On his way home from Salford University, Robin Wilton has the good fortune to pass Strangeways Prison, and is prompted to blog about the panopticon.
POSIWID teaches us to look at the effects of a system rather than its avowed purpose. I have just re-read a post by Scribe (The two faces of CCTV) [URL updated] in which he discusses and dismisses several avowed (and contradictory) purposes of CCTV. Among other things, surveillance is supposed to teach good behaviour to those being watched. In reality, surveillance often merely teaches more devious or secretive behaviour.
Most of the discussion of panopticon revolves around the people under surveillance. But we should also consider the corrupting effect of panopticon on those doing the watching. See my previous posts Guarding the Guardians and Surveillance and its Effects.
In March/April 2005, there was a lengthy debate between Stefan Brandt (from Credentica and McGill) and SuperPat (Pat Patterson of Sun Microsystems) as to whether the Liberty Alliance counted as panoptical. There is a useful index to the debate by Kim Cameron (Microsoft). But this debate revolved largely around the technical features of the Liberty Alliance architecture, and on the hierarchical/network trust relationships. I don't deny that these details are important and interesting, but I don't think they have anything to do with the panopticon.
The panopticon was designed to produce certain effects - certain changes in behaviour in the actors. In my view, this is what is most important in deciding whether to regard something as a metaphorical implementation of the panopticon. I haven't seen any contribution to the Liberty/panopticon debate that identified any such effects.