Machiavelli wrote:
Anyway, if self-reference is not the point, then what is? It seems
like the only thing that could result in a turtle would be a single
clause within a rule that contradicts itself.
Which has happened. Here's a rundown from my Hall of Fame list:
* CFJ 3087 (self-contradictory rule)
* CFJ 2878 (self-reference)
* CFJ 2650 (self-contradictory rule)
* CFJs 1980, 1982 (contract with infinite loop)
* CFJs 1883, 1884 (yes/no questions rather than true/false statements,
before the rules were patched to equate them "the answer to
<question> is yes")
* CFJ 1787 (self-reference)
* CFJs 1594, 1596 (self-contradictory judgements, one of which was
later interpreted as incorrect instead)
* G., cards, June 2005 (playing X to draw Y, then playing Y to
retroactively cancel playing X)
