The Gnarlier Contract does create a paradox that, were it in the
rules, would definitely deserve a Win by Paradox. The issue is
whether the contract is, in fact, allowed to state something like:
6. Whenever a party's gnarlierness becomes twistier (including, but
not limited to, being flipped to twistier), it is flipped to mixier.
and have that actually happen.
Contracts certainly cannot take infinite game actions, by CFJ 1584.
But flipping some internal gamestate is not an game action.
By Justice Pavrita's arguments in the appeal of CFJ 1936, contracts
are at heart a primordial thing that, while perhaps governed by the
rules, are at some lowest level external to them. The Rules could
perhaps restrict the type of gamestates contracts can keep (although I
don't support this), but without any specification of this, it should
be assumed that contracts can keep whatever vague or ambiguous
gamestate they like, just like a foreign nomic. (Indeed, we have had
instances of foreign nomics proposing rules which make them Agoran
contracts.)
I propose that, for the good of the game, contracts should keep that
right (which I judge they do presently have) to have paradoxical or
ambiguous gamestates. Any Rule that depends upon a contract's
gamestate (e.g. the Pledges rule, or allowing partnerships to take
actions) should require that the gamestate be unambiguously in
whatever state is required (i.e. claiming to be a pledge). They
presently do not, "escalating" a contract ambiguity into a Rules
ambiguity. It's not just that these paradoxes shouldn't become
tortoises; they shouldn't be allowed to create a Rules ambiguity at
all.
Since tortoises have to remain such for two weeks, it is worth noting
that (if passed) a prompt proposal could eliminate any Win by Paradox
resulting from this case.
Now, I don't agree with Judge Pavrita's quantum theory from eir
judgement of CFJ 1936. I don't think contracts should be allowed to
force us to make such complicated and arbitrary interpretations of
such concepts as time. Instead, I think the Gnarlier Contract's
switches are simply in an undefined state. This makes UNDECIDABLE a
clear-cut appropriate judgement.
But from Rule 591:
* UNDETERMINED, appropriate if the statement is nonsensical or
too vague, or if the information available to the judge is
insufficient to determine which of the FALSE, TRUE, and
UNDECIDABLE judgements is appropriate; however, uncertainty as
to how to interpret or apply the rules cannot constitute
insufficiency of information for this purpose
I am uncertain as to how to interpret and apply the Gnarlier Contract.
The statement about uncertainty as to how to interpret or apply the
rules implies that uncertainty as to how to interpret and apply
*something* COULD potentially be an insufficiency of information worth
an UNDETERMINED judgement, as long as it is not the rules. The
Gnarlier Contract is not the rules, so UNDETERMINED is also
appropriate.
UNDECIDABLE or UNDETERMINED?
The fact is that, as a bug in the Rules that must be patched with a
Proposal, the ability of contracts to engineer Rules paradoxes is a
valid scam. The scamsters should not be deprived of their Win by
Paradox.
Therefore I (proto-)judge CFJ 1980 UNDECIDABLE.
P.S. If there are already proposals/protos related to this, you'll
excuse me: I haven't read all of that.
