No Subject

How "anonymous" (to what degree) are email messages
sent through the various remailers (Mixmaster,
EFGA, ml.org, etc)?

Have there been any studies done as to the degree
of anonymity obtained from using these remailers?
(I haven't been able to find any.)

And does anyone know of any studies being done
in determining whether or not content-based
analysis - diction and language - is a reliable
way to determine identity?

It seems that if Dick has a certain vocabulary,
uses a certain sentence structure, etc., and
regularly posts using a nym, that this form and
content could be traced to Dick, so the likelihood
of the post coming from him and not Jane would
be that much increased.

::
Anon-To: cypherpunks@toad.com

Anonymous wrote:
 
> How "anonymous" (to what degree) are email messages
> sent through the various remailers (Mixmaster,
> EFGA, ml.org, etc)?

  Everyone who uses a remailer is perfectly anonymous, and thus safe
from having their identity discovered. Remailers are now so advanced
that it is virtually impossible for anyone to discover the user's true
identity.
  Also, the Law Agency Freedom Foundation has verified that no law
enforcement agencies are running remailers, out of respect for the
citizen's right to anonymity.

> And does anyone know of any studies being done
> in determining whether or not content-based
> analysis - diction and language - is a reliable
> way to determine identity?

  This is patently ridiculous. CypherPunks have long tried to figure
out, by use of, as you said, these means, to ferret out schills,
yes, schills, on the list. In totoal, to date, they have been unable
to successfully do so.
  ? the Platypus say heyed to me joust the otter day that hee...hee
warts able two foal Manny peephole awn the CyperPukes lips width
his alanon e-mouse pastes.

  Hope this clears things up for you...

Anonymous (see! it worked for me)
p.s. - and headers are meaningless, too

> Have there been any studies done as to the degree
> of anonymity obtained from using these remailers?
> (I haven't been able to find any.)

Have a look at:

David Chaum: Untraceable Electronic Mail, Return Addresses, and Digital
Pseudonyms, Communications of the ACM 24 (1981) 2, pp 84--88,
http://world.std.com/~franl/crypto/chaum-acm-1981.html

Wei Dai: Traffic analyzing Chaum's digital mix, 1995,
http://www.eskimo.com/~weidai/traffic.txt

Lance Cottrell: Mixmaster and remailer attacks, 1995,
http://www.obscura.com/~loki/remailer/remailer-essay.html


> And does anyone know of any studies being done
> in determining whether or not content-based
> analysis - diction and language - is a reliable
> way to determine identity?

Thomas Horton: Stylometry. In: R. Asher (ed.): The Encyclopedia of
Language and Linguistics, Pergamon Press, 1994, pp 4383--4385.

Joseph Rudman, David I Holmes, Fiona J. Tweedie, R. Harald Baayen: The
State of Authorship Attribution Studies, 1997,
http://www.qucis.queensu.ca/achallc97/papers/s004.html

A somewhat related area:

Ivan Krsul, Eugene Spafford: Authorship Analysis: Identifying The
Author of a Program, 1997,
ftp://coast.cs.purdue.edu/pub/COAST/papers/krsul-authorship_analysis_NISSC.ps.Z