Week 7: Sourcing and Evidence

Three ways a documented case can fail the sourcing test without being obviously fake, and the persistent-plus-dormant reveal held back since week 2.

A striking story and a sourced one are not the same thing.

5 April 2027

Open the slides

Outline#

• three ways a case can fail sourcing without looking obviously fake: a procedure that sounds official but is corroborated by nothing at all, an account that is structurally impossible to corroborate by design, and a legend with no surviving physical evidence at any point in its history — three distinct failure modes, not three degrees of the same one
• case-format literacy: a text write-up (a sculpture that only moves unobserved) and a video/software claim (a secondhand game cartridge whose glitches escalate) look nothing alike on the page, but the same three sourcing questions from the practical apply to both
• the week-2 reveal: the persistent-plus-dormant case previewed then — affectionate and unchanging on the surface, and, rarely, while unobserved, doing something else entirely — gets its full write-up this week, now that you have the sourcing method to read it responsibly rather than just find it striking
• a case’s specific lesson for equipment: a secondhand cartridge whose glitches escalate after the “obvious fix” — wiping and resetting it — turns out to be the documented action that made the case worse. A countermeasure this course teaches you to reach for by default is not automatically safe against every anomaly
• Halloween candy tampering as the bridge to next week: a real, academically tracked legend with fewer than two dozen confirmed injuries across 26 years of press coverage, still driving real defensive behaviour every October — the clearest case this course has for treating a legend’s effect as data even when the legend’s content isn’t

Why this week has no fieldwork#

The phase has changed from on-campus observation to desk study. Nothing here was witnessed by this course directly; everything is read, sourced, and weighed against how well it survives that reading. A striking case that fails sourcing still gets discussed — it just doesn’t get counted.

After this lecture#

This method is what week 8’s global survey and week 9’s convergence analysis both depend on. A case that skips sourcing here doesn’t make it into either.