eCHo CH A MBeRs = DeMoCR AC Y David…

    eCHo CH A MBeRs = DeMoCR AC Y



                          David Weinberger




  "   A democracy needs such "echo chambers," even
      though their discussions inevitably appear like nothing but a



                                                                "
      bunch of homogenous supporters rah-rah-ing each other.




t       alking together is the fundamental political act. While the
        Internet is certainly providing new features and new forums
        for talk, it is not transforming the near-genetic basics of how
human conversation works. In this case (despite the overall premise of
this anthology), the technology isn't changing the nature of democ-
racy so much as clarifying our understanding of democracy. And that
may be no less important.
     Our confusion about the role of conversation in democracy is mani-
fested in the persistence of the question whether the Net is enhancing or
dismantling the political conversations we think essential to democracy.
Rather than opening us up to a wider range of opinion, is the Internet
barricading the doors of belief? Will we use the fact that we have more
control online to hang out exclusively with people like ourselves, or will


                                    
David Weinberger                                                   n    


we use the frictionlessness of web connectivity to engage with people
from different walks of life? Will the Internet become an enhanced pub-
lic forum or a set of "echo chambers?"
     We've been unable to resolve these questions for three reasons.
     First, the Net is too young and is not yet what it will be. We don't
know what effect it will have once its first generation of users has grown
up with it as a ubiquitous part of civic life.
     Second, the empirical research that exists is extraordinarily hard to
interpret. Do we look at the patterns of links between websites? That
doesn't necessarily tell us how the information flows. Do results vary
based on topic? Over time? By demographic? Perhaps we form echo
chambers around political candidates but not cultural topics. Around
TV shows but not movies. Around reality TV shows but not sitcoms?
When we link to people with whom we disagree, are we cursing insen-
sibly at them or engaging in a rational back-and-forth?
     Third, even if we knew which vectors to follow, we would still
have the enormously difficult task of comparing the results to the state
of openness in the real world. As Yochai Benkler, the author of The
Wealth of Networks: How Social Production Transforms Markets and
Freedom, says, the question is not whether the Net will make our politi-
cal discourse perfect, but will it make it better. The law professor Cass
Sunstein reports that only low double digit percentages of links point
to opposing viewpoints, and Benkler is right in responding that he
doesn't know whether that's a cause for rejoicing or despair. To what
could we compare such statistics? To the percentage of space newspa-
pers give over to views that oppose their editorial positions? Typically,
that's a few Op-Ed columns and some percentage of the half-page of
Letters-to-the-Editor that papers run. How often do people read the
columnists they disagree with? How much time in the day do you
spend talking rationally and calmly about matters of state with people
    n                                    eCHo CHA MBeRs = DeMoCR ACY


with whom you disagree? How deep does the disagreement have to go
before you are too angry to talk, or simply see no point in pursuing the
discussion? Have you ever actually sat down for a long, respectful con-
versation with a neo-Nazi or an out-of-the-closet racist, a conversation
in which you're open to having your ideas changed?
     Me neither.
     The question, therefore, is not whether the Internet is closing us
down or opening us up, but rather what assumptions make the persis-
tence of online echo chambers--the same kinds of cliquish gatherings
that have always existed on land--seem simultaneously so urgent and
so hard to resolve.
     This urgency is undergirded by our belief that democracy is a
conversational form of governance. It's not enough (we believe) that
everyone gets to vote. Everyone also has to be able to talk about her
beliefs in public so that those beliefs can be well informed and well
reasoned. Yet when we look out across the Net, rather than seeing peo-
ple engaged in deep conversation, we see clusters of people saying the
most godawful things and, in so doing, giving permission to others to
say even godawfuller things. There's no denying the despair we all feel
when turning over certain rocks on the Net. Hearing sentiments that
are forbidden from the real world public sphere uttered in the perceived
privacy of the Internet legitimates those sentiments. This is worse than
an echo chamber: It is a room full of people egging each other on to the
most extreme and vile opinions. "You think you hate her? Here's how
much I hate her..." is not a helpful trope in a democracy.
     It would be foolish to argue that this never happens. But how
much does it happen? How important are such echo chambers? What
influence do they have on our democracy? And why have so many
people focused on them as the example of the Net's effect on democ-
racy? After all, we could look at hateful real-world groups and despair
for our democracy, but we recognize that such groups are the evil we
David Weinberger                                                   n    


have to live with in order to get the benefits of our freedom to assemble
and to speak.
     Echo chambers loom large in our thinking about the Web, not just
in our thinking about democracy. In part it's because some of the echo
chambers appear on highly popular sites. Thus, they are not equivalent
to marginalized extremist groups such as the KKK or the Stormfront
White Nationalist Community. Yet not all echo chambers are born
equal. Shouldn't supporters of a candidate have a spot on the Web
where they can be supporters together? Is a site an echo chamber if
it fails to rigorously challenge its participants' every view, including a
supporter's most basic commitment to his or her candidate?
     Further, the most prominent political sites--other than candidates'
sites--are not all the hatefests they're often portrayed as by the media.
Yes, participants encourage one another in their beliefs, but not all of
them are devoted to ever-tightening spirals of hatred. At the progres-
sive site HuffingtonPost.com, reasonable disagreements are common.
Present a calm argument against the progressive viewpoint of an arti-
cle, and you're likely to find just the sort of vigorous debate we want
for a healthy democracy, although it may be more rough and tumble
than we'd imagined. Trolls and hand-grenade throwers are ignored,
flamed, or moderated out, because, by definition, they're not looking
for a genuine discussion. Likewise, at the conservative Redstate.com,
reasonable discussion is the norm. (You can find plenty of examples of
awful interchanges, but you can find plenty examples of everything on
the Net.)
     Our picture of the Net as a set of hateful echo chambers is encour-
aged, too, by the premise that the only sites that matter are those with
hundreds of thousands of readers. That's how the mainstream media
works. But the Web is characterized by a "long tail" of sites with rela-
tively few readers. The echo chamber dynamic is facilitated by sites so
large that the commenters are functionally unknown to one another,
     n                                     eCHo CHA MBeRs = DeMoCR ACY


and the way to get attention is to be more outrageous than the previous
person. That dynamic is missing on the smaller sites that, in aggregate,
constitute the bulk of web traffic.
     Nevertheless, our focus on echo chambers, our notion that they typ-
ify Net dialogue, and our taking them at their worst, tell us something:
Our image of what a democracy should sound like is misconceived.
     For example, while we can map the links going into and out of
a site, and we can analyze the political positions of people who write
posts or comment on them, there is little actual data about the readers
of these sites. Perhaps the readers are diverse, even though the writers
and linkers are fairly homogeneous. Perhaps data would show that in
fact we've achieved the democratic ideal on the Web after all: People of
all persuasions are reading sites of every persuasion.
     Pretty lame, eh? Sounds like I'm grasping at straws to defend the
Net? I agree. In fact, that's my point. The previous paragraph is uncon-
vincing because we all agree that people generally don't spend a lot
of time reading that with which they disagree. We know that, on- or
offline conversation simply doesn't work that way. Never did. Never
will. Conversation finds an area of agreement and then explores the
differences. It hardly ever in our lives is an isolated exercise of pure,
unfettered rationality in which we suspend core beliefs in order to
think again about what those beliefs ought to be. Even taking that as
an ideal requires a picture of rationality that is unrealistic. Pure reason
is a better corrective than architect.
     So, what good does conversation really do in a democracy? It
helps us work out differences based upon shared ground. Conversa-
tions shape our existing ideas and occasionally generate new ideas that
are in line with our existing beliefs. We can probably count the times
on one hand that conversation changes our minds about anything
important.
     That doesn't mean conversation is irrelevant or trivial. Even when
David Weinberger                                                   n   


conversation doesn't change minds, it serves other social roles, includ-
ing binding people together so they can engage in effective political
action building trust, community and political commitment. From the
outside that may look like an echo chamber, but that is how people
come to make common cause. A democracy needs such "echo cham-
bers," even though their discussions inevitably appear like nothing but
a bunch of homogenous supporters rah-rah-ing each other. Conversa-
tion among people who are in basic agreement builds relationships and
foments political movement. It also makes possible the rare conversion
of beliefs, and, when done in the public forum of the Net, it leaves
traces by which opposing views can understand--and thus tolerate--
one another better.
     The persistence of "echo chambers" on the Net is not a failure of
democracy. Rather, their continued existence is evidence not only of
the fractures in our society, but of the gap between our ideals of democ-
racy and the mechanics of human social intercourse. We are never able
to stand fully apart from our commitments in order to evaluate them
in the cool light of rationality. If the Net does nothing but help us
accept the primacy of standpoint over reason--while leaving reason
some footholds in the wall of belief--it will have done our democracy
the valuable service of making it more realistic.



About the Author
David Weinberger is a fellow at the Harvard Berkman Center for Internet
& Society. He was an adviser on Net policy to the Dean and Edwards
campaigns. He is a co-author of The Cluetrain Manifesto and author of
Everything Is Miscellaneous.