He describes exactly how he slept in Barbados in imagination — placing New York northwest of him by latitude and longitude, hearing the tropical noises, seeing the sudden red-to-green Barbadian sunset, and feeling his parents and unmarried brothers in the big family house.
Date (happened)
c. October 1933
unresolved
c. October 1933
stated · preferred
Why
Dates the imaginal practice itself, not the later lecture.
Reasoning
The narrative establishes 1933 at its opening and passage 549255 explicitly locates the just-described Barbados sleeping practice in late October, supporting month precision but not a day.
before 1969
bound · possible
Reasoning
Qualifying passage ids [549254] belong to telling lecture nids [1928]. Each listed lecture contributes the selected safe envelope year 1969 (MAX of its non-refuted delivery years); 1969 is the MIN across telling lectures.
before 1971
bound · possible
Why
Corrects the phase4 L6a envelope on this event from before 1969 to before 1971.
Reasoning
Envelope recompute, arithmetic only. The phase4 L6a row on this event states its own rule: each telling lecture contributes MAX of its non-refuted delivery years, and the envelope is the MIN across telling lectures. Recomputed live over lecture nid(s) 1928 that row cites, MAX(non-refuted field_delivery_year) gives 1971, not the 1969 the row carries — the bound was below its own stated rule and therefore UNSAFE: it asserted the event did not happen in 2 year(s) where it may sit. Same premise, same evidence, no re-reading and no re-judgement; only the maximum was taken correctly. The L6a row stands beside this one (append-only). Ruling: #2235 delta 1, architect 378.
Told / reported
4 June 1971
estimated · possible
Why
Lane L7 (#2354) re-derivation. Supersedes in substance, not in storage, the refuted-date told claim(s) on this event; those rows stand per append-only.
Reasoning
Re-derived, not walked down. The existing told claim(s) on this event took their date from a delivery claim on node 1928 that the corpus has since REFUTED, so the corpus was asserting that Neville told this on a day it separately says he did not lecture. That is a contradiction rather than an over-confidence, and lowering the confidence would have hidden it instead of resolving it. The refutation stands and the told claim is the wrong side of it. Node 1928 surviving (non-refuted) claim: 1971-6-4 at `possible`. All claims on the node: 1969-5-26 refuted; 1971-6-4 possible (preferred). Superseded here: delta 0: 1969-05-26 at `certain` (run `corpus v1 / phase3b`). The telling is dated by its lecture, so with the refuted date removed from consideration the surviving claim is what the telling inherits, at that claim own confidence and no higher.
26 May 1969
stated · refuted
Reasoning
|| v1.1/l7/told-close: the delivery claim this was derived from (nid 1928, 1969-5-26) was refuted by the L7 adjudication (#2363). The corrected told claim from the same lecture stands at 1971-6-4. This row is retired rather than deleted so the derivation history remains readable.