Consider the implications for the censor of this technological change. The rise of publishing
through blogs has caused concern in China, Iran, and Saudi Arabia at a minimum, judging
from the reaction of their filtering regimes to block some blog-hosting services wholesale for
a period of time. Assume that all blogs within the persianblog domain are available via RSS
feed. The publisher could create a Web site specifically for the purpose of evading blocking,
listing and displaying all such blocked blogs. This site itself could become a target for blocking by the Iranian government, since any mechanism for making this site known to users
would also make it known to the filtering authorities.
35
But using widely available aggregation tools, a user who wants to read this information
does not need to go to a single URL to access the information published there. Instead, the
user only needs to know the place where the XML feed is located at any given moment—
which need not, ultimately, be at a stable location, so long as the user has a means of being
updated as to its location at any given moment. In this version of the Web—trivial, using
today’s technologies—anyone can make any such blogs they choose available on any Web
page or in an e-mail in-box or on a mobile device.
Another approach that citizen journalists might take would be to seek to bury the blocked
blogs within a much larger number of blogs. The publisher could then establish a site or a
feed that aggregates this larger number of blogs. Then, still using simple technologies, the
readers could either read the full set of aggregated information or could run a filter of their
own against the aggregated group of blogs to distill the information that the publisher wanted
them to be able to access. Though these methods add a layer of complexity that would no
doubt dissuage some Internet users, the net effect would be a publication mode that would
be extremely difficult for the state to filter using current methods.
The state’s censor would still have several options for responding to syndication methods
of dissemination. First, the state could attempt to ban syndication, aggregation, and peer-topeer technologies that might make these circumvention efforts easy to carry out. States have
not, however, tended to pursue such a heavy-handed mode of regulation. Second, the state
could seek to block the sites where the information is published and where the aggregation
takes place. However, the potentially unlimited proliferation of such blogs and aggregator
sites makes this unfeasible. A last option could involve a fallback to more traditional forms of
state coercion—threatening both bloggers, readers, and those who provide them services
with sanction. The difficulty of anonymous access leaves open the alternative of identifying
users after they have accessed banned content. It is this last option that seems most in keeping with previous filtering and surveillance practices, especially since intermediaries closer to
the user can be pressed into service to help.
The enduring point of this glimpse not so far into the future is that as Internet technologies
continue to evolve, so too will state censors have to evolve their methods of Internet filtering
if they wish to keep up. ONI’s early election monitoring efforts in Kyrgyzstan and Belarus,
combined with some of the most recent test results from the Commonwealth of Independent
Internet Filtering: The Politics and Mechanisms of Control
53
through blogs has caused concern in China, Iran, and Saudi Arabia at a minimum, judging
from the reaction of their filtering regimes to block some blog-hosting services wholesale for
a period of time. Assume that all blogs within the persianblog domain are available via RSS
feed. The publisher could create a Web site specifically for the purpose of evading blocking,
listing and displaying all such blocked blogs. This site itself could become a target for blocking by the Iranian government, since any mechanism for making this site known to users
would also make it known to the filtering authorities.
35
But using widely available aggregation tools, a user who wants to read this information
does not need to go to a single URL to access the information published there. Instead, the
user only needs to know the place where the XML feed is located at any given moment—
which need not, ultimately, be at a stable location, so long as the user has a means of being
updated as to its location at any given moment. In this version of the Web—trivial, using
today’s technologies—anyone can make any such blogs they choose available on any Web
page or in an e-mail in-box or on a mobile device.
Another approach that citizen journalists might take would be to seek to bury the blocked
blogs within a much larger number of blogs. The publisher could then establish a site or a
feed that aggregates this larger number of blogs. Then, still using simple technologies, the
readers could either read the full set of aggregated information or could run a filter of their
own against the aggregated group of blogs to distill the information that the publisher wanted
them to be able to access. Though these methods add a layer of complexity that would no
doubt dissuage some Internet users, the net effect would be a publication mode that would
be extremely difficult for the state to filter using current methods.
The state’s censor would still have several options for responding to syndication methods
of dissemination. First, the state could attempt to ban syndication, aggregation, and peer-topeer technologies that might make these circumvention efforts easy to carry out. States have
not, however, tended to pursue such a heavy-handed mode of regulation. Second, the state
could seek to block the sites where the information is published and where the aggregation
takes place. However, the potentially unlimited proliferation of such blogs and aggregator
sites makes this unfeasible. A last option could involve a fallback to more traditional forms of
state coercion—threatening both bloggers, readers, and those who provide them services
with sanction. The difficulty of anonymous access leaves open the alternative of identifying
users after they have accessed banned content. It is this last option that seems most in keeping with previous filtering and surveillance practices, especially since intermediaries closer to
the user can be pressed into service to help.
The enduring point of this glimpse not so far into the future is that as Internet technologies
continue to evolve, so too will state censors have to evolve their methods of Internet filtering
if they wish to keep up. ONI’s early election monitoring efforts in Kyrgyzstan and Belarus,
combined with some of the most recent test results from the Commonwealth of Independent
Internet Filtering: The Politics and Mechanisms of Control
53