Index ← 1729 CFJ 1730 1731 → text
==============================  CFJ 1730  ==============================

    The player Goddess Eris assigned judgements to the CFJs 1716, 1710,
    1711, 1724, and 1725.


Caller:                                 Pavitra
Barred:                                 Taral

Judge:                                  BobTHJ

Judge:                                  Zefram
Judgement:                              TRUE



Called by Pavitra:                      22 Aug 2007 16:54:10 GMT
Assigned to BobTHJ:                     22 Aug 2007 22:31:24 GMT
BobTHJ recused:                         30 Aug 2007 10:31:07 GMT
Assigned to Zefram:                     30 Aug 2007 10:31:07 GMT
Judged TRUE by Zefram:                  30 Aug 2007 11:41:21 GMT


Caller's Arguments:

Upon the adoption of proposal 5140, Eris became a knave,
and henceforth SHOULD NOT have published statements that e believed to
be true. Eris being an experienced player, it is presumed that e would
not have taken any action that the rules specified e SHOULD NOT have,
without also providing some well-reasoned explanation as to why e
chose to break with convention. E provided no acknowledgement of any
kind that e was undertaking a DISCOURAGED action.
It therefore stands to reason that eir published declarative sentences
since the adoption of proposal 5140, such as "I judge UNDETERMINED for
CFJs 1724 and 1725", were intended as false. Thus, Eris in fact did
not so judge, and the CFJs upon which e falsely claimed to assign
judgement are still open.


Judge Zefram's Arguments:

Assigning judgement is an action that is taken by announcement (R2158).
Rule 478 defines "by announcement" thus:

      A player performs an action "by announcement" by announcing that
      e performs it.

Eris's statements in which e purported to assign judgements to the listed
CFJs certainly look like such announcements.  For CFJ 1716:

|The defendant has made an excellent set of arguments for a judgement
|of UNIMPUGNED. I so judge.

For CFJ 1710:

|However, the sums in question do not include Wooble's votes. I find,
|therefore, that the first notice was invalid, and judge CFJ 1710

For CFJ 1711:

|      Secondly, the fact that the message does not quote or even
|include a date/message-ID for the message it is correcting runs
|significantly counter to the rule's use of "to clearly identify". I
|find, therefore, that the second "notice" was also invalid, and judge
|CFJ 1711 FALSE.

For CFJs 1724 and 1725:

|I judge UNDETERMINED for CFJs 1724 and 1725.

If Eris's statements are taken at face value then e certainly did assign
judgements to these CFJs.  The only way e could be taken to have not
done so is by the initiator's argument that these statements are not to
be interpreted in the ordinary way.  Initiator Pavitra's argument can
be read two ways.  I will analyse both.

The more literal reading of Pavitra's argument is that Eris intended to
make false statements, and that this intent robs eir announcements of any
legal effect.  There is an unsupported leap in logic there.  The rules
ascribe very little significance to the intent behind public messages.
Rule 2149 speaks of belief, which is somewhat akin to intent, in making
obligations and recommendations regarding truthfulness, but that is not
the issue here.  Rule 478 is the critical one, and the definition of
"by announcement" makes absolutely no mention of intent.  An appropriate
announcement will accomplish an action even if its author believed the
action was impossible (in which case e intended that the announcement
would be a false statement).  So the intent argument does not change
the analysis of the announcements.

A related argument, which I suspect is what Pavitra really intended to
put forth, is that we must interpret all of Eris's statements through
a filter that inverts truth values.  We have a history of somewhat
flexible interpretation of public messages, but this flexibility extends
to implicitudes and variations in formatting and phrasing; we have no
history of interpreting messages contrary to their explicit content.
Of course, the present version of R2149 is quite new, so there's the
possibility of it introducing new interpretive conventions.

Hypothetically, how might we establish such a counter-literality filter?
A rule could establish whatever interpretive conventions we desire,
subject to rights issues, but we have no such rule.  Common usage
could establish counter-literal interpretation of some phrases, but such
interpretive conventions have always been applied to all persons equally.
Here we require a filter that is applied only to a particular subclass of
player (knaves).  We have no history of such differential interpretation.

Perhaps we could establish a novel and discriminatory interpretive
convention by explicit inter-player agreement.  It is likely that some
players would object, preventing unanimous consent or even consensus,
rendering the status of such a meta-game agreement dubious.  In fact there
has been no explicit agreement of this nature.  Eris, in particular, has
not proposed such a convention.  So the argument for counter-literality
requires that this controversial, inherently confusing, convention of
utterly novel form is established implicitly.  I find that this is too
great a stretch: there cannot be such an agreement.

I will also consider how such a counter-literality filter would operate,
if it were established.  Suppose that it simply inverts the truth value of
every statement made by a knave.  When a knave says "I judge UNDETERMINED
for CFJs 1724 and 1725.", it is argued that we interpret it as really
meaning "I do not judge UNDETERMINED for CFJs 1724 and 1725.", a statement
that is not an announcement of action and so does perform an action.

Of course, it's the rules' interpretation of the message that matters,
not ours, so to be more precise the argument is that rule 478 ("by
announcement") sees the message through this filter and determines that
this is not an action announcement.  But then, rule 2149 should see the
message through the same filter.  It would see a statement of non-action,
which (due to rule 478 denying any action) is a true statement, and so the
message offends R2149's SHOULD NOT.  So this filter, so applied, does not
achieve its intended goal of letting a knave avoid a discouraged action.

For the filter to work as argued, it would have to affect R478 and
R2149 differently.  A statement would have to be inverted for its
interpretation by R478 (for actions) and other rules requiring publication
of information, but then seen by R2149 in its uninverted form.  There is
no basis in the rules for a statement to be interpreted differently by
different rules.  Also, if we were to have such a selective inversion,
it would be more convenient to do it the other way round: invert the
message for R2149, but not for the remaining rules.  That would amount
to interpreting R2149's knave provision in the opposite sense in which
it is written, nullifying its effect.  I see no basis to suppose that
this selectivity is possible in any form.

So, in summary: Eris facially did assign judgements to these CFJs; eir
intent does not affect the interpretation of those statements; there is no
counter-literality filter affecting the interpretation of Eris's messages;
and a filter that would do the job proposed cannot exist in any case.
I judge TRUE.