You are viewing a read-only archive of the Blogs.Harvard network. Learn more.
Skip to content

The Latent Community in Every Webpage

essay by Clay Shirky

On a Monday morning in March of 2006, forty thousand students in southern California stunned teachers and administrators by walking out of school to protest HR4437, a bill in the US Congress that would have mandated a crackdown on US immigrant populations. The largely Latino student group had been inspired to act by a similar protest that had taken place two days before. So many students walked from their schools to City Hall that they blocked traffic as they went, creating a very visible and public display for their cause.

There were several remarkable things about the protest. The size of the walkout alone made it unusual — getting tens of thousands people to take any coordinated political action is hard. Getting high school students to do so, when most of them are too young to vote, is even harder. Being able to do so without the school administration knowing was hardest of all — keeping a secret among 40,000 people has never been trivial. And doing it all in 48 hours should have been impossible — would have been impossible, in fact, even a year before.

The thing that made an instant, secret, and huge protest possible was the spread of new communication tools, especially MySpace and text messages on the phone. Armed with these tools, students were able to coordinate with one another, not just person to person but in groups. Using these tools, the messages they exchanged went to the people who mattered — the other students — without reaching the school administrators.

Making the protest possible, though, was not the same as making it happen. The thing that made it happen was real political feeling — the students had a message they wanted to send. MySpace and texting amplified that message, but they didn’t create it; the real story here is the increased ability of an otherwise uncoordinated group to achieve its shared goals.

Economists would describe this change as a positive supply side shock to liberty. Whenever you improve a group’s ability to communicate with one another, you change the things they are able to accomplish together. The old dictum that freedom of the press only exists for those who own a press points to the significance of the change; to speak online is to publish, and to publish online is to open the possibility of connecting with others. With the arrival of globally accessible publishing, freedom of speech now blends into freedom of the press, and freedom of the press blends into freedom of association.

With this blending of broadcast, conversational, and social elements into one medium, we have entered a world where every web page is a latent community — each page collects the attention of people interested in its contents, and those people might well be interested in conversing with one another as well. In most cases, that community will remain latent, either because the potential ties are too weak (any two users of Google are not likely to have much else in common) or because the people looking at the page are doing so separated by too wide a gulf of time. But in a growing number of cases, being able to synchronize large groups via social media is adding a new feature to traditional media; it is becoming not just a source of information but a site of coordination.

William James, the American philosopher, maintained that thinking is for doing, which to say that our brains don’t exist for purposes of abstract cogitation, they exist to help us decide what to do next. A similar transition is happening to media today: increasingly, publishing is for acting. Instead of just distributing information, many media outlets are also providing ways of gathering users, allowing them to come together as a group.

To put it in military terms, media can create ‘shared awareness’, the sense in a group that not only does each member understand what is going on, but that the understanding is similar among all, and, critically, each member understands this as well. Shared awareness is a useful precursor to coordinated group action, and given how recently we have had not just early adopters but whole societies online, most of the effects of this new leverage for freedom of assembly are still in the future.

Clay Shirky is an adjunct professor in NYU’s graduate Interactive Telecommunications Program (ITP), where he teaches courses on the interrelated effects of social and technological network topology — how our networks shape culture and vice-versa. He is most recently the author of Here Comes Everybody, the Power of Organizing Without Organizations.

Posted 15 May 2008 § Clay Shirky § The Networked Public Sphere ‡ Comments (4)

The Future of the Internet – And How to Stop It

essay by Jonathan Zittrain, response by Reed Hundt

Jonathan Zittrain: The Future of the Internet – And How to Stop It

On the 9th of January, 2007, Steve Jobs introduced the iPhone to an eager audience crammed into San Francisco’s Moscone Center. A beautiful and brilliantly engineered device, the iPhone blended three products into one: an iPod, with the highest-quality screen Apple had ever produced; a phone, with cleverly integrated functionality; and a device to access the Internet, with built-in map, weather, stock, and e-mail capabilities. It was a triumph for Jobs, bringing the company into a market with an extraordinary potential for growth, and pushing the industry to a new level of competition in ways to connect us to each other and to the Web.

This was not the first time Steve Jobs had launched a revolution. Thirty years earlier, the twenty-one-year-old Jobs exhibited the Apple II personal computer, the first machine for hobbyists who did not want to fuss with soldering irons: all the ingredients for a functioning PC were provided in a convenient molded plastic case.

It looked clunky, yet it could be at home on someone’s desk. Instead of puzzling over bits of hardware, Apple owners faced only the small hurdle of a cryptic blinking cursor in the upper left corner of the screen: the PC awaited instructions. Some owners were inspired to program the machines themselves, but true beginners could load up software written and shared or sold by their more skilled counterparts. The Apple II was a blank slate, a bold departure from previous technology that had been developed and marketed to perform specific tasks from its first day to its last day.

Though these two inventions—iPhone and Apple II—were launched by the same man, the revolutions that they inaugurated are radically different. The Apple II was quintessentially generative technology. It was a platform. It invited people to tinker with it. Jobs (and Apple) had no clue how the machine would be used. They had their hunches, but fortunately nothing constrained the PC to the hunches of the founders.

The iPhone is the opposite. Rather than a platform that invites innovation, the iPhone comes preprogrammed. You are not allowed to add programs to the device that Steve Jobs sells you. Its functionality is locked in, though Apple can change it through remote updates. The machine was not to be generative beyond the innovations that Apple (and its exclusive carrier, AT&T) wanted. Whereas the world would innovate for the Apple II, only Apple would innovate for the iPhone.

Jobs was not shy about these restrictions baked into the iPhone. As he said at its launch:

We define everything that is on the phone. . . . You don’t want your phone to be like a PC. The last thing you want is to have loaded three apps on your phone and then you go to make a call and it doesn’t work anymore.

No doubt, for a significant number of us, Jobs was exactly right. For in the thirty years between the first flashing cursor on the Apple II and the gorgeous iconized touch menu of the iPhone, we have grown weary with the unexpected very uncool stuff that came along with the unexpected cool stuff. Viruses, spam, identity theft, crashes: all of these were the consequences of a certain freedom built into the generative PC. As these problems grow worse, for many the promise of security is enough reason to give up that freedom.

In the arc from the Apple II to the iPhone, we learn something important about where the Internet has been, and something more important about where it is going. The PC revolution was launched with PCs that invited innovation by others. So too with the Internet. Both were generative: they were designed to accept any contribution that followed a basic set of rules. Both overwhelmed their respective proprietary, non-generative competitors, such as the makers of stand-alone word processors and proprietary online services like CompuServe and AOL. But the future unfolding right now is very different from this past. The future is not one of generative PCs attached to a generative network. It is instead one of sterile or contingently generative appliances tethered to a network of control. These appliances take the innovations already created by users and package them neatly and compellingly, which is good—but only if the Internet and PC can remain sufficiently central in the digital ecosystem to compete with locked-down appliances and facilitate the next round of innovations. The balance between the two spheres is precarious, and it is slipping toward the safer appliance.

If security problems worsen and fear spreads, rank-and-file users will begin preferring some form of lockdown. A software development kit for the iPhone is just being launched, intending to harness the value of contribution from outsiders while allowing control by and security through Apple: software authors will have to register, and their applications may only be made available through the Apple iPhone App Store, where Apple can control what may be installed on the phones. In turn, this sort of lockdown opens the door to new forms of regulatory surveillance and control. We have some hints of what that can look like. Enterprising law enforcement officers have been able to eavesdrop on occupants of motor vehicles equipped with the latest travel assistance systems by producing secret warrants and flicking a distant switch. They can turn a standard mobile phone into a roving microphone—whether or not it is being used for a call. As these opportunities arise in places under the rule of law—where some might welcome them—they also arise within technology-embracing authoritarian states.

We face a constitutional moment in cyberspace, not because of a watershed moment of oppression by a sovereign, but because of difficult choices of our own making: abuse of our open network and hardware by some among us, and a resulting mass movement towards – indeed, demand for – lockdown. Our future can be kept generative only if we can continue to see the Internet’s invitation to be participants in its use, rather than consumers of it. The path forward is illuminated by the coupling of technological tools – like wikis – that have promoted openness, with social customs and law – like those of Wikipedia – that solicit people to take an active part in building the world they want, rather than simply paying for it and expecting others to do the rest.

Jonathan Zittrain is the Jack N. and Lillian R. Berkman Visiting Professor for Entrepreneurial Legal Studies at Harvard Law School and the Chair in Internet Governance and Regulation at Oxford University. Professor Zittrain is a co-founder of HLS’s Berkman Center for Internet & Society and served as its first executive director from 1997-2000.This essay is drawn from his Introduction the recently published The Future of the Internet—and How to Stop it.

Posted 15 May 2008 § Generativity § Jonathan Zittrain § Reed Hundt ‡ Comments (1)

Market Structure Rules

essay by Reed Hundt, a response to Jonathan Zittrain

The personal computer has been “generative” — Jonathan Zittrain’s felicitous coinage – because of what Andy Grove, Intel’s famed CEO during the Internet’s salad days, called the horizontal structure of the industry. He referred to the nature of the value chain: competitors in microprocessors did not enter the software business; the software firms did not enter the computer assembly and distribution business. These firms could enter, but in each horizontal market they had their hands full in terms of keeping up with the pace of innovation, the impact of competition, and the growth of the overall market.

The old AT&T, pre 1984 break-up at the hands of the Reagan Antitrust Division, was non-generative, as I see it, because it was a vertically integrated business, one in which a single company aspired to sell everything from the client device (aka telephone) to local communications services to networked service to international long distance. The break-up was specifically designed to promote “generativity” by separating the local access market from the rest of the firm’s businesses.

Now the old broken-up AT&T has been reassembled, by dint of perspicacious mergers and successful persuasion of the Bush Administration, into two new versions of Old Ma Bell – these being AT&T redux and Verizon. Yet it is not at all clear that the two new vertically integrated Bells are non-generative. We don’t know yet.

But my suspicion is that they will be generative – that is, they will participate in a maximization of productivity gains for their sector and for the rest of the economy in which they play a vital enabling role – if and only as they face the same forces that made the PC business so tremendously beneficial for Americans. If cable and satellite and independent wireless firms press the Bells to innovate, then whether or not the two firms as a whole are integrated they probably will have to respond to competition at the horizontal layers where the attackers hit them. If cable has content offers to customers that attract subscribers, then the Bells will need to generate new and interesting content. If a wireless firm comes into the market with an advanced mobile broadband offer, then at that layer the Bells will need to respond.

Generativity follows competition, I suspect. Market structure rules: it sets the industry conduct and ultimately performance, to use terms from industrial organization classes.

Tim Bresnahan of Stanford, and previously the key economist on the Justice Department’s case against Microsoft, has identified three key questions to ask about the status of competition in information sector firms. I think these are the three key questions about generativity, and I think they map well against Jonathan Zittrain’s fascinating essay. First, is there divided technical leadership? (There is, where hardware and software in the PC are not made by the same firm.) Second, is there 10x change, or epochal change, such that everyone is hurrying to keep up? (There is in mobile broadband, not so much in fixed line telephony.) Third, is adjacent market entry possible? (Cable going after the Bells exemplifies this point.)

But generativity is not an easy way of life for a big incumbent, with a dividend to pay out and earnings to hit if careers are to progress. Collaborative creativity is all very fine for the upstarts and attackers of the business world, but if it means revenue sharing for the incumbent it takes on another cast altogether. So there are many reasons why an incumbent might not embrace the openness and generative approach of the PC value chain.

With the motive to pull up the drawbridge against marauding collaborators, an incumbent can well look to the threats of privacy invasion and spam attacks and to the demands of government to monitor and survey and find in these themes many reasons to oppose a generative, or “open” model.

Economists are not likely to be of one mind about the relative productivity gains of the two approaches – non-generative or closed and generative or open. Technologists, it has been said, tend to be either anarchists or authoritarians when it comes to their preferred form of government, and while that stereotype is unfair, still the choice of rule of law, and not some scientific or economic paradigm, probably will dictate the outcome. To my way of thinking, democracy depends in large part on the general embrace of the generative or open model – at least, our American republic’s modern version of democracy does so depend. JZ refers to people taking “an active part in building the world they want..” I believe that imprecation resembles the great poet Czelaw Milosz’s “command to participate actively in history.” He was talking about the overthrow of the Soviet Union, a more desperate matter surely than the toppling of a vertically integrated business model. But still the core choice is freedom versus convenience, as has been the case in every revolution, whether benign like the information revolution or something else, like the struggle for liberty against armed oppression. JZ does us all a grand service by explaining the details as well as illuminating the large themes that inform this choice, even as we make day by day in business and governmental decisions that are reshaping the Internet.

Reed Hundt is former FCC Chair and Principal, Charles Ross Partners, an advisory firm.

Posted 15 May 2008 § Generativity § Jonathan Zittrain § Reed Hundt ‡ Comments Off on Market Structure Rules

Learning to Love the Rules

essay by Wendy Seltzer, a response to David Weinberger.

“Rules are norms that have failed,” David Weinberger says in “Tacit Governance.” While his description rings true, the need for formalized governance is a sign of success, not failure. Our challenge is to move from norms to rules, building rules for the Internet that offer the flexibility that allowed norms-based communities to get big enough to need them.

Norms-based online governance works when a community adopts or forms around a technology. Thanks to shared goals, community members are willing to accept tacit limitations for the common good. But community can be oppressive as well as benevolent, and norms can be exclusionary rather than (or even while) productive. When AOL’s connection to the wider Internet released a flood of newbies on established newsgroups, the Usenet “community” responded with jibes of “go home” more often than by welcoming with education in the norms. Old-timers’ defensive reactions undervalued the potential contributions of those not yet inducted, discouraging many who could in fact add to discussions of news, tech, or even Net culture.

While we may look back nostalgically on rose-tinted moments of shared values, we can’t think that the early adopters should have a monopoly on technologies. Norms can leak beyond their community, as when consensus-based black-hole lists filter traffic to those who haven’t joined the consensus. New correspondents need a means to challenge those limitations. We might all learn from newcomers who use a social network “abnormally,” but only if we make room for them to do so.

So when we recognize the need for more explicit rules, we are acknowledging that our population has grown too large or our technology too important to be governed only by the norms of an insider “community.” As Weinberger suggests, the transition from informal norms to formalized rules can be painful. It’s hard to craft good rules, hard to set out in black-and-white the nuances that have evolved unstated, and harder still to abandon those that don’t scale.

Too often, we on the Internet think of governance as someone else’s problem. We cede the rulemaking to others while we focus on the more interesting problems of technology, or culture, or community. If we don’t develop rules from within, building on the norms, we’re liable to get worse imposed upon us. When we leave governance to the governance wonks, we get systems like ICANN, an organization that gets so far into the procedural woods and interest group politicking that it loses sight of the values of the user community it was built to serve.

If we want to do governance right, then, we must do it ourselves. Sure, that requires defining the “we” in the social contract, but that’s less critical than setting up a governance framework that’s open to outsiders. The community is catalyst but not boundary membrane.

Formal must not mean rigid, since there’s often a fine line between disruptive innovation and mere disruption. Rules-based governance need not displace norms or established communities. Instead, the rules are infrastructure upon which multiple communities can flourish with their own stricter norm-sets. A constitution is a minimal structure of rules built for extension.

We may not like rules, but we need them, and if we don’t do them ourselves, others will impose them upon us. We should learn from both the governance and the technical expertise: Plan for a revolution every generation, but build so we don’t need one.

Wendy is a Fellow at the Berkman Center for Internet and Society at Harvard Law School and a visiting assistant professor at Northeastern University School of Law, where she teaches IP, cyberlaw, and antitrust. She blogs at http://wendy.seltzer.org/blog/

Posted 14 May 2008 § David Johnson § David Weinberger § Esther Dyson § JP Rangaswami § Kevin Werbach § Pierre de Vries § Tacit vs. Explicit Governance § Wendy Seltzer ‡ Comments Off on Learning to Love the Rules

Internet Forestry: A Principles Approach to Governance

essay by Pierre de Vries, a response to David Weinberger.

Are forests governed tacitly or explicitly? In the woodland near my house, only the red alder knows how to make its leaves, and the red huckleberry and skunk cabbage pick their ideal spots without having to be told. The shrews and raccoons decide what to eat and when to breed. All the work of being a forest is done by its denizens, from the nematodes on up. And yet humans have had a hand here: broadening deer tracks into footpaths, clear-cutting old growth a hundred years ago, and deciding what to clear and what to keep after the 2002 windstorm.

There is a plethora of governance here, in the homeostasis of every cell, the competition between resident and invading species, the imposition of human needs and desires on the landscape, and the long cycles of regional climate. Most control decisions are invisible to us; those the human managers make are sometimes intuitive, and sometimes painstakingly public.

It helps to think about Internet governance in terms of forestry: both involve the management of autonomous, mercurial systems. Citizens, entrepreneurs and technologists drive innovation on the web, while policy makers try to keep up and shape the outcomes to meet broader social goals – just as the productive work in a forest is done by the plants and animals, with land managers trying to reconcile natural dynamics with social imperatives. Internet/web regulators are like forest rangers: although they have a crucial role and great responsibilities, their power and knowledge is limited. Web infrastructure comes in many sizes, from home networks to the global Internet – just as ecosystems range from window-sill planters to transnational parks. Borders are blurry; local decisions can have global impact. Pakistan Telecom’s attempt to filter local access disrupted YouTube traffic worldwide – just as a backyard gardener introducing an invasive shrub can unintentionally disrupt a regional ecosystem.

Traditional telecom and media, by contrast, are more like commercial farming. There are well-defined fields and crops. It’s clear who is responsible for what, and success is clearly defined. Today’s Internet/web is more like a chaotic patchwork of gardens, forests and national park systems. Control and creativity are decentralized. There are many competing uses and users, and endless arguments about problem definition and success criteria.

How does one govern such systems? By using unchanging principles that remain applicable in the face of unpredictable change. Detailed rules don’t help much, because they are constantly being overtaken by events. Both the internet and forests are complex, adaptive, managed systems. Since the underlying dynamics are the same, insights from one domain can teach us about the other. We can extract principles from the regularities that all adaptive systems share, and learn how to apply them to the internet from related fields like ecosystem management.

All these systems have tangled, nested structures; it’s complicated all the way down. Delegation of authority and trust to those closest to the action is unavoidable. Top-level managers have little knowledge of, and influence over, the crucial details. Further, adaptive systems produce endless surprises, and it is hard (often impossible) to trace effect back to cause. We cannot really explain why we are in our current situation, and we cannot predict the consequences of our actions. Policy flexibility, leavened with a good dose of humility, is the only way to cope with such change. Finally, complexity brings fragility. Resilience can be improved by ensuring diversity of participants, big and small (although the occasional collapse, just like a major forest fire, is necessary for renewal and reinvention).

Such principles can guide policy-making. Applied to the regulation of video services, for example, the delegation principle suggests that regulators look to web participants to generate content in local languages before imposing national production quotas. The flexibility principle implies that they should give content providers a chance to invent web-appropriate solutions to public-interest mandates, such as accessibility for the disabled, and only regulate to the extent that the community does not solve the problem. Finally, the diversity principle suggests that a multiplicity of creators is more likely to generate sustainable social value than a small number of heavily regulated producers; market entry and competition are key.

The governance choice here is between principles and rules – that is, between general guidelines for action and detailed if-then instructions. Principles are not tacit: they are an expression of our shared insights about managing a resource. However, principles are silent about how they should be implemented in a specific situation. This delegates power to a local decision-maker to respond appropriately to conditions that a high-level policy maker could not foresee (or, indeed, cannot see). General principles can become more specific when they guide short-term actions. For example, someone setting high-level policy would implement the flexibility principle by focusing on broadly defined ends and not specifying means, while a regulator would devise specific, but still technology- and business-model-neutral, rules.

Why principles and not rules? Regulators have no choice but to make it up as they go along when coping with rapidly changing systems. Principles guide decisions without relying on rules that will be obsolete before their ink has dried. We should not expect perfection from principles – just better results than could be achieved with either fleetingly relevant rules, or giving regulators complete discretion.

Governance via principles applies not only to the internet, but to everything that it transforms. The Internet/web, in contrast to traditional media, is characterized by modularity rather than vertical integration; self-organization rather than rigid structures; and unpredictable metamorphosis rather than gradual evolution. It has transformed predictable, coordinated industries into highly fragmented, unpredictable ecosystems. A principles approach is therefore needed not only for the Internet/web, but for all parts of society that it transforms. The creative wildness of the Internet is spreading to many previously staid fields of policy making; soon the forest management metaphor will be helpful beyond the narrow confines of the web.

Pierre de Vries is a Research Fellow at the Economic Policy Research Center of the University of Washington, and a Senior Adjunct Fellow of the Silicon Flatirons Center at the University of Colorado, Boulder. He is also a technology advisor to the Washington DC law firm of Harris Wiltshire & Grannis.

Note: This post is based on a recently published paper, Internet Governance as Forestry: Deriving Policy Principles from Managed Complex Adaptive Systems, by Pierre de Vries.

Posted 14 May 2008 § David Johnson § David Weinberger § Esther Dyson § JP Rangaswami § Kevin Werbach § Pierre de Vries § Tacit vs. Explicit Governance § Wendy Seltzer ‡ Comments Off on Internet Forestry: A Principles Approach to Governance

Why Tacit Governance of the Net is an Imperative

essay by JP Rangaswami, a response to David Weinberger and David Johnson.

Human beings are complex adaptive systems. We’re surrounded by such systems, in nature and in society: our immune systems, our bodies, the natural ecosystems around us, the very society we live in. The Net is no different, both in its complexity as well as in its adaptive nature.

Putting forward the case for tacit governance of the Net, David Weinberger argues that explicit governance is often a response to systematic breakdown; that explicit rules are tacit norms that have failed; and that norms derive strength from fuzziness. He also raises three key questions. How do we mediate conversations about governance? Who mediates? And, since “code is constitution”, what are the advantages of a software-mediated world?

David Johnson’s rebuttal, on the other hand, squarely supports the need for explicit governance. He concentrates on the question of legitimacy; as long as there are people being governed, and people doing the governing, goals and values need articulating; through this articulation, governors obtain legitimacy. He goes further: The Net “allows people who hate each other” to interact; such “destructive interactions” need explicit governance.

Two Davids, two sides of an argument pondered by Plato and judged by Juvenal, over two millennia ago: Quis custodiet ipsos custodiet?

I think it’s time to take a leaf out of nature’s book, particularly given the preponderance of complex systems in nature.

Two decades ago I became aware of research related to the existence of explicit parasite-pest pairings for a given plant. The hypothesis was simple: each plant was paired with a unique parasite and a unique pest; as humans migrated and took their favorite plants with them, the pests and parasites were often left behind. Some plants “migrated” with pest but without parasite; others with parasite but not pest; a few with both; yet others with neither. Some plants “took” in their new habitats, others didn’t.

This Blakeian “fearful symmetry” intrigued me: plant, pest and parasite living happily after.

The plant thrived because the parasite protected it from the pest, while the pest ensured that the parasite did not smother the plant. Plant, pest, parasite, working in close harmony according to tacit rules. The result? Symbiosis. Sustainable symbiosis. It occurred to me then that the plant-pest-parasite triple may well form a sensible governance mechanism for other complex adaptive systems, given its success in nature.

I had occasion to revisit these thoughts recently, triggered by an article headlined “Airlines and their regulator too collaborative, says watchdog”. That brought me back to the plant-pest-parasite model, but using less emotionally-charged terms: industry participant, regulator, watchdog.

Explicit governance sometimes fails because of the explicitness. People are tempted to take letter-of-law approaches, doing away with the fuzzy norms. Conversely, tacit governance processes can have an out-of-frying-pan-and-into-fire effect. Governors must behave differently, working very closely with the governed. Sometimes the governors “go native” and become ineffective. Two extremes: head-in-the-sand or mothery-smothery. In both cases the result’s the same, an unsustainable market environment.

Complex adaptive systems need complex adaptive forms of governance. Maybe we need to work on a completely different model of governance of the Net, with “parasite” regulators and “pest” watchdogs. The “parasite” regulator would inhabit the Net and have an obligate relationship with it, encouraging a tacit form of governance, allowing for fuzzy norms, smell tests, sanity checks, gut feels and hunches. In tandem, the external “pest” watchdog would ensure that the regulator stayed honest and objective, in check and accountable.

We may already have nascent examples of this outside the natural world. There may be something very similar operating in the opensource world. Every opensource community has something that purports to be the core, the moderator, the 1000lb gorilla. That moderator “governs” the codebase. However, moderators are themselves moderated by the community, which can withhold contribution at will. The moderator role is often not an elected one but one based on contribution and participation; there is no elected term; a tacit role. With tacit principles virally propagated, similar to You Own Your Own Words or YOYOW in the days of the Well.

There are many people looking at different forms of governance, many we can learn from:

Michael Power and his work on “The Risk Management of Everything”, looking at the problems caused by explicit regulation: the tendency to focus on second-order reputational risks, the tendency to concentrate on the small print and minutiae while discarding valuable yet vulnerable professional judgment.

Cass Sunstein and his work on the issue of “substitute risks“, the “hazards that materialize, or are increased, as a result of regulation”. For “regulation” we can read “explicit governance”. There is also something attractive about the argument that those who take responsibility for the avoidance of catastrophic outcomes, and take actions based on that responsibility, should be held accountable for the consequences of those actions.

Stafford Beer and his work on the Virtual System Model decades ago.

Tacit governance is an evolutionary process, with “natural selection”. Things that work get strengthened, things that don’t get jettisoned. When it comes under attack, the system adapts. Swarm behaviors are supported, with probabilistic, rather than deterministic, approaches to governance, soft-handed, even-handed. Built around weak interactions between participants. Agile, adaptive, responsive to external stimuli.

The Net is unique. Weinberger touches upon at least one unique aspect, mediation by software. Johnson touches upon another, the capacity for continued destructive interaction. There is much we have to learn about the Net. But we’re not going to learn about it if we place explicit governance models in the way. Because that’s what they’ll be. In the way. You only have to look at what happened during and after the “Kathy Sierra” incident early last year; regardless of the specifics of the incident, we have to learn from the communal response. An explicit Blogger’s Code of Conduct was put forward, and was about as successful as plumbic parachutes.

We’re going to have to learn more about identity in the Net; about intellectual property rights in a Net-influenced world; about the internet itself, about net neutrality, multi-tier speeds, traffic shaping. About how to keep paths unpolluted. None of this will be possible if we impose explicit governance models. They just won’t be adopted.

The Net needs tacit governance. Tacit governance with a difference, potentially based on our learnings from the world of biology rather than just physics. Yielding sustainable symbiosis.

JP Rangaswami is Managing Director of BT Design for BT Group. Prior to joining BT in 2006, JP was Global CIO at Dresdner Kleinwort Wasserstein. He is a Fellow of both the Royal Society of Arts and the British Computer Society. He is the author of Confused of Calcutta.

Posted 14 May 2008 § David Johnson § David Weinberger § Esther Dyson § JP Rangaswami § Kevin Werbach § Pierre de Vries § Tacit vs. Explicit Governance § Wendy Seltzer ‡ Comments (5)

The Virtues of Being Explicit – A Reply to “Tacit Governance”

essay by David Johnson, a response to David Weinberger.

David Weinberger’s essay suggests that we are better off, at least in the realm of “governance” of online activities, when we don’t (or don’t have to) write things down. His point makes sense with regard to certain types of online groups, engaged primarily in discussion rather than collaboration, formed via voluntary participation by those who already share (or can easily adopt) common values, making collective decisions primarily about their own internal affairs. But it doesn’t provide much guidance regarding how to deal with conflicts among online groups that have different values and goals and that interact with and affect each other, or with groups that want to take effective action together in both the online setting and the real world. In those more complex settings, explicit governance is often not a “scar”, as he would have it, but rather the embodiment of our highest ideals about how to live together in a diverse world.

One must begin by asking “governance of whom? By whom?” The core question is always legitimacy. The core setting is always one in which one person doesn’t want to behave in the way preferred by others. And, as Weinberger himself hints, gradual progress towards *self*-governance represents the history of political and social enlightenment. But the libertarian instinct to avoid all rules founders on the fact that we engage in socially relevant acts in groups. The very creation of a strongly mutually-imagined group necessarily creates a set of goals and values (the goals and values of the group itself) that differ to some degree from those of every individual participant. We compromise in order to collaborate. Indeed, we can hardly be said to have goals and values of our own independent of those we adopt (or live with) in the context of relationships with others, in social groups. The problem of “tacit governance” thus involves the question when and how the values and goals of the group itself should be articulated. In many contexts, there is great value in being explicit.

Clear articulation of some particular set of values may be the only way to trigger our collective imagination, to galvanize productive cooperation towards a shared goal. It was not a sign of failure of some norm-based regime of “gentle negotiation” when the authors of the Declaration of the Rights of Man and Citizen tried to write down clearly what they hoped would become true about the relationship among citizen and state. Words have power. Clarity usefully constrains. It is only by being explicit about what the social group aspires to that we empower ourselves to play the roles necessary to make it so. As in physics, social “work” only occurs when force is applied in a particular direction, subject to constraint.

If the online world consisted mainly of groups engaged in discussion, minding their own business, Weinberger would have a persuasive point. But the Internet is far more powerful than that. It allows people who hate each other, or who want to impose various forms of harm on each other, to interact. To govern such destructive interactions, we need some combination of decentralized filtering (based on reliable forms of pseudonymous identification). We need to enable groups to avoid receiving packets from those who have a bad reputation or who have not been properly introduced — a net in which every actor is accountable in this sense to all others. There is no such thing as a tacit filter – your packets are either on the list to get through (or be blocked) or they are not. There is no virtue in vagueness when the goal is to empower online groups, acting together to serve their own goals and values, to decide who should be able to join and who should be excluded.

Moreover, we need to be explicit about the principles under which any given group (including local sovereigns) should proceed when deciding whether to defer to the internal decision-making (governance, if you will) of other groups. Should the practices and customs of a virtual world be treated with respect or second-guessed based on laws made by non-participants? Even if the customs of the online group are tacit, the principles that should be used to decide whether to defer to that form of self-governance must be made explicit if we are to develop a globally valid way of enabling diverse social organizations to operate. (I would suggest the principle of “congruence” – insisting on a substantial overlap between the group whose welfare is considered when establishing a rule or custom and the potentially distinct group whose welfare is affected by the resulting practices and rules.) We have to explicitly articulate such a principle in order to abide by it.

There is a key difference between online conversation and the use of online groups to coordinate group action. Maybe we don’t need many rules about who can say what next, especially if everyone is free to tune out. But we sometimes form online groups in order to accomplish real world goals that range from changing laws to making money. These groups are powerful in part because individual participants can play complementary roles. They are most effective when people can promise to play a particular role and be held accountable by the group for having done so. A social role involves a promise made with reference to an explicit set of group goals. We are best at making promises, and sticking to them, when they are explicit. The difference between a promise and a general expectation, perhaps differently interpreted, is the difference between success and failure in many social contexts. I’m sure Weinberger doesn’t favor tacit marriage, tacit employment, or tacit fiduciary duty on the part of his financial institution. When what we do in social groups matters enough, the costs of being explicit about our promises to one another are very low compared to the value created by the resulting increase in trust.

Groups work most effectively together when every member of the group can see themselves in relation to the others, when the group can see itself. We are just beginning to develop effective ways to use graphics to create such a social mirror. But we know that the most powerful approaches will involve the tendency of graphics to disambiguate. We will be able to indicate that “my argument responds specifically to *this* point because I have placed it *there* on the screen”. Or we will become more motivated to contribute to a group project as we see precisely how many other people are pulling their weight or how close we are to achieving our (explicitly stated) group goals.

Governance of various types has been “co-extensive with the rise of civilization” precisely because it is what allows us to work together in groups, to accomplish goals none of us could aspire to accomplish on our own, to make investments of time and attention and money that would be hopelessly risky unless some explicit and reliable forms of control of others reduced the risk of defection. We must articulate our highest aspirations in order to achieve them. The default condition of society is not a cozy (tacitly norm based) community. It is instead a world of less-connected (more socially isolated) individuals, a world with fewer effective social organisms, a world with less diversity and individually empowering choice about where to invest our scarce attention and effort. The explicit articulation of group goals, of participant promises to one another, and of principles on the basis of which diverse groups can coexist, are all necessary to create an increasingly complex and interesting world.

David Johnson is currently serving as Visiting Professor at New York Law School. He also devotes substantial time to the development of new types of “graphical groupware” software products. His previous legal practice focused primarily on the emerging area of electronic commerce, including counseling on issues relating to privacy, domain names and Internet governance issues, jurisdiction, copyright, taxation, electronic contracting, encryption, defamation, privacy, ISP and OSP liability, and intellectual property. He helped to write the Electronic Communications Privacy Act,
was involved in discussions leading to the Framework for Global Electronic Commerce , and has been active in the introduction of personal computers in law practice. Examples of some of his recent work can be seen at http://dotank.nyls.edu.

Posted 14 May 2008 § David Johnson § David Weinberger § Esther Dyson § JP Rangaswami § Kevin Werbach § Pierre de Vries § Tacit vs. Explicit Governance § Wendy Seltzer ‡ Comments Off on The Virtues of Being Explicit – A Reply to “Tacit Governance”

Steering to the Edge of Trust

essay by Kevin Werbach, a response to David Weinberger.

David Weinberger’s essay demonstrates that what we do know, but don’t formalize, can actually help us. His analysis is characteristically deep and provocative. However, the parts he leaves out demand a closer look. The real action lies at the interface between tacit and explicit governance mechanisms for cyberspace.

That hidden norms can organize online behavior should not surprise us. As David points out, “governance” derives from the Greek kubernetes, or steering. The prefix “cyber” shares the same root. Norbert Wiener and his colleagues in the mid-20th century coined the term “cybernetics” to describe the implicit control mechanisms — the tacit governance — that produce order without external direction. In other words, the rule of norms is inherent in the very definition of cyberspace.

Yet there is more to the story than that. Norms and rules need each other. That is, after all, why the Founding Fathers of America produced the Federalist Papers. Their aim was to show how the norm of liberty and the explicit rules of the Constitution would serve each other.

The same is true in cyberspace. Online interactions are necessarily mediated by digital communications systems. David emphasizes the role of software, which is extensible in ways no prior media allowed. Yet even software, plastic in its functionality, must at times be rigid in its interfaces. Otherwise, the small pieces that make up the Internet would never be joined, however loosely. Anyone can edit the open source code of the Firefox browser, but if the resulting application can’t interpret the very explicit rules of HTML markup, users won’t see the Web.

Moreover, a great deal of cyberspace involves not software, but infrastructure. And that infrastructure is chock full of rules. The days when researchers built the Internet as a communal activity are long gone. Backbone networks and domain name registries link together through written contracts: explicit governance through and through. The same goes for technical standards. As norm-driven as the Internet Engineering Task Force may be, its function is to systematically surface and formalize those norms into rules. There are right and wrong ways to write the technical documents known as RFCs, to populate the header fields of an Internet protocol data packet, and to advertise routes across interconnected networks. And those infrastructure rules matter. Bad route advertisements recently took YouTube off the Internet worldwide, after Pakistan attempted to block it locally.

The Internet’s rules rule because the old norm-based approaches no longer serve the huge and diverse Internet community. Gone are the days when @Home, a cable broadband service with no subscribers but good relationships, was allocated as much address space as all of China. Moreover, the norms of those building networks and those using them are increasingly mis-aligned. Many of the staunchest defenders of tacit governance in cyberspace are the loudest advocates of explicit non-discrimination rules governing how network operators deliver their packets. Network neutrality is, at bottom, an effort to surface a norm and make it into a rule. It may well be a worthwhile endeavor. Either way, it shows the danger of an uncritical preference for tacit over explicit governance.

Where, then, is the dividing line between norms and rules? It is the edge of trust. If I can trust my compatriots, I need no formal rules to rein them in. Trust is strongest when dealing with those we know or share values with. Small groups, like the engineers who built the original Internet or the inner circle of top-tier Wikipedia editors, are more likely to trust each other. The larger the sphere of governance, the more likely interests will diverge, and the less likely that norms will suffice.

Yet bigger is not always worse. Larger and more heterogeneous communities are the best hedge against the tyranny of the majority. As James Madison, an earlier Publius, explained in the Federalist No. 10, those who represent the interests of many are less likely to be emissaries of the special interests of a few.

The cyber-solution to this governance dilemma is to fight the constraint that produces all the tensions: scarcity. Abundance trumps governance. There is no need to worry about resource allocation when there are more than enough resources to go around. And those who find their norms ill-served can choose a more suitable environment, because the costs of forming new groups and institutions are so low.

The good news is that cyberspace – if we let it – can be the greatest engine of abundance the world has ever known. From the billions of search clicks that Google pairs with targeted text ads to the millions of WiFi devices using shared wireless spectrum to the hundreds of thousands of books along Amazon.com’s long tail, abundance is the driving force of the Internet economy. It should be an abiding goal of Internet governance as well. Furthering the historical analogy, it was territorial expansion, to the Western edge of the continent and beyond, that channeled and checked the tensions of the nascent American constitutional republic.

If cyberspace is to be well-governed, therefore, it must grow. We must resist the temptation to look back nostalgically to the frontier homesteading days, when norms dominated because so many of them were shared. Let us, as David urges, embrace the Internet’s wondrous chaos. At the same time, though, let us sing the praises of its well-designed rules. The shared enemy is not structure, but exclusivity and other barriers to choice and connectivity.

Kevin Werbach is a leading expert on the business, policy, and social implications of emerging Internet and communications technologies. Werbach is an Assistant Professor of Legal Studies and Business Ethics at The Wharton School, University of Pennsylvania and the organizer of the annual Supernova technology conference. He is also the author of werblog.

Posted 13 May 2008 § David Johnson § David Weinberger § Esther Dyson § Kevin Werbach § Pierre de Vries § Tacit vs. Explicit Governance ‡ Comments Off on Steering to the Edge of Trust

Governance – Tacit or Explicit?

essay by Esther Dyson, a response to David Weinberger.

The whole point of the net is that it is decentralized and heterogeneous: One size need not fit all. Thus there is no need to resolve the question of whether tacit or explicit rules are better for online communities. But you can ask when to use each…and how do they interact? You could ask those same questions about offline communities and get some of the same answers, but there are two important differences online: Online communities can have much greater scale and reach. And they are much easier to join or leave (and rejoin under a new identity), so they have a much higher proportion of strangers.

So, when should we use tacit rules? As David Weinberger points out, they are wonderful. People share common goals, they police themselves, and everything works. And indeed it does, in small groups of people who know one another and share values. (I define “community” as a group in which a member who leaves is missed.)

Indeed, says Weinberger, “the rise of explicit rules is a sign of failure.”

Perhaps so, but we learn from failure. Those explicit rules are a result of learning…and they make learning easier. Without them, we couldn’t easily share that learning with people who had not lived through the failure of tacit governance.

That’s why over time, offline, we have developed rules for dealing with “foreign” communities, with outsiders and invaders. The development of explicit rules surely helped us to deal with foreigners— not just as invaders or invadees, but as equal but different communities who could have their own, possibly tacit rules internally, but would observe certain more general, explicit rules in dealing with outsiders. Those explicit rules make it easier to operate effectively with strangers – so that they do not make mistakes that might cause friction or worse between communities. You can’t expect a stranger to follow tacit rules.

Let’s look at what happens when we attempt to scale up with tacit rules, even for insiders. In thinking about this, I have the advantage of having spent the weekend at a seminar with a group of Russians. In Russia, there’s a proliferation of laws, but the overall system of governance is mostly tacit in practice. (That’s not to say that there is not a lot of excruciatingly explicit paperwork, but most of it is irrelevant.) This tacit system – of connections, unspoken rules, shadowy powers – leads to all kinds of maladies. Those in power can act as they like almost with impunity. Those without power but with an understanding of the rules can mostly stay out of trouble.

But those who don’t understand the rules, or who question them, can lose their freedom or even their lives. (As Russian politician Boris Nemtsov once pointed out [in paraphrase], “Yes, there is freedom of speech. But that does not necessarily mean freedom after speech.”)

Tacit rules are inherently hostile to outsiders and to trouble-makers – and ineffective with them as well. Weinberger’s tacit rules are benign, but most tacit rules are not so benign. And if they were explicit, they could be more easily condemned and repudiated.

Even if the rules are good, their implicitness makes them harder for newcomers to understand. Yes, newcomers can observe and learn, but the burden that imposes should be acknowledged. And finally, tacit rules are harder to spread, since they can’t be easily transferred to other communities. That’s unfortunate, because ideally communities can learn from one another, either by copying one another’s rules, or by having members who bring effective rules with them.

Communities: Intelligent design or evolution?

Add another factor: Online communities are easy to enter and leave, which means that communities with bad rules – whether tacit or explicit – can be abandoned without much harm. There are no burned-out neighborhoods left behind, though there may be bitter memories in the hearts of some members. This allows for competition among rule sets, leading to the survival of good communities and the destruction of “bad” ones, by whatever measure.

But there’s another, possibly better form of evolution, rather than survival of the fittest communities: That’s learning and change within a community. Tacit rules may be effective, but they aren’t that easy to change.

In fact, we need rules, but we also need an explicit grammar by which they can be changed. Explicit rules, by their very explicitness, can be expressed and can be changed.

That is, what kind of group consent is required for the change? Can they be changed to benefit only one or two parties? Can they be retroactive? If someone lived by the rules and now they have changed, what is owed to those who were bound by the old rules? (This too is visible in Russian society: What does the state owe to a loyal factory manager who worked hard but never got equity? When the factory was privatized, did the factory manager have no rights at all versus a foreign buyer wading in with wads of cash? Or, more likely, did the tacit ability of a legislator’s cousin to know when the property was to be put up for auction?)

Scaling laws

Finally, let’s consider scale. In a large community, participants are not likely to know one another. If the stakes are small – if the site is devoted to discussion or content rather than transactions (including the “transaction” of disclosing someone’s secrets – that doesn’t matter much. But if people are putting money or other valuables, including their reputations, at risk, then they need some kind of accountability and reputation system. That again needs to end up fairly explicit. What are the rules of membership and of disbarment? To what extent do those rules depend on external rules – such as the requirement of entering a credit card number as a light credential? Like it or not, the danger of some individuals compromising the system are just too great. Just ask eBay, which spends a dismaying proportion of its resources fighting fraud.

Of course, it’s possible for a small, tacit community to come up with ratings or other badges that its users can wear in other, looser communities. And those “certifying” communities themselves can have reputations that let members of other communities know whether to trust them. With luck, everyone can just get along without consulting the rules, but the knowledge that they are there provides protection for everyone.

Esther Dyson is a long-time catalyst of start-ups and new ideas, primarily in information technology, but also in health care and in private aviation and space—all markets disrupted by privatization, decentralization, and the impact of IT.

Posted 13 May 2008 § Charlie Leadbeater § David Johnson § David Weinberger § Esther Dyson § JP Rangaswami § Kevin Werbach § Pierre de Vries § Tacit vs. Explicit Governance § Wendy Seltzer ‡ Comments (2)

Tacit Governance

essay by David Weinberger, responses by Esther Dyson, Kevin Werbach, David Johnson, Wendy Seltzer, JP Rangaswami, and Pierre de Vries.

Since governance is, like speaking, co-extensive with the rise of civilization, it’s curious that it has such a bad name. Or perhaps it’s not so curious. Governance, as an explicit social structure, codified and implemented, arises when tacit governance fails. At its best, explicit governance is a response to a breakdown. It rarely restores a society to its prior, unbroken state.

Governance is made explicit as a scar. Scars are useful. They can even be honorable. But they generally mark wounds. The lack of explicit constitutions and explicit rules often is a sign of health.

The vastest stretches of the Internet’s surface are as yet unblemished by explicit governance. Tacit governance, however, is the surface of the Net. The Net is most of all a new social space in which people gather in groupings familiar and odd. All human intercourse has some form of governance, for otherwise the participants have no way to talk. Conversing (in its broadest sense) requires not only a common language, but also some set of core expectations about the boundaries of the conversation. Those expectations steer the conversation; “governance” comes from the Greek for steering.

The expectations that steer human intercourse are rarely laid out, in the real world or on line. Since our interactions always occur within some context, we assume the norms of that context: In Boston, we’d be fine with our cab driver spontaneously expressing support for the Red Sox, but we would be surprised if the cab driver pulled over, unasked, to show us how well she plays the tuba, no matter how well she plays it. There is no explicit rule about this because there doesn’t need to be. If, however, cab drivers start regularly giving curbside tuba performances, a “You are entitled to a tuba-free cab” rule will be posted. Rules can also be useful guides to norms when we can anticipate cultural strangers may be coming along for the ride.

The fuzziness of norms is their strength. We need the looseness of norms to enable us to be with one another in surprising ways. The narrower, more explicit, and less ambiguous the norms, often the deader the social interaction: “Come now, Marjorie, you know that we raise our hands before speaking.” Norms are not rules that have yet to mature. Rules are norms that have failed.

Governance and rules often arise when norms not only fail to cover important cases that have arisen, but when – as is typical – they don’t provide ways to resolve conflicts about themselves. For example, informal discussions, on and off the Net, sometimes find themselves talking about the discussion itself. Person One says: “Can we please stay on the topic?” Person Two says: “I didn’t realize we had an agenda.” In the real world, Person Two may just peel off and talk to Person Three a few feet away. On line, where the discussion itself may have some ontological claim – it’s got a name, a place, a history, maybe a membership – People One, Two, Three, all the way to Person n, may try to work out the norm for the sake of the discussion. Perhaps a rule will emerge — “Kindly keep your comments on topic” – and perhaps the group will fork or die. Once enough rules have been hammered out, a FAQ may emerge, or an introductory message permanently pinned to the top of the discussion board. Some rules may call a ruling class into existence: “Maria is the Flame Mistress. Use harsh words and you will answer to Maria.” The rules may get codified not only in the sense that they’re written down but that they develop their own cryptic expression – for example, Godwin’s Law, or the outburst “WP: BITE” at Wikipedia that reminds experienced Wikipedia’s not to bite the newbie’s—a shibboleth by which the old hands recognize one another. Knowledge of the rules can itself become part of the fabric of the group.

This is one reason why so few groups on or off the Net begin with an explicit constitution. The task of creating explicit agreement is so onerous and so fraught with unexpressed expectations that groups usually need the cohesion – the implicit connections – provided by living together via norms to survive the constitutional moment.

Explicit governance that arises in response to a failure of norms is inevitable and organic, although at times it can be disruptive and acutely painful. Governance that is imposed from without is almost always mainly harmful. It thinks clarity and precision are virtues,
when in fact they drive out the gentle negotiation by which problems are solved and, more important, by which groups become more than merely well-regulated collections of individuals. This is true online and off.

But there is something special about tacit governance online: It is mediated by software, and software comes with some abilities and not others. Code is constitution. Of course that’s true of real world media as well, but the media that we’re used to in the real world have been so limited that their implicit governance has felt more like limitations than possibilities. How can you talk about the affordances of a telephone system for social interaction without beginning with its overwhelming limitation: You can only talk, it’s really designed for talking to just one person at a time, you generally reach someone by interrupting her. Eventually you may get around to considering sending faxes and navigating phone trees by pressing numbered buttons, but the essence of the telephone is expressed by its overweening lack of ambition. Net applications, on the other hand, tend to be rich in possibilities. And even when they are not, we inhabit them with our own inventions. For example, if you were to write a user manual for Flickr, it’d have more sections than any normal person would want to read. And, even so, that’s not enough for us. Flickr lets us annotate photos by drawing boxes on them with notes attached. If you come upon a photo with concentric boxes drawn on it, there’s a very good chance you’ve come upon people who are arguing about some feature of the photo, using Flickr’s affordances in a way Flickr never anticipated. Flickr’s affordances are a type of tacit governance. So are the nested boxes effortlessly invented by its users. But it will not seem like governance until someone nests boxes in an “inappropriate” way, and someone else draws a box around them all and says, “Dude, stop inserting your spammy boxes in the middle of our conversation.” The moment at which Flickr has to post rules for using nested boxes is the day that the nested box norm has failed.

Every new rule is a scab covering tender flesh. Or so we should hope.

When governing bodies look at the Internet, they see its unruliness. But most of the swirl of the Net is in fact governed by rules so deeply implicit that even to surface them would disrupt the creative social work underway. The overwhelming preponderance on the Net of tacit governance over explicit is a sign of the Net’s depth, importance, humanity, health and success.

David Weinberger is the author of Joho the Blog and a Fellow at the Berkman Center for Internet and Society at Harvard Law School.

Posted 12 May 2008 § Charlie Leadbeater § David Johnson § David Weinberger § Esther Dyson § JP Rangaswami § Kevin Werbach § Pierre de Vries § Tacit vs. Explicit Governance § Wendy Seltzer ‡ Comments (4)