==============================  CFJ 1426  ==============================

    Absolving (removing) a Blot as per Rule 1986 is not the same as
    Expunging Blots as per Rule 2017.

========================================================================

Caller:                                 G.
Barred:                                 Taral

Judge:                                  Steve
Judgement:                              FALSE

========================================================================

History:

Called by G.:                           14 Jan 2003 00:07:43 GMT
Assigned to Steve:                      14 Jan 2003 00:57:09 GMT
Judged FALSE by Steve:                  16 Jan 2003 03:05:52 GMT

========================================================================

Caller's Arguments:

First of all, Goethe made two attempts to remove a number of
Blots from Eris at a time when Eris has more than zero but less
than 1 Blot.  Either attempt may have been successful, but not
both, for Goethe was only permitted one such successful action
with 0 Support in a given Agoran Week (R1986).

If CFJ #1 is FALSE, and Absolving and Expunging are the
same thing, then Goethe's first action of absolving 1 Blot
from Eris (also quoted in Evidence below) succeeded in
reducing Eris's Blots to zero, by the last paragraph of R2017.
If so, CFJ #2 is trivially FALSE, CFJ #3 is trivially TRUE,
CFJ #4 is trivially FALSE, and CFJ #5 is trivially FALSE.

However, I argue for a judgement of TRUE in CFJ #1.
Expunging is a very specific mechanism defined in Rule 2017,
that is only triggered (as explained in R2017's last paragraph)
when the Rules explicitly state that a Blot is Expunged.
Rule 1986 does not use the word Expunge, and the fact that
(remove) is placed in parentheses where (expunge) could be used
highlights that this is a different method of reducing Blots.

Further, the definition of Absolve (simply removing a Blot)
has a higher precedence than the definition of Expunging a Blot
(R1986 is Power-2 vs. R2017 at Power-1).  Thus CFJ #1 should
be judged TRUE, and Absolving is a distinct method of removing
a Blot which is more powerful than Expunging.  And in cases
where removing a Blot from a Player would result in negative
Blots, the rounding-to-zero method triggered by Expunging in
the last paragraph of R2017 is not directly triggered by a R1986
Absolution.

If so, and CFJ#1 is TRUE, what happened in Goethe's Absolution
attempts?

At the time of the attempts, Eris had 0.68 Blots.  R1986 states
that an Acolyte may, as a single action Absolove (remove) exactly
1 Blot from a Player, provided e has enough Support for the
number of times e performed the action.  Goethe, in eir
first attempted action, had the required support, and therefore
removed 1 Blot from Eris and Eris has less than zero Blots.

Now R2017 specifically states that "The Stain of each entity is
at all times a non-negative number."  However, R1986 (Power=2)
*HAS A HIGHER PRECEDENCE* then R2017 (Power=1).  Therefore, the
fact that an Acolyte with the proper Support may remove 1 Blot
from an entity has a higher precedence than the fact that Blots
must be non-negative.

Thus, the Caller strongly argues for this interpretation, from
which follows a judgement of TRUE for CFJ#2, FALSE for CFJ#3,
TRUE for CFJ #4, and TRUE for CFJ #5, as an Immaculate Player
is one whose Stain is exactly zero (R1962).  Under this
interpretation Goethe's second Absolution attempt, to expunge
exactly 0.68 Blots, trivially failed due to requiring more
than 0 Support.

If the judge believes that Absolution is not Expunging, but
that the non-negative nature of Blots in R2017 somehow has
precedence over the ability of Acolytes to Remove Blots in
R1986, then some more interpretation of Goethe's actions are
necessary.  It is possible that some portion of Goethe's
first attempt was successful, but this is unlikely, in just
the same manner that game precedence holds that an attempt
to transfer "20 Stems" when you've only got 10 does not
transfer 10, but rather fails entirely.  If the first attempt
thus failed, it is possible that the second attempt (specific
attempt to reduce Eris's Blots to zero) succeeded.  But this
is also unlikely: just like previous CFJs ruled that half a vote
couldn't be specifically cast, R1986 doesn't allow fractions
of Blots to be expunged.  It is more likely under this line
of reasoning that both attempts failed, and Eris still has
(positive) 0.68 Blots.

========================================================================

Caller's Evidence:

Goethe's attempts:
>   I announce my intent, as an Acolyte, to Absolve (remove) 1 Blot
>   from Goddess Eris, with 0 Support.  Having received the
>   required Support, I Absolve 1 Blot from Goddess Eris.
>
>  That might not have worked, so:
>
>   I announce my intent, as an Acolyte, to Absolve (remove) 0.68 Blots
>   from Goddess Eris, with 0 Support.  Having received the
>   required Support, I Absolve 0.68 Blots from Goddess Eris.

-Acolyte Goethe


Eris's Blots [Jan 09 Herald's Report]:
> Goddess Eris [...] 0.68


Rule 1986/5 (Power=2) [Excerpt]
Role-Based Powers

      [...]
      An Acolyte may absolve (remove) 1 Blot from any Entity by paying
      Fee in Indulgences, or receiving the Support of a number of
      other Acolytes, equal to the number of such Absolutions the
      Acolyte has previously performed in the current Agoran Week
      (minimum zero).
      [...]


Rule 2017/0 (Power=1)
Blots

      A entity's Stain is a measure of that entity's cleanliness,
      measured in Blots.  The Stain of each entity is at all times a
      non-negative number.

      If the Rules state that an entity gains, is assessed, or is
      penalised some number of Blots, then, as soon as possible after
      the Herald is informed of the change, e shall record an increase
      of that amount in the entity's Stain.  (If the entity was an
      Unready Player at the time of the change, then the Herald shall
      instead record an increase of half that amount.)

      If the Rules state that some number of a entity's Blots are
      expunged, then, as soon as possible after the Herald is informed
      of the change, e shall record a decrease of that amount in the
      entity's Stain.  (If this would result in a negative Stain, then
      it instead results in a Stain of zero.)


Rule 1962/3 (Power=1)
Immaculate Players

      An Immaculate Player is a Player whose Stain is zero.

      Not being Immaculate is a Win-Preventing Condition.

      Being the only Active Immaculate Player is a Win Condition. Upon
      the report of such a Win, the Herald shall expunge N Blots per
      Player, where N is the Stain of the Active Player(s) with the
      lowest non-zero Stain.

========================================================================

Judge Steve's Arguments:

R1986 refers to Blots being 'removed'; all other Rules (1664, 1988,
1505, 2018, 1962, and especially 2017) that talk about getting rid of
Blots (to use a neutral term), refer to Blots being 'expunged'. What is
the significance of this difference?

My judgement is that the difference is not significant, and that we
should regard 'remove' in R1986 simply as synonymous with 'expunge', in
line with R754. Any dictionary will confirm that the two terms are
synonymous. This seems to me to be the simplest and most straightforward
interpretation of the situation.

However, in a series of ingenious posts, Goethe has argued that the
difference is a very significant one, that what we have here are two
independent and distinct procedures for getting rid of Blots, Absolution
and Expunging, each complete in themselves.

Goethe gave six arguments against synonymity (see <1>). The most obvious
of these is just the difference in terminology itself. Why does R1986
use 'remove' when it might just as easily have used 'expunge'? But this
argument doesn't really lead anywhere. Of course it's possible to point
to cases where a difference in a single word marks an important
difference in procedure, e.g. 'pay' vs 'pay out'. But equally it's
possible to point to cases where the use of alternative terminology
makes no difference: 'pay' vs 'transfer', or 'enact' vs 'create'. The
general principle I bring to bear on such cases is anti-formalism: other
things being equal, we should not be fussy about precise forms of words
where that can be avoided. The burden of proof lies with the person who
wants to show that two otherwise similar expressions should be treated
as having different effects.

The key argument offerred by Goethe to try and meet this burden seems to
me to be his first:

| 1.  If you removed the definition of "expunge" from the Ruleset,
|     the phrase "absolve (remove)" is still a clear definition of
|     what absolution is, the removal of exactly one Blot.  This
|     suggests that it is an independent mechanism.

In subsequent posts, Goethe has clarified this argument: the use of
'remove' in R1986 is to be regarded as the definition of a complete,
independent and distinct procedure for getting rid of Blots, one that
could stand on its even if R2017 were repealed.

However, I think this key argument fails. R2017 specifically defines the
consequences of expunging Blots in terms of Recordkeeping responsibilities
incurred by the Herald. No separate procedure is defined for removing
Blots. The inference is a strong one, I think, that the consequences of
removing Blots are precisely those defined for expunging Blots in R2017.
The argument that removal is a complete, independent and distinct
procedure for getting rid of Blots collapses.

========================================================================