Doing Justice to Others 35Th e attitude of the Greeks illustrates the position all too clearly. Prior to
the conquests of Alexander the Great, there was scarcely a pretense that the
noble ideas of Plato or Isocrates would apply to relations with non- Greeks.
Plato made it clear that his precepts about moderation and justice in war did
not apply to confl icts with the barbarians. With them, there should be “war
to the death.” Aristotle was of a like mind, famously contending that bar-
barians, as inferior beings, were naturally fi tted to be the servants, or slaves,
of the Greeks. Th ere would appear to be no record of a Greek polis ever
concluding a treaty with a barbarian state.
An absence of respect or aff ection for exotic foreigners need not, however,
rule out the devising of de facto working relations on the basis of some
discernible shared values— that is, on the basis of something that could,
without undue distortion, be called the rule of law. Th is fact was vividly
demonstrated in the 1970s, when social scientist Robert Axelrod ran a con-
test in which various strategies of utility maximization, in the form of rival
computer programs, were pitted against one another. Two forms of conduct
were available— cooperation and defection— and no consultation or prior
agreement between players was allowed. Th e game was structured so that
mutually cooperative behavior was the best outcome— but the problem was
that an individual who played cooperatively risked being exploited when
playing against a defector. Th e challenge, in essence, was to devise a strategy
that best promotes cooperation while at the same time ensuring against ex-
ploitation by an opponent who might choose to defect at any time. Each
player’s preprogrammed strategy was matched against each of the others for
a duration of two hundred moves.
Th e winning strategy (or program), by a wide margin, was the simplest
one, which operated on the single principle of pure reciprocity— that is, to
be cooperative on the fi rst move, and then mechanically to replicate, in each
subsequent round, what ever the other side did on the previous play. Even
when that strategy was employed by only one of the players, it produced the
highest score for its user. When both sides employed it, the results were
spectacularly good— a fl awless record of mutual cooperation.
It should be noted that this happy outcome has nothing what ever to do
with morality or altruism. Each contestant had no goal except to obtain the
highest score for himself alone. But the optimal way of achieving this purely
self- interested aim is to settle into a pattern in which each side consistently