The proposal to limit "noise" posts by using re-usable tokens is interesting because it has the effect of filtering and at the same time remains responsive because the "moderator" passes judgement after posting has occurred. In essence, the poster is betting that the moderator will approve the post retroactively. The purpose of limiting the tokens available is to give them value so that there is a cost associated with losing the bet. However, distribution of the tokens has the same problems that centrally planned economies have distributing bread. Should people wait in line, receive bread at random, or get their bread through connections? None of these solutions is attractive. The right way to implement the scheme is to use dollars for tokens because the USG has already solved the problem of giving the tokens value. The "moderator" returns the money to people when posts are worthwhile and keeps it otherwise. This means that the "bet" can be quite a bit larger than a dollar because responsible posters will get their money back. Starving graduate students will not be discouraged from posting. Abusive posters, anonymous posters, or spam artists will have to pay a substantial fee for the privilege. Of course, there's no reason to have just one moderator on the list. A moderator could just be an e-mail account that forwards mail to the list if valid payment is received. (Moderators should PGP sign their messages so mail forging won't work.) Readers could filter on which moderators have approved the kinds of posts they want to read. The effect of this scheme is to allow newcomers, infrequent posters, or anonymous posters to get through killfiles without taking the time and effort to develop a reputation. Peter Hendrickson ph@netcom.com