
March 4, 1994

The Unscrupulous Diner's Dilemma and Anonymity in Cyberspace

By: David R. Johnson

We are beginning to understand, on a scientific basis, the fundamentals
of group behavior -- and this new understanding has implications for the
law of cyberspace.  This month's Scientific American includes a
fascinating article on "the Dynamics of Social Dilemmas" -- a study of
cooperation featuring the example of a group that has agreed to split
the meal check evenly and then must decide, individually, whether to
order lobster or hot dog. The bottom line: collaboration occurs in
groups that are small, last a long time, and know a lot about each
other.

At a glance, this finding seems a disaster for the prospect of
civilization in cyberspace, where communications span the globe,
clusters of online citizens meet only sporadically and for brief
periods, and the malleable nature of the medium can cloak much of the
information about others online. How can Counsel Connect, for example,
hope to foster collegiality in the legal profession's electronic
guildhall if the scientists are right? Won't the participants in
transitory, impersonal communications across large groups engage in
"free riding", spoofing and exploitation -- and subject us,
collectively, to a tragedy of the electronic commons?

To the contrary, it turns out that electronic communications foster just
the type of trust and collaboration the scientists predict for small,
stable and personable groups -- they give rise to community -- IF
certain ground rules apply. If users are allowed to form separate areas
of the system, then even large systems appear, functionally, to consist
of much smaller groups of "regulars." The local pub (or cracker barrel,
as some would say) emerges naturally.  If users are required to
communicate under their own identity, then there is less temptation to
exploit others or abuse trust.  Particularly in the context of a
professional communications network, the participants behave responsibly
(even thoughtfully) because they have to expect to continue to be found
at the same electronic address throughout their professional careers.
And, over time, even this text-based medium is rich enough to convey as
sense of personal acquaintance with those we have never met face to
face. Repeated selfless contributions to the community will be
remembered and can be rewarded over time by the group -- and antisocial
conduct can be punished.

If the electronic network creates new, tightly integrated social groups
capable of eliciting collaboration and punishing defection, then these
groups may exercise more powerful social control than those that apply
in the "real" world -- and there may be less need for a more formal type
of "law" in cyberspace. Everyone knows that complex contracts count for
less when parties must deal repeatedly with each other and become highly
interdependent.  The US has more lawyers than any other country, the
theory goes, because we are so much less integrated as a social group
than the homogeneous cultures of other countries. As we all move our
commercial transactions to "EDI" (electronic data interchange) and
portions of our cultural and social lives become electronic, perhaps we
will all live in more homogeneous online cultures and the need for
lawyers will decline.  There might be a need for protection of the
rights of individuals against such groups -- but diversity coupled with
freedom to find a group compatible with each individual's values may
even make such formal rights less necessary. The real conflicts in
cyberspace seem more likely to occur between groups, across social
boundaries.

Presumably, the scientists' findings apply to interactions between
groups as well. We can expect collaboration to arise between distinct
cyberspace cultures that must deal with each other on a continuing
basis. The optimal conditions can be approximated IF the system
operators can effectively speak for their members and control their
members' activities. I predict that we will see the rise of compacts
between SYSOPs regarding such matters as accurate labelling of files,
enforcement of rules privacy, and system security. To some degree, the
global Internet is already based on an informal set of such agreements.
But any such agreements will require the collaborating systems to be
able to enforce their own local rules locally -- and that implies an
ability to identify (and discipline) anyone who deviates from such
rules.

The ultimate implication, I believe, is that to achieve a civilized form
of cyberspace, we have to limit the use of anonymous communications.
Many early citizens of cyberspace will bitterly oppose any such
development, arguing that anonymous and pseudonymous electronic
communications are vital to preserve electronic freedoms and allow free
expression of human personality.  But the problem with that view is that
we all collectively face the diners' dilemma -- we must collaborate in
groups to build a rich social fabric, and we know that the ability to
act anonymously, sporadically, in large groups brings out the worst in
human character.  One thing is clear: the architecture of the network
will have a profound impact on the nature of the societies that form
there. Now that we know what's at stake, it's time for the sysops who
control key policies (such as those associated with the identity and
identifiability of participants) to build in the structures that hold
the greatest potential for altruism, collaboration and responsibility.


