In article <+cmu.andrew.internet.cypherpunks+4kz3qSq00UfAM0yv9n@andrew.cmu.edu> Mr. Nuri writes:

that 9 of the last 10 flamewars on this list were actually carefully orchestrated, *manufactured* by a single person interested in making this point, and teasing people that refused to believe that rampant dischord can be sown through a barrage of pseudonyms.

Enough already. We all know that discussion groups are subject to disruption by flamers, and that anonymity can reduce accountability. What is more interesting is whether the use of pseudonyms adds any new possibilities.

how can you be so sure that the cypherpunks lists is really what you think it is? a bunch of people from around the country independently interested in crypto? an agent provacteur, or agent saboteur, could create a vastly different perception regardless of the input of other people.

Ah, conspiracy theories again. A conspiracy of one does have certain advantages over the old-fashioned approach of gathering several like-minded people. Communication costs are reduced, and the problem of trust is eliminated. On the other hand, a conspiracy of one isn't much good for anything but playing games on the net -- cattle mutilation is really a two-man job, and infiltrating the U.S. Government is right out. Being a good Medusa may be easier than controlling the banking system, but it seems to be quite hard. Your history suggests that stylistic analysis may be used to link nyms. A sufficiently skilled writer might be able to avoid this. An individual known to me was wanking around with pseudonyms (from nyx, I think) on rec.music.industrial in '92-93; he did a decent job of stylistic variation, but was noticed on the basis of the response patterns of his articles (and nailed by nyx usage logs, but that's another matter). There are several Usenet examples of people trying to use inapparent pseudonyms without lasting success. The classic response, of course, is that the real conspiracies are too good to be detected, and/or they off the investigative reporters. This is not falsifiable, but we can compare pseudonym conspiracies to meat ones: the risk is lower, particularly in an regime of unconditionally-secure pseudonyms from which the Medusa can simply walk away. But you don't get real-world power or money. You get to put in a lot of effort for a chance at unduly swaying public opinion on one forum. (And if you screw up, you get a lot of attention -- this may be a major draw for some.) But there are easier and more effective techniques: press releases and paid advertising, for example. Who needs pseudonyms? Side note:

another very interesting effect to measure is the following: if there is already a lot of mail on a list, people tend to post less.

I would say the opposite. -- Eli Brandt eli+@cs.cmu.edu