Nicklas Lundblad's paper on Privacy in a Noise Society (pdf, via Politech) provides an interesting framework for thinking about privacy in policy and tool design. He contrasts the collective vs. the individual expectations of privacy. If both are high, the result is Privacy Society (e.g. EU legislation), which has high costs of administration and interpretation of privacy laws and investments in privacy enhancing technologies. If both are low, the result is a Surveillance Society (think Orwell), which has high costs of collection, classification and structuring of data and archiving, format conversions and storage. Because information is abundant, he believes that the costs are so high in both cases that a collective privacy border prevents pragmatic policy. The conclusion is that we are in a Noise Society, with a high expectation of collective privacy and low individual privacy, which pragmatically suggests policies the focus on abuse of abundant information: In a Noise Society, the suggested guiding principle is to avoid attracting attention, or anonymity loves a crowd, which implies avoiding use of encryption and explicit resistance to the system using tools that help blend into statistical norm or generate noise. I find this bothersome, not only because it disrupts community dynamics and discourages diversity, but while a policy that is based on enforcement of abuse rather than constraining regulation -- the social costs are too high. The fourth quadrant is discounted far too easily: For the sake of completeness it is also possible to include a strange and unusual kind of society where the collective expectation of privacy is low, and the individual level of privacy is high. He suggests this is the relm of science fiction (with a wrong example of the Borg as having high individual privacy) or saunas. A Network Society of low collective and high individual expected privacy would represent the emerging decentralized structure of the web as Invisible Villages. The policy implication is the same as with a Noise Society, regulate abuse, not design, for enforcement and cost containment. For tool design, the implication is fulfilling the demand for user control of expression of identity, relationships, group formation and information shared. Now that's the strange and unusual kind of society we should strive for.