The Genius of Pictures
Yesterday's COPA decision represents a milestone in the era of data visualization: citation to histograms created by a politically diverse group looking together at a screen. And those histograms turn out to have been critically important to the majority opinion in the case.
David Johnson, who wrote the program that encouraged the group to come up with those histograms, should be feeling very proud today. (He's now an esteemed professor of law at New York Law School.)
Here's the story: One part of COPA that wasn't declared unconstitutional created an (unfunded) Congressional commission to study technologies and methods designed to reduce access by minors to “harmful to minors” material on the internet. The Commission's report is here.
Congress directed the Commission to evaluate the accessibility, cost, and effectiveness of protective technologies and methods, as well as their possible effects on privacy, First Amendment values and law enforcement.
The COPA Commission worked hard. It met in person several times (all without funding). It held regional meetings. It examined a wide range of child-protective technologies and methods, including filtering and blocking services; labeling and rating systems; age verification efforts; the possibility of a new top-level domain for harmful to minors material; “greenspaces” containing only child-appropriate materials; internet monitoring and time-limiting technologies; acceptable use policies and family contracts; online resources providing access to protective technologies and methods; and options for increased prosecution against illegal online material.
It was hard work even coming up with this list of technologies and methods to examine. And then an even tougher task was confronted by the Commission: how to “evaluate” all these disparate things along the axes suggested by Congress. David Johnson and I were staff to the chair of the Commission, Don Telage, for this effort, and I'd like to thank our former law firm, Wilmer, Cutler & Pickering (now WCP Hale Dorr) for supporting our work on this task.
David came up with the idea, and coded the program, that allowed the Commission to complete its work.
What David did was to build a software program that displayed a separate screen for each technology and method that was the subject of the group’s study. On that screen, he put a histogram consisting of bars that could be higher or lower based on a set of scores for effectiveness, cost, accessibility, privacy-compatibility, and impact on First Amendment rights. The screen also contained a numeric rating given to the technology by every member of the Commission. The overall state of any given bar in the histogram, on any given screen, reflected the average views of the group.
We passed out scoring sheets to the members of the group (which included people with a wide spectrum of different views – everything from prosecutors and anti-porn crusaders to civil libertarians). We had them tell us their ratings individually, without compromise or log-rolling or speeches.
At the next physical meeting, we showed the Commission the resulting pictures. Each member of the Commission saw how his/her ratings compared to those of his colleagues. This allowed us to focus on scores that were quite different from the others, and gave those scorers a chance to persuade the rest of this group that he/she was right. Sometimes this led to changes by scorers. Sometimes, the group agreed to disagree. But the most remarkable effect of this shared screen was that the group worked well together. No one talked about irrelevant subjects or tried to derail the discussion. The question was the picture on the screen.
The Court cited the results of this histogram work, saying what the ratings of the Commission were for use of filtering technologies, and citing with approval the overall work of the Commission. The answer that was reached by the group: filtering technologies are more effective, with less adverse impacts, than any of the other technologies and methods considered. And that's what the Court found.
The COPA Commission, chaired by Don Telage, was made up of the following people:
Stephen Balkam, Internet Content Rating Association
John Bastian, Security Software Systems
Jerry Berman, Center for Democracy & Technology
Arthur H. DeRosier, Jr., Rocky Mountain College
J. Robert Flores, National Law Center for Children and Families
Albert F. Ganier III, Education Networks of America
Michael E. Horowitz, Department of Justice
Donna Rice Hughes, Author, Kids Online/Founder, Protectkids.com
William M. Parker, Crosswalk.com
C. Lee Peeler, Federal Trade Commission
Gregory L. Rohde, Department of Commerce/NTIA
C. James Schmidt, San Jose State University
William L. Schrader, PSINet Inc.
Larry Shapiro, Walt Disney Internet Group
Srinija Srinivasan, Yahoo! Inc.
Karen Talbert, Nortel Networks
George Vradenburg III, America Online, Inc.
Nethead/Bellhead — Noticing DHS
I have been working on the questions I'd like the panels to debate on September 28. The first panel will deal with justifications for regulation.
Consider the following text from a May 1998 World Information Technology and Services Alliance (WITSA) paper, Statement on Convergence:
“WITSA believes there are three legitimate justifications for regulation. The first is to foster or stimulate competition. In most cases, the WITSA believes that general competition rules will suffice, and markets should be allowed to operate free from government intervention. However, when dominant suppliers exist in a given market, or when market forces - for whatever reason - are not able to ensure discipline, then some form of regulatory oversight may be required.
A second justification for regulation is to protect and enhance the public good. However, use of this condition must be made lightly and sparingly. Instances where this justification may traditionally apply may even be disappearing in the face of competition. For example, the build-out of telephony networks to provide ubiquitous access to basic services has long been considered a legitimate reason for regulation. However, this view is now being challenged with the introduction of the new technologies.
A final justification for regulation is the equitable allocation of scarce public resources. Spectrum, for example, has long been considered a scarce resource and regulation was used to insure its distribution in an orderly and reasonable manner.”
Consider also this statement from the Department of Homeland Security, filed in the IP-enabled services proceeding:
“[National Security/Emergency Preparedness] NS/EP considerations provide a compelling rationale for applying a certain amount of regulation to IP-enabled services. The purpose of such regulation would be to ensure the prioritized availability of certain communication services to Federal, state, and local officials and first responders in times of emergency or national crisis.”
What is/are the justification(s) for FCC regulation of IP-enabled services, as those services are defined in n.1 of the NPRM? How persuasive are these justifications?What are possible responses to these proposed justifications?
So — that's the first panel. I'll be sending out emails asking for links to materials that should be posted in connection with the conference. We're planning to have a must-go-to conference web site that collects these resources.
In working on these paragraphs, I've been struck by what DHS is asking for. Scott Bradner pointed me to this a week ago, but it has taken me some time to actually climb the learning curve (thanks, Scott!). I've linked to their statement above.
They say, “In times of emergency or network congestion, [National Security/Emergency Preparedness] NS/EP priority treatment may be required for certain communications such as electronic mail, instant messaging, video feeds, or video conferencing sessions. The Commission's rulemaking process must keep this in mind.” (This is on p. 8 of their filing.)
And, “In the event of crisis, NS/EP national leadership must receive end-to-end priority treatment over other users. . . . NS/EP traffic must be identified with its own class of service — above and beyond “best effort.” (p.9)
And, on p. 10, “It is the intent of DHS to take advantage of the technology developed by the industry to achieve its objectives of assured NS/EP communications during crises. The NCS intends to continue to work with the industry through voluntary and contractual arrangements, subject to Congressional budget constraints, to support NS/EP services and features [including standard priority markings of NS/EP communications]. If these voluntary and contractual arrangements are insufficient to achieve assured NS/EP IP-enabled communications services, the NCS would request the FCC consider imposing regulatory constraints on all providers of IP-enabled services. . .”
So: One of the possible reasons we need regulation of ip-enabled services is to assure priority of national security communications. And, whatever we do, we should not interfere with CALEA for ip-enabled services.
The FCC is clearly under enormous pressure.
Overstatement and IICA
There are reasonable people walking on this earth who will say that the IICA is not a big deal. Rather than jump down their throats, I'm going to suggest that we slow things down, have some hearings, and try to get to the bottom of what's going on. Three kinds of arguments, all of which interrelate, are being thrown back and forth:
1. The bill will cripple the development of new technology. Counterargument: The RIAA says that the legislation “was meant to be narrowly tailored to address companies that build technology focused on illegal file sharing.” And Sen. Hatch says the bill “protects existing legitimate technologies and future innovation in all technologies – including peer-to-peer networking,” and maintains that “all agree that non-piracy-adapted implementations of P2P could have legitimate and beneficial uses.”
Let's assume, for the sake of this argument, that both sides have good points. But there is a great deal of fear on the IT sector side, and there's no limiting language in the bill that focuses on “illegal file sharing.” It's all about copyright infringement, which can be said to occur all the time in all kinds of digital and analogue devices.
Because it's hard to go after direct infringers, it makes chokepoint sense for the content industry to go after machine and software maufacturers. (Note: under the bill, Xerox copying machines look like much more attractive targets. Those guys know they're making money from illegal copying. They constantly advertise their copying services.) The content industry has of course already done this, and they're disappointed with the district court decisions in Grokster and Napster. But they did very well at the appellate level in Napster and Aimster, and they drove ReplayTV out of business. Surely secondary liability is alive and well.
Given the apparently well-founded fears of the IT industry, and the existing, ample, judge-made doctrine on secondary liability, why create a brand-new, seemingly unlimited cause of action that any copyright owner can bring against any thing or person or business he doesn't like? It's worth having a hearing or two on this subject, at the very least.
2. The bill will broaden secondary liability for copyright infringement in ways we cannot predict. Counterargument: This is a merely incremental statute, and nothing new. “In the Criminal Code, Congress made it a federal crime to willfully infringe copyrights or to distribute obscene pornography or child pornography. Congress also made it a crime to induce anyone – child or adult – to commit any federal crime.” And, “The Inducement Act will simply import and adapt the Patent Act’s concept of “active inducement” in order to cover cases of intentional inducement that were explicitly not at issue in Sony.”
Again, there are certainly good sources for “inducement” in both the criminal law and patent law. But both are arguably inapposite sources: criminal law has a higher standard of proof and requires the exercise of a prosecutor's judgment about what's worth suing on; patent law requires specific intent to induce infringement of a specific patent — and it may be that “inducing” can't be found where the article is capable of substantial noninfringing uses. So where are we? We're puzzled and confused. Again, it's worth having hearings.
The bill will render Sony irrelevant. Counterarguments: The bill says that “Nothing in this subsection shall enlarge or diminish the doctrines of vicarious and contributory liability for copyright infringement or require any court to unjustly withhold or impose any secondary liability for copyright.” This ”savings” clause is said to save Sony, and MPAA says that “Enabling technologies have nothing to worry about as long as they are not inducing other people to violate the copyright law.”
The people on the other side of this argument can point to the emptiness of DMCA's statement that fair use wasn't affected by that statute. Yes, the doctrine of fair use isn't affected by the DMCA, but the anti-circumvention rules make fair use impossible. It's a difficult problem: “Fair use” is itself a court-created, after-the-fact balancing of many relevant factors.The fair use balancing does not occur until someone has brought a case charging infringement, infringement has been found, and the infringer has raised the affirmative defense of fair use. In the anti-circumvention arena, all of this back-and-forth is inpossible.
Similarly, so the argument goes, the IICA's creation of a new kind of secondary liability, triggered by “acts from which a reasonable person would find intent to induce infringement based upon all relevant information about such acts then reasonably available to the actor, including whether the activity relies on infringement for its commercial viability,” would leave judge-made theories of contributory liability in place — but no one would ever use them again.
Plaintiffs would have this brand new, clean-as-a-whistle cause of action to use, untainted by all that Sony history. They could show that infringement is important to the commercial viability of a particular machine (like the copy machine), and then, poof, there would be enough “intent” to satisfy the statute.
So where are we? We're worried enough to hold a hearing. We don't need to scream or claim that the other side is witless or evil. We just need to think this through.
INDUCE Act introduced
Here is the revised bill [pdf], introduced late last night. Floor statement is here [word].
The content industry would like to overrule Sony, and sees an opportunity to do so before all legislative activity ceases for the election. This is the Hollings bill in another guise. It would potentially make some legitimate technology liable for secondary copyright infringement — things you love, like the Apple iPod.
The industry will say “If you're not with us, you're against us,” and if you're against this bill you must be in favor of child pornography. The bill's proponents will claim that this is all incremental — an application of patent standards that have worked well for years. But the INDUCE Act is much more than that: it is no less than an attempt to ensure that any equipment manufacturer that makes money in an atmosphere in which some copyright infringement may be occurring will itself be liable for infringement.
It's very easy to show direct copyright infringement — much easier than it is to show patent infringement. So indirect liability will also be easier to claim under this act.
I'm not saying it's the end of western civilization. I am saying that this is bad legislation that will have a negative effect on a part of our economy — the IT sector — that contributes much more in terms of jobs and revenue to the country's economic health than the content industry does.
INDUCE act
There's some good news and some bad news. The good news is that “counsels” has been dropped from the bill. (No copy of the bill is available at the moment.)
The bad news is that there are strong rumors that the bill will be introduced today or tomorrow, that there will be a hearing about it on Thursday or Friday, and that it will go to the floor quickly on a unanimous consent request — a fast track.
For those who suggest that patent “inducement” claims haven't brought western civilization to a halt, three responses.
1. You're right.
2. It's much much harder to prove patent infringement than copyright infringement. You have to be serious to bring a patent claim. You have to hire experts. It's a big deal. Copyright infringement claims, by contrast, are a piece of cake to bring. So the risk of unpredictable consequences to innovation etc. from secondary copyright “inducement” claims are higher — because there will likely be more of them.
3. Sony depends in part on a borrowed patent analysis (staple article of commerce capable of substantial noninfringing uses), so you may be claiming that the INDUCE act is just one more borrowing. But what you'd be doing is effectively overruling Sony by this much broader borrowing — and creating liability rather than limiting it. Different.
This proposed bill is part of a concerted effort [pdf] to make sure that P2P software is viewed as illegal stuff.
Nethead/Bellhead: Progress report
Plans are marching on for Nethead/Bellhead: The FCC Takes On the Internet. Date: Tuesday, Sept. 28. This is a one-day conference, under the auspices of Cardozo's Floersheimer Center for Constitutional Democracy and co-sponsored by the Yale Information Society Project.
So far, representatives of MCI, BellSouth, Verizon, Level3, and Vonage have said they'll come. We've got several people from the FCC signed up. People from Public Knowledge, CDT, Cato, and PFF are coming. David Isenberg, Kevin Werbach, David Weinberger, Jeff Jarvis, and Dave Farber are coming. I'm still working on Google, Intel, AOL, and MSN and have high hopes. Jon Weinberg, Jim Speta, and Rob Frieden are coming.
Excited as I am about who's coming, I'm more focused on the substance. In its IP-Enabled Services rulemaking (WC Docket No. 04-36), the FCC appears to be asserting that although there isn't a compelling rationale for applying traditional “economic” regulation to providers of IP-enabled services (paying for universal service, paying interstate access charges, filing tariffs, etc.), they're considering using their ancillary jurisdiction to apply “social policy” regulations.
The rulemaking asks for comments about which social policy regulations make sense for which IP-enabled services. Here's footnote 155:
“For example, one might question what it would mean to apply E911 obligations on an Internet retailer . . . . Similarly, some obligations may only be sensible in the context of VoIP service. However, to ensure that whatever distinctions we ultimately draw among different IP-enabled services are sound as a matter of law, technology, and public policy, we decline in this Notice to foreclose any particular approach, and therefore frame our questions in terms of all “IP-enabled services,” though some may only apply to particular types of service.”
This rulemaking is not just about VoIP. It's about all IP-enabled services — including email and IM. All of these services are arguably “information services” in the FCC's lexicon, and thus subject to FCC's jurisdiction.
FCC recently found [pdf] that pulver.com's Free World Dialup (FWD) was an information service. Although some commentators had asked FCC to make clear that FWD was just a software application, not a service, FCC took a different route. As I understand it, things that use wires or radios in the US fall into one of two categories in the FCC's eyes: you're either a “telecommunications” entity or service, or you're an “information service.” Since Pulver isn't a telecommunications carrier, it has to be an “information service.”
Pulver had tried to argue that, no, he wasn't offering an “information service” — he was just offering a software product online that created a namespace. But the FCC stuck to its guns:
“[T]he statutory definition of an information service speaks only to the offering of various types of computing capabilities via telecommunications, not the offering of telecommunications itself. The fact that FWD’s computing capabilities, as described above, are available to its members via “telecommunications” – i.e., the telecommunications underlying its members’ Internet connectivity; the telecommunications connecting Pulver’s FWD server to the Internet; and the telecommunications underlying the Internet backbone itself – is sufficient to meet the statutory definition of “information service.”
With me so far? This means that any computing capabilities offered via a connection to the internet are “information services.” Email, IM and the DNS fall in this category. Although these things might not have to be tariffed, because they're not “telecommunications services,” they might have to be subject to several different kinds of social policies — including making their designs subject to FBI approval, making themselves accessible to the disabled, facilitating E911 services, and being subject to privacy rules. That's what the IP-enabled services rulemaking is about. It covers both applications and services.
The argument may be that these social policies are lightweight in comparison to treatment as a common carrier — and that protection from a huge variety of state rules is worth giving the FCC authority to promulgate these social policy regulations.
Potential panel subjects are (still in the planning stages — send advice):
1. What market failure suggests the need for “social policy” regulation of IP-enabled services? If market failure is not the basis for this suggestion (to the extent such a suggestion is being made), what is the rationale?
2. FCC's ancillary jurisdiction to adopt the broadcast flag rule is under judicial review. What are the arguments in favor of/against exercise of FCC's ancillary jurisdiction in the IP-enabled services rulemaking?
3. Should the disability accessibility requirements in Section 255 and 251(a)(2) of the Telecommunications Act be applied to all IP-enabled services?
4. What, if any, consumer protection requirements — including privacy obligations, billing rules, prohibitions of obscene or harassing communications – should be imposed on IP-enabled services?
5. What, if any, obligations to contribute to universal service funds should be borne by the providers of IP-enabled services?
So: Send advice and tips, and join us on September 28.
Internet governance
For the last two days, I've been at the Berkman Center as part of a group talking about internet governance.
John Palfrey just brought some light into the room by talking about a Net Dialogue project that Stanford and Berkman are working on together. The plan is to make the layers of net “governance” visible by using wikis and other visualization/data tools — what agencies work on what layers of the protocol stack, what issues are where. Thank goodness. John says Mary Rundle is working hard on this and will be launching the site in July.
The problem is that internet governance, as a term, is useless. Developing nations are worried about connectivity; the FBI is worried about pre-approval of all IP-enabled applications; there's a group of people wandering around the world going to WSIS meetings; and ICANN (which has very little to do with internet governance) attracts enormous criticism.
I'm frustrated with the unfocused discussions about this topic. I'd like to be building things, I'd like to be talking to people one-on-one, I'd like to be learning more about the FCC's role in the world. But developments like John and Mary's suggested web site cheer me up, and I'm willing to continue as long as people like them stay involved. The Accountable Net continues to be a very strong and useful idea, and I'll keep pitching it.
INDUCE Act
Here's something to worry about: The INDUCE Act of 2004 [pdf].
The logic is that P2P applications inevitably lead to exploitation of children. With me so far? So the act is called the “Inducement Devolves into Unlawful Child Exploitation Act.” I'm not even sure that's how “devolves” should be used. But the crimes here go far beyond the title.
The Act (to be proposed tomorrow by songwriter Sen. Hatch and others) amends the copyright law to say that anyone who “induces” copyright infringement is himself/itself an infringer.
“Induce” means intentionally aids, abets, counsels, or procures. So you can't even hire a lawyer if you're doing something risky.
This is amazing. Now we're waaaaaay beyond contributory and vicarious theories of liability, which are court-created and pretty darn broad on their own. See Napster 9th Circuit, Aimster 7th Circuit. It's not even clear what the limit to this is — “aids” could mean that even something that would have been fair use under the Sony Betamax decision is now an illegal inducement.
And no one can talk to you if they think there's the slightest risk of copyright infringement liability.
We're back to the CBPTDA – another hugely broad way of making sure that no unauthorized machines ever enter into our lives. If there was ever a moment to organize (see prior post) this might be it.
“Copyright. It's Not Just the Law. It's All Law.”
[thanks to Fred von Lohmann for pointing to this]
Reflection
So where are we? Blogs seem primitive, but they're changing. Everyone's looking for ways to follow up on whatever online energy spurred the Dean campaign on. Technorati is pulling things together for us. Some people have recognized that blogrolls are just like social software lists of frients (which results in mixed feelings). And academics are finding different communities to work with.
These developments, for all their excitement, feel insular. It seems as if we have a long way to go in recognizing what online life is and what effect it is having on the world. We seem to be so focused on what the big guy (the blogger rock star) is thinking, rather than on what interesting complexities we've created by way of the groups we're involved in. I love the blogs I read, and I go back to them every day. I'm often inspired by them. But I also often feel that I'm witnessing a strange big-media move (get everyone watching the same thing) in a local-media place. After all, the internet should make it possible for large groups to act on the same thing, not just link to it.
It's not that I'm against crowd-pleasing events. Everyone wants to be part of an order that can be understood — we like drum beats, marching bands, parades, views of ocean waves, and bass beats. Some people like to drive down the road with the radio blaring, just to feel part of some rhythm larger than their own. Synchronization is central to our lives.
But I'm worried that the groupness of all this online journal-keeping is (paradoxically) being underplayed. We watch and imitate, somewhat passively, but we don't make it possible for online groups to speak more powerfully than we can as online individuals. There's something missing, some sense of animating purpose or spirit that produces results. When there's something big to complain about, why don't like-minded local writers have a place to go to leverage their voices?
Maybe to have a trade association you need a trade. (Someone wise told me this.) Maybe without the leverage of compensation flowing in one direction or another, the many voices stay just that — multiple, clever, smooth, bantering, never ceding any shred of autonomy.
Zen and ICANN
Take another look at Zen and the Art of Motorcycle Maintenance. It's an inquiry into values, and it reminds us that the most important thing is Quality. What's Quality? Let's say it means “excellence.” Robert Pirsig's thesis is that Quality is the source of everything we know.
The book also reminds us that it's dangerous to think that rationality is the most important part of life. Thinking things through isn't necessarily the same as improving something.
This ties into the ICANN budget. It's based on a strategic plan that hasn't been made public. Its statement of constituency concerns aren't necessarily based on actual constituency statements. Whether or not ICANN is successful in promoting choice and competition will apparently be measured by how the community feels about this issue at the end of the day. It provides for an enormous effort in “compliance,” but doesn't say what that means or whether ICANN will be more receptive to new registry services. And it doesn't provide for a backup plan in case the registrars decide not to go along with the increased funding requirements imposed on them.
But that's all logical quibbling. Let's say that ICANN has answers to all of these questions and satisfies most people that it intends to do the right thing. What about Quality?
The thing about Quality is that you can tell when it's there, and you miss it when it's not. You can take Quality apart, and point to elegance, competence, coherence, and intelligence as some of its elements. But it's a holistic sort of thing. It's either there or it isn't. And it's terribly important.
When you try to explain ICANN to someone else (particularly someone who is thinking about running a new TLD) it's hard to include in your description the idea that ICANN is a quality organization. As a group, ICANN seems to be devoted to process. Yet it seems to be hard for any one of these processes to result in a finished product. (If you look at the ICANN home page, most of the listings concern processes that aren't over.) Its meetings continue to be in-person, all-over-the-world, lengthy happenings in which process is discussed – odd in a technical coordination body. It has opened up very few TLDs, and has used its de facto control over the root to mandate all kinds of things. Its organizational chart will no doubt continue to grow, but it's hard to point to particularly elegant, competent, coherent, and intelligent things it has done.
On the other hand, ICANN was formed to be a forum for discussion, and it is that. It was formed as an alternative to government control of the DNS, and it is that in a de facto sense. Maybe its quality is found in these two elements: it's better than the alternative, and it's a place to talk. If that's the case, ICANN should focus on building up these two competencies and become more of a Quality organization.
It doesn't appear to be pleasing people on the rationality front.