CPSR conference brings people and Internet together
by Andy Oram
The Internet never looked this way from Harvard Square before. The
2002 annual meeting of
Computer Professionals for Social Responsibility
this past Saturday left the 75 participants enlightened and wildly
excited about giving control over information to poor people all around
the world. I arrived home at ten o'clock at night and told my wife,
"You're lucky I didn't sign up to spend a month in Malawi installing
The title of the annual meeting was
"Shrinking World, Expanding Net",
a title that nimbly conveyed the dual (and perhaps dueling) trends
within an Internet that is quickly becoming a commodity.
On the one hand, Internet access is being extended to geographic
regions and demographic groups where recently it was considered
unfeasible. As access spreads, the new nodes take on characteristics
totally foreign to the original users in the developed world:
characteristics adapted to poor connectivity, low bandwidth, problems
with literacy, and a diversity of cultural conditions.
On the other hand, as people realize the Internet's importance,
pressures increase to impose some predictability on it, while the
pursuit of democracy and community development online gains support.
Here is a summary of the day's events, including the ceremony awarding
to networking engineer and ICANN Board member Karl Auerbach:
The workshop was expertly assembled and carried off in the belly of
the beast, Harvard University's Kennedy School of Government, by
Kennedy School professor L. Jean Camp and a dozen student
volunteers. (To their credit, the Kennedy School co-sponsored the
workshop.) If anything on this weblog makes you interested in working
with CPSR, check our
list of topics or
Elsewhere, perhaps, debate still rages. Do poor people need
advanced information technology? Can they make proper use of it? Is it
possible to deploy it in remote areas?
At the annual meeting we went beyond these questions. Instead, people
who actually spent time in India, in Malawi, in the Dominican
Republic, and elsewhere discussed what they learned about the value of
communications and computers, and how they brought these things to
local residents in meaningful ways. Throughout all the speakers talks
ran the critical thread: understand your users and their
needs. Work with these needs in creative ways.
Liby Levison, for instance, while stationed in the capital of Malawi,
experienced frequent telecom failures and pitifully slow connections.
She learned here an interesting piece of meta-design: that
underdeveloped areas need entirely different technologies for
information retrieval. There limitations made it unfeasible to use the
information retrieval strategy that we use in the developed world day
after day: enter a search term into a search engine, browse a few
dozen results, request a home page, follow a link to a resource,
etc. In rural Malawi, the Internet connection would be down before you
were half done.
To respond to the needs of Internet users in these areas, Libby
developed a deliberately low-tech system with deep ramifications. Her
TEK (for "time equals knowledge") system works a bit like Web2mail,
making use of the store-and-forward aspect of email to provide
robustness in a non-robust environment. A person enters a search term
and is emailed the Web pages corresponding to the most
promising search engine results.
There are more interesting design choices in this system than meet
the eye. TEK strips out graphics (depending on the user's choice),
information-poor pages such as portals or home pages that have mostly
links, duplicate pages, pages in inappropriate languages, and so
on. It also deals with lost mail through a protocol that acknowledges
received mail and retransmits lost mail after a timeout.
Iqbal Quadir, as a financial executive in New York, decided to try to
provide cell phone access to the poor in his native Bangladesh. To
find a base for action, he approached the
famous for its microcredit for poor entrepreneurs (mostly
women). Iqbal persuaded the bank, with some difficulty, that a cell
phone could be just as useful as a cow or a generator in forming the
engine behind a successful business. Cell coverage is now offered to
30% of Bangladesh's territory, reaching 50% of the population.
Across the subcontinent on the West coast of India, Daryl Martyris of
World Computer Exchange
distributes recycled computers running GNU/Linux to schools throughout the
state of Goa. Hardware costs (as well as software costs, of course)
are cut to the minimum by hooking minimal clients up to a central
Linux Terminal Server Project (LTSP)
software. After-hours use by paying adults is popular. Of course,
fixing broken systems is a problem, but a small coterie of students
has been trained to fill WCE's needs.
Hani Shakeel described a research project in the Dominican Republic,
which happens to have amazingly advanced communications and computing
equipment in hundreds of villages around the island, thanks to a
former president's effort to win popular support. In the village Hani
chose, these centers are very popular with school-age children, who
use the visual aspects of Web sites to overcome any limitations they
suffer in understanding languages.
Hani designed an asynchronous bulletin board so that people could come and
go at their convenience. He also integrated text, graphics, and voice
in such a way that people could use whatever medium was most
convenient. He even used a text-to-speech synthesizer to allow
illiterate people to hear text messages.
As we heard the various needs of people in different areas and stages
of development, we became increasingly receptive to Judy Brewer,
director of the W3C's
Web Accessibility Initiative,
as she talked about accessibility. This is not simply a matter of
accommodating the disabled (although even that could open up the Web
to another ten or twenty per-cent of the public); measures taken for
the disabled always have value for other populations too, especially
poor people who deal with literacy problems, low bandwidth, poor
equipment, or other limitations. Measures that promote accessibility
also promote device-independence, a common concern in modern design.
A framework for understanding the growth of communications
infrastructure in underdeveloped areas was offered by Annalee
C. Babb. First one has to provide a physical
infrastructure. This allows information acquisition and communications
such as email. Next one needs to provide a financial
infrastructure with legal guarantees, and then a security
level offering privacy, secure transmissions, and authentication. Now
the people can develop online markets. The fourth layer is an
administrative that attempts to protect intellectual property
rights where appropriate "without stifling new intellectual property."
So far, we have familiar aspects on online life. But Annalee went
further and offered one more levels of access. Operational
access reflects people who are creative, who can exploit the telecom
infrastructure to produce something unique to their culture and
hopefully of value on the world market. This is level where democracy
and power reside.
A good note on which to leave the issue of universal access is to ask
"Who gets to collect and use information?" The ramifications of this
subtle question were laid out by Calestaous Juma, founder of Kenya's
African Centre for Technology Studies. He pointed out that Western
agencies collect an enormous amount of data about Africa that would be
of value to local people there, such as rainfall patterns. This
information is stored, however, in Washington. The people most
affected by the data do not decide what to collect and do not get the
information in a timely fashion (if at all). In addition to
eliminating the barriers of cost and technical access, agencies have
to consult with local users and figure out the best way to collect and
The highlight of the day for me was Dr. Patrick Ball's keynote on the
use of Information and Communications Technologies (ICTS) for human
rights. It was so eye-opening that I am saving a description of it for
an upcoming article. The talk was a true revelation for those who
never previously saw a relationship between free software and human
rights; it was informative even for those who did.
Robert Guerra described
the relatively new CPSR project he started to link computer experts
with human rights groups. He described it as "bringing the people with
the knowledge together with the people with the needs."
Privaterra reflects the key insights of the day: that one cannot help
people simply by dropping technology in their laps, but should
evaluate their organizations holistically and design specific
solutions. Along these lines, Privaterra helps human rights groups set
up encryption, firewalls, VPNs, and backups. It also brings back
lessons from developing countries to the developed world.
Doug Schuler is a long-term CPSR member dedicated to representation
and community-building; he has initiated more such new types of
representation than most members have even thought of joining. Doug
has written about community networks and played a key role in the
Seattle Community Network,
coordinated conferences on
working groups at CPSR.
Over the past couple years he has focused on pulling together
discussion, community building, and giving a voice to the previously
silent through his
Public Sphere Project.
What all these things have in common is people participating in
decisions about their future.
But CPSR cannot entirely lose itself in the idealistic construction of
new public arenas, it also to deal with existing ones. Thus, Robert
Guerra described how we joined with some 80 other non-governmental
organizations from around the world to present issues of public
interest to the World Summit on the Information Society, a meeting
started by the
International Telecommunication Union
and approved by the U.N.
CPSR itself has gone global over the years. While we always had a few
members outside the United States, we've just recently had the
resources and visibility to start some chapters in other countries. We
heard from a Peruvian member, Katitza Rodriguez, about their
initiatives in providing wireless Internet access to rural health
clinics, promoting Internet access to public information, and leading
the debate on the .pe domain. A Japanese member named Nobuo Sakiyama
reported on that chapter's intervention into the National ID debate, a
Carnivore-like device that intercepts email, and a government-funded, national Internet filtering system with a license that rules out reverse engineering or criticism.
Because it controls such a central Internet resource, and because this
year's Wiener Award went to one of its most prominent critics,
ICANN deserves its own section.
ICANN is many things: a trough at which lawyers and consultants line
up to slurp greedily from public funds, a madhouse where complex
subdivisions of subdivisions of organizations strive to make their
voices heard and are indulged or ignored at the Board's whim, and--not
least--a powerful standards-making body whose decisions have a
long-range policy impact on the use of the Internet.
CPSR chair Hans Klein, in an afternoon presentation, pointed out that
ICANN had turned the Law of the Ungovernable Internet into the Myth of
the Ungovernable Internet. Although he said that the board's recent
elimination of public representation was a classic case of a public
body being captured by a private interest, he was cautionary but not
totally pessimistic on the question of whether ICANN could be opened
up (or replaced).
At his evening talk, Wiener Award winner Karl Auerbach, one of the few
people who has the honor of getting on the ICANN board through public
election (and even the election is the butt of semantic quibbles)
talked about the ways ICANN has frittered away legitimacy and support,
such as by fighting ridiculous battles with national governments. He
described its sheer incompetence in managing the major resource
entrusted to it (the Domain Name System) as well as in its basic
actions as a business entity.
More fundamentally, Karl described the ambiguous position ICANN
occupies in between a public and private organization, possessing
governmental functions but run like a corporation. Although the
U.S. Department of Commerce could rein it in or dissolve it, they are
stymied by their confusion over its claim to by a private corporation
(and the ideology that says governments should not interfere with
private corporations). He also laid out his suggestion for breaking
ICANN into four parts along natural fault lines: one part for IP
address assignment, one for protocol numbers, one for technical
administration of the Domain Name System, and one for policy issues
related to the Domain Name System.
On a day like Saturday, everything seemed to fit together. But in
writing up the day, some fine presentations didn't seem to go in any
Tu Tran, winner of CPSR's annual student essay contest, delivered a
quite professional talk about computer forensics, looking at it from
many points of view. Courts are increasingly allowing searches of
computers for evidence related both to crimes and to civil suits.
Not all searches require a warrant: the court may allow a search
without a warrant if the person did not have a "reasonable
expectation" that the information would remain hidden. You could lose
this "reasonable expectation" through something as trivial as sending
the information to a colleague over email; now another party can
demand the information. Of course, some critical safeguards remain:
they have to show probably cause that the information pertains to a
case and give a precise description of the item to be found.
And how easily can information be found? Everyday encryption programs
are good for most purposes, but can be cracked by a determined
opponent. This includes, obviously, the U.S. government, when a
journalist in Afghanistan got his hands on a computer and hard disk
formerly owned by an Al Qaeda member. Thousands of files were
retrieved, although Ms. Tran did not reveal the contents of those
files. (Soft porn? Metallica songs?)
Deleting files, as most administrators know, offers practically no
protection against retrieval. Zeroing out a disk is little better
(although it's pretty good when done twice) and even reformatting
removes just pointers to files rather than the files
themselves. Tran's recommendation: if you want to prevent data from
being retrieved, drill a hole in your disk.
Carlos Osorio presented research questioning the very foundations of
software licensing. Far from being a form of piracy, the spread of
unlicensed software in new markets creates a bigger market for
This is not simply a matter of familiar network effects. Native users
offer the best possible marketing. Why spend hundreds of thousands of
dollars exporting marketing staff and a canned strategy to a place
with a different culture, when end-users copying your software can
talk it up with all the friends and colleagues for you?
While Carlos suggested several ways proprietary software companies can
make their licenses more appealing--such as offering good customer
service or releasing new versions frequently--he ultimately
recommended the ultimate approach as the most natural approach to
gaining markets: distributing free and open source software.
The conference was long but never tedious; tiring but not
exhausting. Several of us came away with new energy as well as new
ideas of where to apply it. I felt great pain thinking that many of
the areas where good work is being done may soon see it all swept away
by the storms and floods caused by global warming. But as much as we
can bring people to the Internet and the Internet to the people, we
can increase discussion of this and other critical issues facing us
What could you do to make a difference in the world?