Re: more ideas on anonymity

_________________________________________________________________
             FROM THE VIRTUAL DESK OF SANDY SANDFORT
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Today, Theodore Ts'o raised the specter of libel and slander to
justify some forms of censorship for anonymous remailers.  He
assumed, rhetorically, that if you believe in total freedom of
speech, than you must not believe in libel or slander.

He is correct, of course.  Nevertheless, his implied conclusion
is in error.  He would have us give up free speech to uphold
libel and slander.  I say, let's give up the artificial concepts
of libel and slander and uphold our freedom of speech.

Can people be harmed by speech?  Maybe yes, maybe no.  But if so,
that is the price that must be paid to avoid a far greater harm.
Damnum absque injuria--there are some loses for which there is no
remedy at law.

If your girlfriend impugns your manhood, there may be harm.  But
should her speech be ILLEGAL?  I think not.

I'm sure that what NBC did was NOT libel or slander.  It may have
been fraud, though.  In any event, if a free society chose not to
make what NBC did illegal, that in no way implies that NBC was
"perfectly justified" in faking an explosion.  It just says
society takes any remedy out of the hands of the state.  Of
course, NBC's acts have already damaged its own reputation karma.
Serves 'em right, too.

One last note about practicality.  We live in a world with pay
telephones, which anyone may use to threaten anyone else,
anonymously.  Shall we eliminate pay phones?  Today in the United
States, a uniformed agent of the government will deliver your
anonymous threat to your victims door for only US$0.29.  Shall we
eliminate the Postal Service?  (The answer is "yes," but for
other reasons.)  So what's the point?  Why should we be any
harsher on the networks than we are on all the other anonymous
channels?

There is only one realistic and moral solution to threats of
violence.  Punish the perpetrators, not the messenger.  Even in
today's world, crimes are rarely solved by wiretaps, forensic
science or any of that.  Perps gets fingered by someone they
know; perps confess; perps screw up and accidentally give
themselves away; or the intended victim blows the perp's head off
during an attempt to carry out the threat.  Again, punish the
perpetrators, not the messengers.

     S a n d y                         ssandfort@attmail.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

From: ssandfort@attmail.com
   Date: 1 Mar 93 02:36:10 GMT

   There is only one realistic and moral solution to threats of
   violence.  Punish the perpetrators, not the messenger.  Even in
   today's world, crimes are rarely solved by wiretaps, forensic
   science or any of that.  Perps gets fingered by someone they
   know; perps confess; perps screw up and accidentally give
   themselves away; or the intended victim blows the perp's head off
   during an attempt to carry out the threat.  Again, punish the
   perpetrators, not the messengers.

I agree 1000%!!!!  But in order to punish the perpetrators, it is first
necessary to *identify* the perpetrators......

							- Ted

>> One last note about practicality.  We live in a world with pay
>> telephones, which anyone may use to threaten anyone else,
>> anonymously.  Shall we eliminate pay phones?  Today in the United
>> States, a uniformed agent of the government will deliver your
>> anonymous threat to your victims door for only US$0.29.  Shall we
>> eliminate the Postal Service? 

This is different than anonymous remailers currently.  Why?  Because
pay phones and letters *cost* *money*.

So, I have a proposal.  Have someone *sell* signed keys, for real
money.  The only purpose of the signature is so that someone can
verify that it's been paid for.  It can be bought out of band.  For
example, I can email the signing agent my public key, and paper-mail a
money order with no return address.  Maybe in the future, we can just
use digital money.  When he gets my money, he signs my key and sends
it back to me, possibly via some remailer or whatever.

Now, when I send a message via a remailer, the mailer checks to see if
the message has been signed by a key which has been signed by the key
signing agent.  If it has been, then it forwards the mail.  If not, it
doesn't.

Now, here's the trick.  If too many people complain about a particular
sender (see below), the key gets revoked by the signing service.  This
means flooding newsgroups, slander, etc., *costs* *real* *money*.
This is a big encouragement not to abuse the net, because it costs
money.  Perhaps we even have different levels of signing, based on
costs.  Pay $1, and the remailers will let you send to alt groups
only.  Pay $10, and you can send to alt, soc, misc, talk, and some set
of high-traffic, low-signal lists. Pay $100, and you can post
anywhere, including individuals.  These are all example numbers, of
course.

 - People can easily remain pseudonymous.
 - There is a monetary incentive not to be a bonehead.
 - Revocation is possible.  There's no punishing wrongdoers, but
people aren't likely to put much credence behind anonymous claims.
 - This works like the Real World.  People with enough money can
create false identites for themselves, anyway.
 - Remailer owners can choose which signing agents to trust.  Other
than monetary schemes might be able to be set up.  For instance, if
you can get five real, trustable people to vouch for your pseudonym,
you win.  If you screw up, their reputations become tarnished, as
well.

I guess all this is is a positive reputation scheme, bootstrapped by
money, which is more important to some people than reputation, anyway.
I mean, someone willing to go through the bother of spending $10 to
get his key signed is either very bored, or has to have something to
say.

Now, back to complaints.  What constitutes a complaint?  Each
certifier would probably have to decide this on their own.  Some might
choose never to revoke a signature.  Some might do it only after some
number of complaints.  Some might come up with systems of arbitration,
with judges, voting, and everything.  The remailers can vote with
their feet: whichever signing agents match the policies they like,
they trust.  And recipients (like newsgroups, mailing lists, or even
individuals) can choose to accept messages only from remailers they
"trust".

So, we have a fairly complex system with multiple degrees of freedom,
where endusers can pick and choose from among the policies they like
best.  This is sort of like credit card purchases.  Hertz is willing
to rent me a car based on Visa's evaluation of my creditworthyness.
Similarly, large-public-mailing-list@site.domain can decide to accept
messages from a given remailer, based on their published policies.
Actually dealing with signatures isn't their problem.  It's to the
signing agent's advantage to be trustworthy, or else nobody will pay
attention to his signatures, and nobody will give him money to sign
their keys.  Poof, a free market.

It isn't simple, but I think one thing we can all agree on is that no
solution to these problems we are facing will be both adequate and
simple.

		Marc