From: fc@all.net (Dr. Frederick B. Cohen)
Date: Fri, 25 Aug 95 04:23:52 PDT
To: an366601@anon.penet.fi
Subject: Re: The Theory of Nymity (value of True Names)
In-Reply-To: <9508250244.AA26841@anon.penet.fi>
Message-ID: <9508251117.AA13712@all.net>
MIME-Version: 1.0
Content-Type: text


> |   /\  |\| /~ L~ anonymously spewed forth
> L_ /~~\ | | \_ L_ as follows:
> 
> The Theory of Nymity
> 
...
> The cypherpunks definitely pioneered forms of nymity, and passionately
> champion the general issue, particulary its ramifications in
> cyberspace; however they refuse to confront the issue openly and
> overtly, suggesting they perceive some sort of stigma or taboo
> associated with their practices.

They seem to be openly confronting the issue to me.  There is a stigma
attached to anonymity - haven't you seen the lone ranger?  If you're
honest, why do you wear a mask?

> They also refuse to contemplate
> negative or potentially disastrous social effects of the associated
> scenarios, asserting all forms of nymity are either indistinguishable
> or essentially morally neutral practices that invariably extend the
> rights of the individual in the face of the oppressive State or
> massive corporations.

I think they contemplate the issues.  I think that most of them (us?)
realize that:

	They are the ones likely to be most harmed by a lack of open and
	honest communication. 

	By identifying themselves they might open themselves to undue
	risk of harm in the form of harassment, arrest, search and
	seizure of personal property, etc.

	The only way they can bot help preserve individual freedom for
	others and remain free themseleves is by saying some things
	anonymously.

> They see all forms of nymity as merely one
> unified hacksaw of the serf, useful and effective in cutting away
> chains of bondage in the Sisyphean struggle for privacy.

They certainly see the advantages over the disadvantages.

> Detweiler confronted the diverse implications of nymity by elevating
> the subject of "nymity" to a study complete in itself. Abstracting
> from his writings, Detweiler considered the concept of a "nym 
> spectrum" involving the key concepts of receiver of a message,
> sender, and identity. At various points on the nym spectrum, the
> receiver has varying degrees of knowledge about the identity of the
> sender.

There are times when identity is important, and there are times when it
is not.  Identity is not important when disclosing a danger - if you
yell "FIRE!!!" in a burning building, nobody is going to say "Hold on
there...  before I get out, what's your name and address, and how can
you prove that you are who you are?" They recognize the danger you
identified and get out of the building.  When I buy a can of fruit juice
from the local store and pay with cash, if they ask my name and address,
I tell them (politely) to stuff it.  You don't need to know my identity
to trade a fungible for goods or services.  That "need to know" phrase
is important here.  Identity is only important to reveal if there is a
legitimate need to know it, and if that need to know outweighs the risks
associated with the knowledge.

...
> Detweiler defined the various forms of nymity based on the knowledge
> of the sender, say party C. This is his critical distinction that
> cypherpunks denied, arguing that all the forms of nymity he saw a
> difference between were really interchangeable and indistinguishable,
> and therefore identical. They do not recognize any relevance of the
> "knowledge" or mental state of the receiver in  regards to messages.
> (In a sense the  philosophy is similar to behaviorism in asserting
> the invalidity of internal mental state.)

I think the cypherpunks recognize that by being anonymous they risk
their message being taken less seriously, but they are willing to take
the risk because most people on this list seem to evaluate the content
of the message more than the sender anyway.  Contrast this list to other
lists and you will see far fewer personality issues here because, in
part, the moderator doesn't inject his/her personality into the
discussion all the time. 

...
> The cypherpunks argue that there is in principle no difference between
> Detweiler's "pseudoanonymity" and the classic "pseudonymity", stating
> that anywhere there is an "actual" nym, a receiver cannot be sure it
> is not really a formal one, and vice versa (i.e., any nym is
> potentially a pseudonym or true name). In fact they say there is
> fundamentally no distinction to be made between formal and actual
> nyms.

No - perhaps some of the cypher punks would argue that, but certainly
not most.  Many cypherpunks use pseudonyms and we trust them to some
extent because of the repeated usage of the pseudonym.

> The critical question is of course is how "actual" and "formal" nyms
> are defined. What does it mean to "identify" a sender with a nym?
> Detweiler adapted to the ambiguities in the following way. He defined
> the "actual nym" (or, interchangeably, the "true name") as an
> identification for party A if for every context where an actual nym
> applies, that party is identified as A. Call this the "actual nym
> scenario".  The definition appears to be somewhat circular but he was
> able to derive conclusions from the premise.

Hogwash.  A rose by any other name would smell as sweet.  Names are just
conveniences created as a part of language because we are able to use
nouns to identify objects.  That guy who runs all.net is just as good as
Fred Cohen in terms of identifying me, and I rarely use my full name,
and there are otyher people in the world with exactly the same full name
as mine - even including the Dr. (a mathematician I believe).

> Suppose that a context was established in which the "actual nym"
> scenario applied, and a message was identified as from "A". Then "A"
> is the actual nym of the sender. Suppose that the context is not
> "actual" (the converse will be considered the "formal") and the
> message is identified as from B. If there is some  "additional
> information" that B is a "true name",  then B is the actual sender,
> and no other party sent the message (forms of the "additional
> information clause will be considered below). Otherwise no conclusion
> can be made about the actual sender.

A lot of conclusions can be made about the sender based on the content
and context of the message. 

> In general, in an "actual context" the actual sender is always the
> formal sender. In a formal context, the actual sender is dissociated
> from the formal identity of the message,  but "independent
> information" may pinpoint the actual identity.
...
> Obviously contradictions can arise within these definitions if the
> sender "lies". Detweiler simply observed that a logical theory could
> be built up based on his premises from which  conclusions could be
> drawn. Furthermore, there is the fundamental observation that in a
> formal nymity system, "true" conclusions about actual identity can
> only be derived from message contents  if participants "don't lie". A
> core premise of the theory is  that an entity can express statements
> such as "my true name is [x]" in the communication system. (This
> is one form of the "additional information" clause above.)

Baloney.  If someone claims to be me (likely a masochist) and I am
listening, I can identify it for all to see.  As long as the society
doesn't tollerate excessive amounts of this and fights back against the
lack of integrity, it is likely that any person who does this will be
detected.  The point here is that identity in the form of a name is only
a tag we attach for convenience of associating actions with individuals. 
It doesn't matter what the name is - I don't know the real name of
HOBBIT, but I know that s/he says a lot of sensible things. 

> The cypherpunks believe that if the communication system includes only
> the formal context, there is no such thing as a "lie".  Detweiler
> however maintained that as long as the actual system exists (or more
> particularly the overlay of a formal system over an actual one), lies
> can exist, although they may be undetectable within the formal
> system. (In this sense it is analogous to a Godel Theorem for Nymity,
> making an observation about a phenomenon that "exists" but  is
> "undetectable" within the formal system, with the parallelism of
> "mathematical" and "metamathematical" statements mapping to "formal"
> and  "actual" communication systems.) 

Lies are relative things, not absolutes.  There is no "actual" system
and there never was.  Detweiler assumes that the way s/he identifies
people is the "actual" system, but that's only one perspective.  Many of
us have grown up knowing about more than one system of identities, and
apparently Detweiler missed this one.

> The Detweiler Thesis
> 
> Detweiler had a preoccupation for considering the "community", which
> is defined as a formal or actual context in which communication takes
> place in both ways between a group of entities (i.e. entity A may be
> a receiver of sender B and vice versa for all members of the group).
> He explored the implications of both the formally and actually
> identified communities, and consistently objected to the actual
> identification scheme as at least uninteresting  and at most too
> constraining, mirroring the quintessentially cypherpunkesque
> position. However Detweiler diverged from the philosophy by
> repeatedly emphasizing a basic premise, which is currently unprovable
> speculation  much the same way that the  Church-Turing thesis is: in
> a formally identified community system  where participants don't
> "lie" about actual identity, the communication of the "community" is
> of "higher quality".

The first error is usually the basic assumption.  This one is not a very
accurate reflection of the reality that many of us in the world live in
today.  The fact is, our political system is full of people who use one
and only one identity and have done so for their entire lives, and yet
it is jam packed with lies and crude (As well as sophistocated) attempts
at tricking people with communications.  The cypherpunks, who have far
more uses of pseudonyms have far fewer deceptions and far better
communications in my experience.  Many things that cannot be proven
are actually false.  This counterexample demostrated that Detweiler is
not just unprovable, s/he is wrong.

...

All of the confirmations Detweiler may be able to come up with can't
overwhealm the one counterexample just cited, however, I can cite many
more counterexamples.  By the way, counter to Detweiler's proposition,
plenty of scientists lie about results, and many University professors
who claim to be scientists lie about the true state of their work so
that they can propose research grants that are certain to be successes
(it is a common practice to propose work you have already completed so
as to assure that you will be successful and thus get the next grant). 

...
> The idea of a "true name" is a very problematic and perplexing 
> concept in the same way that the concept of "absolute space" is
> troubling to the theory of Newtonian mechanics, which Einsteinian
> relativity sought to remove, starting with the premise that "there is
> no preferred reference frame". The cypherpunks cite the absurdity of
> the "true name" concept in an analogous argument and attempt to
> discredit Detweilerian theories on this basis. However the theory is
> not based on "true names" but the existence of "entities", hence this
> cypherpunk position translated to its most basic form, becomes,
> essentially, "unique communication sources (such as 'humans') don't
> exist"  (or analogously in the Einsteinian metaphor, "mass and energy
> do not exist").

I have discounted his/her thesis on the basis of its lack of accurately
predicting or reflecting reality and on no other basis.  It may be an
interesting theory, but in practice it seems to have been refuted.

...
> The essence of the idea of "true identification" is that there is a
> one-to-one mapping between "entities" and "true names". The
> representation of the "true name" is irrelevant. Detweiler certainly
> did not make the absurd claim that a "true name" had any special
> syntactic properties. He also did not claim that "true name
> registries" had to be erected to provide the feature, although they
> could support it. At the core of the concept is the idea that every
> entity in a "truthful" system must make a choice as to their actual
> name identification and not "lie" about its properties in messages
> that refer to it. 

So legal name changes are not allowable? Two people can't have the same
name? This doesn't map to the current reality.  It sounds more like a
government imposed identity number system - like the one in Nazi Germany
that contributed to the extermination of 6 million Jews.  I'm against it.

...
> Detweiler countered by suggesting that communities with communications
> based on trust and honesty and the contrary not only both exist, but
> that discrepancies between the two probably existed as well and
> furthermore were worthy of study. (Again, he conjectured that the
> "dishonest" communications forums led to "disharmony" without further
> defining the term.) All these distinctions lie in the area Detweiler
> denoted under the heading "morality" which again the cypherpunks
> generally deny exists in an abstract communications system.

He said - she said.  It's just a bunch of baloney.  Lies are often the
basis for harmony and being "too honest" generally gets you very little. 
The important thing about communication is that you get other people to
think about the things you think about.  It's just a way of affecting
brain states in other entities and nothing more.  Truth, lies, and video
tape can all do this.

> Hence the key formal ideas of the overall theme that communications
> systems could lead to significantly different scenarios based on the
> "honesty" or "truthfulness" of members of a community with respect to
> identity had been addressed for the first time by Detweiler,

I appreciate Detweiler's attempt, but it's a failure (oops better try
only a partial success) because it doesn't reflect the reality.

> but at
> great cost to Detweiler's credibility within the cypherpunk circles,
> which have rebuffed, ridiculed, and excommunicated him.

If you can't take the heat, stay out of the chemistry lab.  I say lots
of things that people don't like and gripe about.  If Detweiler can't
shake it off and come back with something better, s/he's overly
sensitive.  What do you expect? That people will embrace your
characterization of them just because you think it's interesting? If
Detweiler is really right, then s/he should fight back with the
strongest weapons that exist in this forum - the facts.  As far as I can
see, the facts are against Detweiler, but s/he may prove me wrong in
open debate, and as far as I can tell, the cypherpunks will agree with
Detweiler if s/he is right.

> Detweiler
> tended to take this as evidence that the core cypherpunk philosophy
> was not about seeking privacy, which he went to pains to demonstrate
> existed in his "honest" systems, but rather a sort of denial of the 
> existence of morality in cyberspace-- that the question of "whether a
> message 'lies' about it's authors identity" is inherently
> meaningless, a premise he strongly rejected.

On this, Detweiler is clearly incorrect from my view.

I don't want to waste further space on this.  I'm already sorry I spoke
up on it and I haven't even sent the mail yet. 

-- 
-> See: Info-Sec Heaven at URL http://all.net
Management Analytics - 216-686-0090 - PO Box 1480, Hudson, OH 44236