Benkler, Jenkins, Sunstein: The good, the bad, the ugly

(Note that the order of names in the title is not meant to imply a specific relationship to the adjectives in the subtitle. I went with alphabetical.)

Last night I grabbed the red line over to MIT after work and saw a wonderful event put on by the MIT Communications Forum, entitled: Our World Digitized: The Good, the Bad, the Ugly.

Fuzzy photo of Sunstein, Benkler, and Jenkins

Audio from the event (Real Audio format) has already been posted, so if you find my chaotic “liveblogging” style notes below interesting, be sure to check it out. (They will post video on the seminar page soon as well).

My *very rough* notes:

HJ: This is not about the Utopian versus the Dystopian – it’s about what criteria can we even use to measure the success or failure of online democracy.

Jenkins started with three quotes (two from Sunstein and one from Benkler) which define the needs of a healthy society.

HJ: Do you agree with each other on the criteria for valuing free expression? What current practices promise to encourage or deter achievement of these goals?

CS: Self-reflexive culture, individual freedom – these all sound positive, though I have some questions about what that might mean. To grade the internet or the culture on this is difficult, but I think I’d give it a C-. The babel and the excellence, the chaos and the order, the brilliant insight and the cruelty are not – from the standpoint of a self-reflexive culture – what we deserve. So much of what is out there is the opposite of self-reflexive.

YB: One place we may differ is the degree to which the practical constrains on action determine how we evaluate the level of normative life lived as a practical matter. ? Trust and reciprocity are agreed goals. But the practice of pervasive collaboration on the net helps reinforce these values in a way very different than the mass media method of demonstrating authority in a centralized, controlled fashion.

Another difference is the question of participation – the CS quote suggests too much passivity on the part of the citizen.

Coherence versus practical independence – it is the ability to act that I am focused on rather than the notion of correctness coming from a mass media – messier, but more capable of correction.

CS: ok, maybe C- was just difficult grading. Passive, authority driven, non-participatory mass media was worse – at least for those who have the ability (means, confidence, digital literacy) to be participants in the construction. The qualities I describe as being required for “A well functioning system of free expression” are necessary but not sufficient – there are other things (active participation) that are necessary.

Absolute “freedom of choice” in the information sphere verges on committing the error of Chicago School economics – in a well functioning system you don’t construct a daily me which is your information cocoon – you see things you did not choose to be exposed to.

In a heterogeneous democracy like ours, the ability to self-sort has dangerous implications – shared experience and unanticipated exposure to information is necessary to make democracy function.

YB: I think this is the closest I will ever get to the University of Chicago economics department.

Power is the core distinguishing feature between what I see as my position versus what Cass has laid out. I would resist any alternative that is not based on the actually available alternative actions. What we can’t get away from is that the production of shared experiences in the mass media context – the sharing of experience was very much in the hands of a government controlled agency or a mass commercial entity. (As someone said, “Newspaper men are the people who write on the back of advertisements”).

This is a common citizen experience which is already dependent on a series of governmental and commercial entities and power systems.

A greater degree of freedom requires a greater set of capabilities both in terms of access to raw equipment as well as skills / digital literacy.

What if what makes our democracy better is a question of thousands of people who can set the agenda, rather than hundreds. There is still an elite, but it is a much broader set of people who can push on power in different ways. As an example I’d use the influence the Net Roots have had in Democratic party politics – that is a form of participation and making common cause which is much more democratic than the previous alternative.

Question 2: The question of the digital enclave and the “babel hypothesis” – what is the real danger of fragmentation.

YB: The question is certainly one on which the Berkman Center and many other people are still investigating. Mapping the shape of discourse on the net through link analysis – the goal being to see who links to whom and when. Our data suggested that what we saw was not a “daily me” (perhaps a “daily you”) – what we have is a structured public sphere – central nodes, with a fairly shallow network.

The daily me was a prediction which has turned out to be not the universal practice – instead we have many different kinds of ways in which news is filtered and distributed, from Google News to social networks and collaborative filtering.

CS: Boulder / Colorado Springs discussion groups – liberals talking to each other, conservatives talking to each other. What happened is that in every case the individuals got more extreme, got less diverse internally. After a pretty short discussion they came into line with one another. That’s an empirical claim.

The theoretical claim is that if people are using the internet in this fashion – and replicating the experiment – that suggests we have a problem. The belief that freedom of choice (ability to select their own information, ability to reproduce their own believes) in and of itself leads to a better democracy is what I am concerned about.

The DailyKos as an example: 1% of its readers are republicans. Study involving the blogosphere and linking – the plurality of the cross linking (liberals linking to conservative blogs) was people heaping ridicule on each other. This doesn’t suggest a deep, cross-linked public sphere.

CS quotes YB to make his point – people cluster, people link to like opinions.

YB: Three steps back. How should we feel about the net as a structure for the public sphere. Well, compared to what?

I question the degree to which the structure of the net itself is particularly polarizing (as opposed to Fox news and talk radio) – we have many examples of significant polarization – say the 1994 election – which one certainly can’t blame on the net.

I completely agree with the person you quoted. This is also how we operate outside the net. That strikes me as a plausible description of how we have always been.

CS – brings up Jane Jacobs and the Life and Death of American Cities – the amount of serendipity one gets regularly has an impact on the clustering. The infrastructure of the internet can encourage or discourage that serendipity.

YB – YouTube. Think about the kind of panorama of human life you find there as something which was never possible before. Even the email with the “can you believe that!” – with the link to the other side’s argument – that is a form of interchange which was not possible before.

CS- to what extent are we seeing the colorado-ization of youtube. Are people who are Obama fans obsessively watching the “No You Can’t” McCain video, and the McCain fans watching reverend wright?

Question 3: Moving from YouTube to Wikipedia. What do you see it telling us about civil society and about the collective production of human knowledge.

CS: I love wikipedia. The fact that participation – while not as huge as is sometimes portrayed – is huge. Hayek on the dispersed nature of human information – Jimmy Wales says you can’t understand wikipedia without knowing Hayek. For wikipedia to work, there have to be norms in place for it to work. If you look at the history on various individuals you see the cruelty, the invasion of privacy, the lies – but the hierarchical nature of the organization is what enables that.

YB: Wikipedia is very interesting. When it was culturally irrelevant, or only relevant to those editing it – it was very much norms based. Now that it has become relevant to many people it requires different procedures to protect itself. But I worry about calling this structure hierarchical – I would call the processes behind wikipedia fundamentally democratic, open, and relatively transparent. What I find most amazing about wikipedia is how it can be so great while also being imperfect. The distance it has come since 2001 when Jimmy Wales put up 900 stub entries on a platform which allowed anyone to edit – is amazing. As an existence proof – “and yet it moves” – to drive us to understand that we need a new model of understanding human collective behavior – wikipedia is fundamental.

CS: Seigenthaler incident – where his wikipedia site was vandalized to say he had been accused of having been involved in the assassination of the Kennedys. The sad fact is that the number of living persons who have been subject to vandalism is quite high. Often only for a short time, but the reality is that even for a short time it can have an impact. But Britannica doesn’t even have that. at all. The elaboration through collaborative development of social norms – as in the case of the wikipedia editors norm of ‘dignity’ in historical biographies – is compelling and does point to development which could be helpful in other contexts in the future.

Q: What skills / tools do citizens need? What is expected of citizens in the new digital age?

YB – The belief that you can impact the agenda changes the way you behave. How are the skills involved in affecting the agenda being learned? Practice, practice practice. Some people criticize youth culture for not being political in the use of new media they miss the point that kids play, and in play learn skills they will use as an adult. It is the attitude and the behaviors that are important. This is where education comes in – learning by doing, how to seek and find information, and act upon it, and have an impact on the world – this is very important. To teach both the technical skills and the attitudes. In this regard it is much more difficult for me to be sanguine about the even distribution of these skills and attitudes. Part of the problem is our overall risk averse culture – the assumptions about what kids can and cannot do at age 9 and 10 today versus 30 years ago. It is like the city (referring back to Jane Jacobs) – there is more risk and more danger there, but we would not want to give up what the city makes possible.

CS: There is an important distinction between citizen and consumer. We have different expectations of the citizen in reflecting on legislation, for example, Some celebrations of the internet in the early days seem to belie the difference between the two. Something like a spirit of charitable understanding – which assumes that the person with whom I disagree is acting with integrity and doing the best they can with the information they have available.


Audience Q: To what extent do the tools available change how we communicate, if we are “wired” in certain ways?

CS: One virtue of the internet is its potential to encourage curiosity over clustering, and to a large extent the success of this depends on education writ large.

YB: I wonder about the assumptions about how much of us is hard wired. The basic question of the relationship between the tools we adopt versus the way in which we change while using those tools – I think we’re also very plastic. Not perfectly plastic, but plastic and perfectable or capable of constant change.

Audience Q: About Prof. Benkler symposium on cooperation – what have you learned, or how has it changed your understanding?

YB: I learn every week. A few examples – paper published in Nature about punishment and its use as an example of cooperation, another on AI and collaboration in task based environments. Lots of case studies of how people collaborate – the DailyKOS as a collaborative effort of lots of people working together to create a discursive environment. A presentation last week on couch surfing as a phenomenon. What we do is try to understand what goes in to making collaboration successful. When you are in environments which are not as thick or engaged as traditional communities with rich physical history – experimental settings or online in the absence of rich history – what motivates collaboration in those settings and how is it different than what many studies of traditional communities (irrigation districts, for example) have show.

Audience Q: Is there a possibility of unintended consequences – that as communities become large and successful do they end up creating less depth of communication? You can’t talk to all the people all the time – just as you don’t say hello to people in the street in the city.

CS – Yes – this is why there is importance in large scale social shared experiences.

Audience Q: You both discuss Wikipedia but other mediawiki attempts, like WikiNews, have been not as successful. Why not?

CS – Lostpedia is highly successful. Where people have an intense interest in both production and output and have good norms, they can be successful. Perhaps WikiNews is too broad – what community are they serving. In one sense, the internet itself is wiki news. Perhaps Wikipedia and Lostpedia trigger norms of cooperation because the users sense that what they are providing is valuable and is not available elsewhere.

YB: Also the time cycle is different in news – the time it takes to work things out, and that there are many other outlets for news reporting which is highly opinionated – you need a critical mass of people who believe passionately about objective reporting who are not already in the professional media.

Audience Q: Thinking of the news media in other arenas they have much greater sense of polarization and don’t have the ideal we do of objective neutrality or don’t have it the same way – 36 parties in the dutch parliament – makes you used to relativizing and making alliances in a way the two party system does not. What are the longer term implications in terms of how we organize ourselves politically? How do we get beyond two parties?

CS – Is the fact that we have a two party system what encourages / drives / inclines / requires a great degree of polarization? I don’t accept that. You can imagine a situation in which 12 parties each think of the other 11 as wholly confused and lost. You can also imagine scenarios in which 2 parties see each other as diverse and some of them are better than some people in my party, etc. I don’t think it is as simple as saying other multiparty scenarios are necessarily better than ours.

YB: I can certainly speak to that scenario having grown up in Israel where there were dozens of parties each of which had a monopoly on truth. The two party system if anything compresses the variety of options – pushes us toward the center if anything.

Audience Q: I feel a bit like I’d like to speak up for polarization and clustering. Given how little power people feel they have, the fact that they might group together and find consensus and act on it may be better than not doing so. This doesn’t mean there isn’t value in the public sphere and in the broader perspective, but there is something valuable in clustering.

CS: The irony of course is that those of us concerned about like minded people only listening to each other have to be open to this objection or risk proving our own concern. But there are some empirical studies which show that people who talk to like minded people actually become less politically active. Engagement is often fueled by the construction of the Boulder vs. Colorado Springs – the other which we argue against is important. There’s also first order diversity and second order diversity. In other words diversity within and diversity across – ie, Brigham Young exists and Berkeley exists, rather than having each university need to have it’s own BYU and Berkeley inside. The second order diversity (each pool talking to itself and building clusters) actually creates more opinions over all – having a BYU which argues against a Berkeley causes each to create more persuasion and more ideas. You basically get larger argument pools from which people can then draw. The worry, of course, is that you get large pools of people who cannot engage productively with people who disagree.

YB: If one sees democracy as a pluralistic competition of different views, you are fine with clustering. But if you have a view of democracy which requires some more substantive culture, commitment, learning that is shared by the whole – if you believe there is a need for this shared public sphere then you worry about the lack of the center and shared experience.

HJ: Into the lightning round. Greater brevity, on both sides.

Q: Can collective intelligence / production replace the need for institutions in developing countries?

YB: I was with you until you said “replace.” Collective intelligence can improve development in many areas related to education – which is a lot, but not everything. Access to basic needs like food and shelter? No. But access to open innovation networks supplementing weak education? yes, it can contribute.

Q: Lostpedia, acknowledgment of fragmentation where people obsess over frivolous things – but then play can also be practice for other kinds of civic engagement. What is the role of frivolous activities? Are they more than preparatory play? Is there some value in that play itself?

CS: Both. Lostpedia is an intrinsic good to those who enjoy it. So those of us concerned about citizenship should not overlook the role of play. Some of it is preperatory and some of it is valuable in its own right.

HJ: A follow up – digital enclaves make sense where sites declare themselves ideological to begin with. Is it possible that non-political sites (Lostpedia) might be sites for a more diverse conversation?

Q: How do you reconcile your critique of the deliberative majority in your book with your respect and hope here for the dignity standard of wikipedia.

CS: We know that deliberative bodies have a number of major failings: cascades, groupthink, reinforcement of initial mistakes, shared information crowding out uniquely held information. The advantage of wikipedia is that these mistakes are suppressed – partly due to the sheer number of people, due to the opportunity for correction by disconnected individuals, and partly due to the motivations of those individuals participating.

The dynamic for the wiki, and wikipedia in particular, is very different than in other deliberative bodies (like blog networks or forums of like minded individuals).

Q: What about the increasing permeability between the offline world and online world. For example the group Anonymous, which hopes to combat the Church of Scientology – this is a small minority collaborating to amplify their voice. Should we be worried about these types of effects, as we celebrate the empowering effect these tools have overall.

YB: If the meaning of should we be worried is should we pay attention to these things as we design our systems, of course. If the meaning of “should be worried” is that we should think of these new tools as a bad thing overall, I say no. This is fear of the freedom of the net – what it makes it possible for people to do, in groups, using the internet.

CS: I have a very positive reaction to the question. Powerline, for example – a blog network with some very smart conservatives. They’ve said some false things about people, which spread quickly and widely – there is a way in which the network can create real destructive behavior and reinforce it in a way the mass media did not.

Q: If you think of each site / network as a city with its own rules. Should we be looking specifically at certain sites in terms of how democracy is or isn’t changing at specific places rather than the web as a whole?

YB: The question isn’t really whether the metaphor is good or bad but in what contexts it is useful. Sometimes it is useful to think of an individual set of sites as a city and in other cases it is more useful to think of the broader context in which they are embedded.

Q: What about virtual worlds which have a visual / avatar component – what impact will that have?

YB: Too early to tell. The truth will be emerge in the practices which arise around the tools.

Q: What about real, video-based, face to face interaction.

CS: I think the reality is that it makes very little difference. The Boulder and CS experiment described above was face to face.

YB: Humanization does seem to encourage people to be more collaborative. But I am reluctant to accept that video will fundamentally change this equation given the practices we see online which compensate – have we already gotten all the humanization that you get, and the video has only marginal impact.