status: https://faculty.washington.edu/kerim/nomic/cases/#4010
(This document is informational only and contains no game actions).
=============================== CFJ 4010 ===============================
The mentioned replacement in proposal 8898 was effectively
applied.
==========================================================================
Caller: snail
Judge: ais523
Judgement: FALSE
==========================================================================
History:
Called by snail: 19 Feb 2023 05:12:47
Assigned to ais523: 21 Feb 2023 01:05:11
Judged FALSE by ais523: 27 Feb 2023 17:11:13
==========================================================================
Caller's Evidence:
> On Feb 18, 2023, at 10:40 PM, Janet Cobb via agora-official wrote:
>
> In proposal 8898:
>
>> Amend R2675 (Dream of Wandering) by replacing the paragraph that starts
>> with "- Gardens" with:
>>
>> - Gardens: Immediately after a wandering, the Base Rockiness of each
>> Gardens Dreamer is increased by 1.
>
>
> The text to replace is a list item, not a paragraph. I am interpreting
> this as not being applied.
>
> --
> Janet Cobb
>
> Assessor, Mad Engineer, Rulekeepor, Stonemason
Caller's Arguments:
Light arguments for TRUE: a list item is a type of paragraph, and there's
nothing else the term "paragraph that starts with '- Gardens'" could refer
to.
--------------------------------------------------------------------------
Gratuitous Arguments by Jason:
I believe we've generally held that list items are either parts of
paragraphs or a separate thing, rather than being paragraphs themselves
(even if formatted as such). Also, "it can't possibly mean anything
else" isn't enough, given that we've rejected things like "Amend Rule
NNNN/R", even though that couldn't possibly mean anything else.
--------------------------------------------------------------------------
Gratuitous Arguments by G.:
CFJ 3778 found that list items could have whole line breaks inserted
between them and removed because they were not significant. This is
not true with paragraphs. If the section of text with ' - Gardens" is
taken to begin a paragraph, and is followed by additional list items
where the whitespace could be removed, the replaced paragraph would
include all of those line items. Or at least it is unclear where the
paragraph ends.
Additional CFJ to consider - In CFJ 3451, Judge Tiger explicitly
finds that the R105 standard is even stricter than "everyone knows
what was meant and there's only one reasonable interpretation" (maybe
R105 should be weakened a tad, but that would be a legislative not
judicial correction):
Judge Tiger wrote:
> By the natural-language interpretation of ambiguity alone, I would deem
> that using the wrong ID, which is not used by any other rule, and the
> correct title, is clear enough. Ambiguity implies multiple potential
> meanings. There is no sensible way for the "incorrect" number 2455 to be
> an indication of another meaning, or an attempt at obfuscation of the
> proposal's effect, when there is no rule 2455 but there is a 2445. I
> deem this case to not fall under the precedent of CFJ 1625.
>
> However, the second sentence of the paragraph specifies that any
> variation, other than "inconsequential variation in the quotation of a
> rule", constitutes ambiguity. I take this to be an extra safety measure:
> even if, as in this case, the intended meaning was never really
> ambiguous to us (to me), the rules require a higher standard of clarity.
https://faculty.washington.edu/kerim/nomic/cases/?3451
https://faculty.washington.edu/kerim/nomic/cases/?3778
--------------------------------------------------------------------------
Additional discussion in DIS:
snail wrote:
> Looking at this CFJ (3778), it seems to say the opposite about line
> breaks within a paragraph:
>
> CFJ 3452 ruled that paragraph boundaries should be determined based mainly
> on grammatical structure rather than layout. Following its reasoning, "A"
> above would all be considered a single paragraph, since it's a single
> grammatical sentence; therefore, there are no "paragraph breaks" to
> contend with and the changes *[inserting whole line breaks within a
> paragraph]* are definitely insignificant.
>
> Grammatically, each list item looks to be its own paragraph. The list
> items following the "- Gardens" list item are not able to have all of
> their whitespace removed, as this would contradict CFJ 3778: "[there
> is] a prohibition on merging or splitting paragraphs".
>
> If there was any ambiguity of whether the list items are all part of
> one paragraph, or each their own paragraph, the proposal resolves that
> ambiguity by referring to one of the list items as a paragraph.
G. wrote:
> Interestingly, this argument had the opposite effect on me. Before
> you said it, I thought "we've all agreed these are list items not
> paragraphs, the question is whether a proposal mistakenly referring to
> an otherwise clearly-specified unit as a "paragraph" breaks things.
> Now I'm thinking it's ambiguous whether they are paragraphs or list
> items to begin with (that is, more ambiguity not less) and I don't
> think a proposal has the ability to clarify that just by assertion.
Janet wrote:
> CFJ 3910 found that text has a rich structure. If a proposal could
> change the structure of a text merely by calling it something, that
> could be an amendment (and thus prohibited unless it met the R105
> communication standard).
--------------------------------------------------------------------------
Judge ais523's Arguments:
This CFJ basically comes down to whether rule 2675 contained "[a]
paragraph that starts with '- Gardens'". If it doesn't, then the
instruction in the proposal is impossible to follow and therefore its
replacement couldn't have been required.
It's clear that the rule in question contained a *list item* that
started with "- Gardens", so the question basically boils down to
whether a list item is a paragraph. This is a distinction that has been
causing trouble for rich text engines / markup languages for some time;
the issue is primarily that there are some situations in English for
which a list item clearly isn't a paragraph (such as when a single
sentence continues across multiple paragraphs). There are also
situations in English where list items need to be considered paragraphs
for consistency (it is possible to have a list for which some of the
items consist of multiple paragraphs; in this case, the remaining items
would typically be considered to be single paragraphs). It's worth
noting that markup languages normally make explicit distinctions
between the two cases, and even render them differently; for example,
the CommonMark specification (https://spec.commonmark.org/0.30/*loose)
specifically defines "loose" lists where the elements contain
paragraphs versus "tight" lists that don't, and HTML allows for
different markup in the two cases (see the attached demo) – this is
evidence that these are real distinctions that actually exist in the
natural-language English, because markup languages would be unlikely to
go to the trouble of trying to capture this sort of distinction unless
doing so were required to perform their job of reflecting the structure
of written text.
The list in question has no clues that would help us distinguish
between the two cases here (other than the double-spacing between the
list elements, which suggests paragraphing but is not a sure indicator,
and might or might not be relevant under rule 2429). Instead of
resolving this problem immediately, I consider a subtle distinction
that exists in the "the list contains paragraphs" interpretation,
between "each list item is a paragraph" and "each list item contains a
single paragraph". This makes a difference because in addition to text,
the list contains bullet points (here a "-" marker) that mark the list
items, and the distinction affects whether the bullet point is part of
the paragraph or not; in the latter case, the bullet point serves to
mark where the list item starts, and the paragraph starts after that.
And the latter case seems to make a lot more sense here: if a list item
contained two paragraphs, the bullet point clearly wouldn't be part of
the second, so it would be inconsistent to consider it part of the
first (except if ruling paragraphs to be defined entirely by the line
breaks within the rule's text and ignoring its structure, which would
be inconsistent with previous precedent, e.g. CFJs 3452 and 3778).
This means that for the purpose of this CFJ, it doesn't matter whether
the list is considered to contain paragraphs or not; if the list item
in question contains a paragraph, then that paragraph starts with
"Gardens" (as opposed to starting with "- Gardens"). If the list is
considered to be a paragraph on its own, it starts with "- Wandering"
(or "The following is" if it's part of the same paragraph as its
introductory text). So in neither case does the proposal successfully
identify the text to replace.
One other point that needs considering is: is this an "inconsequential
variation in the quotation of an existing rule" (i.e. can we consider
the replacement as replacing a paragraph starting with "Gardens",
misquoted as "- Gardens")? If it is, it would be a special case which
rule 105 does not consider to be an ambiguity. However, in the case
that this CFJ is considering, the variation does not seem to be
sufficiently inconsequential: not only is there ambiguity around
whether the list is formed of paragraphs or not (which is not a
variation in quotation), it would affect whether the post-replacement
list item started with one bullet point or two (because even if rule
105 considers the bullet point in the *text to replace* to be
inconsequential, no such allowance is made for the *replacement text*,
so the first bullet point in the proposal gets ignored but the second
one doesn't; a difference between the expected and actual results of
the proposal that defies common sense to this extent would be
considerably consequential).
I judge CFJ 4010 FALSE.
Judge ais523's Evidence:
[Arbitor's Note: the "attached demo" mentioned in arguments can be found
at https://www.mail-archive.com/agora-business@agoranomic.org/msg44084.html
It is left out of this archive record due to html sanitizing.]
==========================================================================
