with extreme views might find each other in the first place and recruit other impressionable people, and preventing *that*, I would submit, is a blatantly unconstitutional goal.”
[8/c] True, but not relevant. Nobody should even be considering taking away any non-convict’s freedom of association. But that doesn’t mean that the internet isn’t making it easier for kooks to flock together, and it doesn’t mean we shouldn’t worry about it. There are almost certainly certain kinds of kooks that are relatively harmless in ones and twos, and that are rare enough that, without the internet, they’d never come together in dangerous numbers. But in the internet age, these previously harmless kooks can come together and become dangerous. Perhaps very dangerous, perhaps not. We don’t know. But it might be worth thinking about the possibility and keeping our eyes (and minds) open. The printing press probably had a similar effect, though, so none of this means that the sky is falling.
 “If the concern, on the other hand, is that *society as a whole* will become more polarized as a result of Internet speech, I think the fears are greatly overstated. The blogosphere continually provides a check on people's more extreme claims. It continually throws people together who have clashing views. Its architecture allows a wide dispersion of views to contend, a phenomenon which should not be confused either with an echo chamber or with group polarization.”
[9/c] There are really two questions here: (a) is the blogosphere balkanized? And (b) does the blogosphere have an effect on discourse and society at large? Balkin’s argument is that the blogosphere won’t contribute to the balkanization of discourse and society at large because the answer to (a) is ‘no.’ Again, I suggest that Balkan is probably right but not obviously so, since it isn’t really clear that the blogosphere provides such checks, throws such people together with sufficient frequency and with the relevant effects, etc. The architecture of the blogosphere does ALLOW views to contend, but the question is: are they doing so? Or, rather, are they doing so productively? Or, rather: are they doing so at least as frequently and at least as productively as they were before the advent of the web and the blogosphere? And does the blogosphere have a positive effect, a negative effect, or no appreciably effect at all on discourse and society at large? I assert: unless no
meduim of communication has an effect on society at large--which simply cannot be true--the blogosphere has some
effect. (Not that Balkin woud disagree with that claim.) The question is: what kind of effect?
 “I'm not trying to be a Polyanna here. I'm not claiming that no group polarization effects could ever occur on the Internet, or that Internet speech is necessarily going to make the world a better, safer place for democracy and/or reasoned discussion. What I am claiming is that fears that the Internet was going to produce a significantly greater tendency toward group polarization seems wrong. I think, in fact, that people's fears and anxieties about loss of control over the traditional public sphere governed by mass media have been projected onto the Internet.”
[10/c] I hope Balkin is right and I suspect he is. But as I noted above, we may miss the forest for a tree if we focus overmuch on balkanization per se
. Much of Balkin’s optimism is based on his claims that balkanization won’t produce polarization and that blogic hostility isn’t caused by balkanization. But even if he’s right (as, again, I suspect he is), that could still leave us with the problems of balkanization and increased hostility. But is Balkin right when he claims that there’s no cyberbalkanization at all? Maybe. But why speculate? We could get evidence by counting links on sites and analyzing traffic patterns on the web. My guess is that it won’t be too long before something like “web studies” comes into being in academia and elsewhere, and, so, it won’t be too long before we have some real data that bears on these questions. I have a knee-jerk reaction against such hip new hyperspecializations (and intellectuo-balkanization in general), but in this case I might make an exception.