Three Mistakes in the Moral Mathematics of Blogging – A Reply to Clay Shirky

Since I seem to have reignited the debate about justice and networks (with a good bit of highly-connected “amplification” by Joi and danah), I feel a need to reply. Most of all I'd like to thank Clay (1 2 3), Joi (1 2), Stewart, danah and the many commentators for the inspiring conversation.

Derek Parfit – from whom I borrowed the title of this post – writes towards the end of his ambitious book Reason and Persons (1984): “[Our many false beliefs about justice and ethics] did not matter in the small communities in which, for most of history, most people lived. In these communities, we harm others only if there are people whom each of us significantly harms. Most of us now live in large communities. The bad effects of our acts can now be dispersed over thousands or even millions of people. Our false beliefs are now serious mistakes."

I’d like to point out three mistakes that are common in the current debate about the justice of networked forms of organization. No doubt much hinges on how we spell out the concept of a just institutional scheme. But before we even go at that issue it’s good to clear some ground by avoiding these three mistakes in the moral mathematics of blogging.

The First Mistake - “Natural Social Institutions”

The first mistake – lets call it the “Natural Social Institutions” view – is the simplistic but widely held view that the patterns resulting from the operation of freely forming networks are acceptable because the rules of operation of these networks are in some sense natural. “Diversity plus freedom of choice creates inequality, and the greater the diversity, the more extreme the inequality,” Clay writes. “In systems where many people are free to choose between many options, a small subset of the whole will get a disproportionate amount of traffic (or attention, or income), even if no members of the system actively work towards such an outcome…[I]t arises naturally.”

Diversity plus freedom of choice creates inequality, yes, but how much inequality comes out in the wash is determined by a complex mix of institutional arrangements – including informational feedback mechanisms – as well as other factors influencing individual linking behavior. Clay has acknowledged as much by pointing to David Sifry’s Technorati Interesting Newcomers List and later by sketching several possible strategies in modifying the power law distribution. But Clay avoids the mistake only part of the way. He still gives the “natural power law” a kind of moral priority in his picture. The reason this is a mistake is that there is no way in which we can meaningfully say that “the blogging world without the Technorati Interesting Newcomers List” is in any way natural, or the baseline from the point of view of justice, in comparison to “the blogging world with the Technorati Interesting Newcomers List.” Neither has a special claim to be the baseline of moral analysis. It’s not as if there is one distribution and then we tinker with it. In order to answer the question of justice we need to agree on some further point of view from which to judge the justice of the rules and the resulting distributions.

The Second Mistake – “Links from Nowhere”

The second mistake – lets call it the “Links from Nowhere” view – claims that link choices are made under full information about available options and fully formed values or preferences over those options. We should also reject this view. Autonomous linking choices are always informed by incomplete information and incomplete values and preferences. There are in fact no links from nowhere. Recognizing this mistake helps us understand the role of informational agencies - like the Technorati Interesting Newcomers List – i.e. nodes that feed information about the network back into the system.

The second mistake is closely related to the first and it is, no doubt, a bit of a strawman i.e. it is not held in this extreme form. The basic point is that the nodes in the freely forming network of blogging are agents that make choices with incomplete information, and use different tools to deal with this incompleteness. The kinds of informational agencies and personal information tools we create will have an impact on linking behavior. This point is in line with the last of Clay’s strategies in influencing the distribution, namely, different ways of making “the system more dynamic.” Note that while the debate on inequality refers frequently to the descriptive theory of networks it is odd that this theory borrows the language of individual choice without having any account of human choice and incentives within the theory itself.

The Third Mistake – “Forced Compensation”

The third mistake – lets call it the “Forced Compensation” view – claims that the only way to address the unacceptable degree of inequality that results from the operation of a freely forming network is to “force” people to change their linking behavior. This is a far too narrow view of the means available to influence the distributions that arise. In reality a number of factors influence the overall distribution of links. Clay writes that “[b]ecause it arises naturally, changing this distribution would mean forcing hundreds of thousands of bloggers to link to certain blogs and to de-link others, which would require both global oversight and the application of force. Reversing the star system would mean destroying the village in order to save it.”

Clay has recently considered more ways that the distribution can be altered. In other words he has taken Steven Johnson's question seriously: “What architectural changes would fight against the power law trend, without doing it in a command-and-control kind of way?” But I think Clay is still partly in the grip of a narrow mistaken picture. At the very least we can think of four objects that causally influence the kinds of distributions that arise in the blogging world. First are groundrules, or the basic rules governing linking behavior (e.g. caps on links) or the kinds of informational agencies that are to be promoted (e.g. more Newcomer lists of different kinds). Second there are the non-coercive regularities that arise strategically for individuals playing the blogging game e.g. the practice of reciprocal blogrolling or the rapid repeat posting of hot items. Third, there are conventions that cannot be justified strategically but are the “traffic rules” of the blogging world including permalinking and other features. Fourth, is the ethos or the “spirit of the bloggin age” if you like, the regularities in preferences governing personal choices. These could possibly be the ethos celebrating a concern for newcomers or the ethos celebrating yet unpopular nodes of exceptional fitness. It is clear that a wealth of parameters can be tweaked to influence the outcomes of a freely forming network.

Where do we go from here?

Clay writes: “The interesting and hard question is “Since there is to be inequality, how shall it be arranged?” I think we are going to see an explosion in work designed to alter the construction and effects of this inevitable inequality…and I am optimistic about this change, as I believe the concentration of real thought and energy on what is actually possible, as opposed to cycles wasted on utopian declations, will be tremendously productive.” I can only agree and I too am optimistic. As we go forward we need to think not only about the distributional effects of different architectures and tools, about the roles of different amplification mechanisms to use Joi’s phrase. We need to also focus on the hard normative questions:

What arrangements of inequality are preferable over others from the point of view of justice? How do we justify to each other the rules, architectures and tools we adopt in the blogging world?

In answering these questions we should look back to understand the present. John Rawls put the task description well: “The task is to articulate a public conception of justice that all can live with who regard their person and relation to society in a certain way. And though doing this may involve settling theoretical difficulties, the practical social task is primary.”

A public conception of justice for freely forming networks. That could be our shared goal.

Comments

[cbrayton]

This is a really excellent analysis. You should ping http://topicexchange.com/t/social_software/ with it. However, I don't understand the point of citing Rawls against the Shirky point of view: Shirky seems like the most Rawlsian of Rawlsians to me, but begs the wider question of barriers to participation. The blogosphere is overwhelmingly Anglophone, for example (though blogs in Persian, French, Polish, and Brazilian Portuguese, are signficant minorities). The fact that Francophone sub-Saharan African bloggers command little attention in the "market" is not due to some power law militating in favor of English as an international lingua franca (as Shirky has argued), but to the fact that Africa's teledensity is a tiny fraction of the developed world's. The question of justice is not the opportunity to make the Technorati 100, but the ability to publish at all to an audience equipped to receive your content. In Rawlsian terms, what self-interested rational calculator, deliberating behind the veil of ignorance, would apply the maximin principle in favor of a network in which complete lack of access to the network is the minimum outcome? If the point of our utilitarian moral calculus is to add as much utility to the network as possible, what rationale could there be for excluding potential participants? What sense can the question of choice have without univeral suffrage, so to speak? So I submit that the question of inequality among participants is beside the point. Inequality among participants and nonparticipants is the more pressing moral question, isn't it? But Shirky considers the Digital Divide a pernicious urban legend (http://www.shirky.com/writings/half_the_world.html). Robert McChesney: "In a class-stratified, commercially oriented society like the United States, cannot the information highway have the effect of simply making it possible for the well-to-do to bypass any contact with the balance of society altogether?" (http://jcmc.huji.ac.il/vol1/issue4/mcchesney.html)

[Tom Steinberg]

It is worth remembering here that even for the most ardent marxist, inequality is not a social problem in itself - it is only a problem in terms of what outcomes it causes.

Increasingly, I find myself looking at the tail end of the blog power law curve as the interesting part, and the peak as more of a curiosity resembling the 'most phoned telephone numbers in country X'.

I wholeheartedly approve of the idea of identifying a conception of justice for free forming networks, but the first question in that task has to be - "What harm results, or might result from inequality in this sphere?"

[Jeff Warren]

I agree with and support the theoretical basis for what you are describing- I'd be even more interested in starting up and trying out some techniques. For example, the Technorati list is manually compiled, I suppose? What about an daily updated graphic of the popularity of an open list of blogs (could this be generated via Perl?), which could be displayed with various statistics. Say, add stats for whether people frequent the site, or how much time they spend there. Find a new metric for determining the quality of a blog.

Again, how to gather the information becomes an issue, but I think that the compelling part about such a plan is that it offers ways of influencing the power law trend in new ways. Granted, it's easy enough to say that the inter-linking system of promoting blogs lends itself toward an injustice, and it seems likely that a new system would, after reaching a limited equilibrium, become just as unjust. But just as Clay described the power law trend as a cycle, where new tools eventually lead back to unjust popularity distributions, perhaps a continual migration to new tools is how the system remains dynamic. One could certainly make that case in government and world history.

[antoin o lachtnain]

I think that the question 'how should the network be arranged?' may be the wrong question, or we may be asking it before we are ready to deal with it, because it is so attractive as a mathematical problem. The right question is 'what is the network for?' If we knew what the network was for, then we could pick out the right arrangement for it.

In practice, the network of weblogs is for lots of different things. Most weblogs in the world are just to keep friends informed. That's all they are. They need a pretty flat, freeform structure. On the other hand, some weblogs are part of an idea-forming and decisionmaking process. In that case, good ideas need to be able to float to the top of the pile periodically to be implemented, and the discussion has to happen within a time framework. The network should be designed to give exposure to good, practical, well-expressed ideas, rather than simply giving exposure to people who might have been prominent in the past. (Exactly how this might be done, I'm not quite sure.)

[Jennifer Leonard]

Marko,
Great analysis. Perhaps you'd consider posting to the newly launched Civiblog (http://www.civiblog.org) out of the Citizen Lab at the Munk Centre for International Studies, University of Toronto. The purpose is to create community and open dialogue for all persons working/studying/volunteering in 'global civil society' and so this line or discussion is particularly relevant. Thank you!
Jennifer

[Latasha]

Great ingsiht! That's the answer we've been looking for.