Open the live control panel | Research manual
(Kinds of second-order theme I expect to find, from the stage A1 notes. Dated.)
exemplar rule: (how you will choose among quotes that all pass a theme's gate, written before the
first theme is opened; the default is the line that meets the gate most directly, with the
passing quotes not chosen named in the exemplar's rivals. Stage B1 §3.)
- Continuity at the top, drops at the bottom, a model all the way down; the lower levels will
describe decline and continuity. (The operator's expectation, recorded here so stage E1 can
say whether it held.)
- (The analyst's own, from the reading.)
- Local adequacy: each analysed room receives a model plus validated adaptations accounting for
its evidenced organizing; unresolved claims and insufficient evidence are reported separately.
- Shared adequacy: clusters must explain their members jointly through a stated condition or
mechanism. Define what counts as a distinct shared adaptation and what would show that the
global model requires unbounded revision rather than limited branching. State the grounds and
denominators before seeing returns; no numerical cutoff is imposed by the instrument.
- Report the fixed-global verdicts (does-not-compose included), particulars, and corrections
met twice as diagnostics. Repeated corrections are candidates for clusters, never failures.
Count ledger lines and distinct shared changes separately. Report both pre- and post-adaptation
fit. Do not call exact local fit evidence of global composition by itself.
- What each level's verdict will rest on (contract §1, two kinds of claim): for the population
pattern, the forms and the particular histories in turn, whether the claim rests on the words
of the record or on the shape of the corpus, what evidence of that kind would support it and
what would count against it. Written before any claim exists, so that a later validation can
be read against the kind declared here.
- Composition fails where adequately observed cases require only unrelated particulars, shared
branches do not explain their members, or global claims remain incompatible with validated
local evidence without a defensible scope decision. An insufficient record yields uncertainty,
not demonstrated failure. Compare this decision with Kozlowski and Klein's separate distinction
between composition and compilation; neither term is a success grade.
Declare brackets from documented 2017 event times, with sources and timestamps. Distinguish
placement time from aftermath. Do not assume canvas expansions from another edition of r/place.
If only opening and placement closure are documented, use those anchors. Additional brackets
need an independent event record or a declared analytical rule, never the findings.
organizing_end_hour)A coded turn counts as coordinating when its first-order concept is an act toward the canvas or
toward other members' acts on the canvas: a call, an assignment, a plan, a report of placement,
a defence, an alliance, a correction of another's placement. Recounting, joking about, or
mourning the event is not coordinating. (The analyst refines the wording from the stage A1
lexicon before B1 opens. After that it changes only by a logged departure, and every figure
computed under it is then reported twice, under the rule before and after the change.)
(The clock contrast on onset_first_code_hour, with the members of each side named.)
w = 4, 2, 1, 0.5 for a meeting at third-order, fourth-order, dimension, bare target; g = 1, 0.5,
0, -1 for same op, same kind, different kind, the two opposite pairs.
- Concept floor prompting review: (number) first-order concepts.
- Off-platform pointer count prompting review: (number).
- What evidence is needed to decide the organizing claims, and which gaps or conflicts leave them undecidable.
The thresholds do not force a verdict. A well-observed room may remain decidable despite many off-platform mentions, and a rich record may lack one decisive event. State the evidence limit for each cannot-tell verdict.