Week 4: Week 4 lecture: What 'stacking' actually means

Why no single document proves a decision was pre-made, and how a small set of documents together can still raise the burden of proof — read through two new documents from the pilot's own record.

15 March 2027

One document is rarely enough#

Week 3 asked you to read a single agenda for what it argued. This week’s question is different: what does it take for several documents, none of them individually damning, to add up to something a single document never could? The honest answer has a ceiling built into it. A pattern across documents raises the burden the “genuinely open” reading has to meet — it does not retire that reading. Treating a stack of circumstantial signals as equivalent to a confession is the single most common overreach this course marks down, on either side of the argument.

The story continues: a working session before the working group#

The Steering Committee confirmed the ClarityPulse implementation timeline on 8 March. What happened in the week after is new material — two documents that were not part of Week 1’s fact/inference exercise, not because they hadn’t happened yet (the whole pilot is a closed, years-old case; every document in it already existed before you read any of them), but because Week 1 released only the ten entries it needed for that week’s exercise. These two were not previously released to students, and are newly added to this week’s evidence packet: a calendar invite, sent 10 March, for a small working session between Tomas Herrera, Priya Nandakumar and ClarityPulse’s delivery team, and the agenda for the 13 March LSS Rollout Working Group the invite refers to as “Saturday.” Both documents are reproduced in full on the session page — read them there before continuing, since the reading below assumes you have.

What stacks here, and what doesn’t#

Read alone, the calendar invite is unremarkable — technical delivery planning ahead of a large meeting is a completely normal reason for a small group to talk first, and the course does not ask you to treat every pre-meeting conversation as suspicious. Read alone, five minutes for an agenda item with an attachment already prepared is also unremarkable — plenty of agenda items are genuinely simple. What starts to stack is the combination: the small group’s stated purpose is explicitly “so we’re not designing it live,” the attached sequence in the 13 March agenda is not described as a proposal open to reordering, and no version of the sequence that differs from the working session’s outcome appears anywhere in the record. None of that proves the 40-person Working Group’s discussion was theatre. It does mean that if you wanted to argue it wasn’t, you now have to point to something in the room on 13 March that the documents alone don’t show — which is exactly the gap the honest answer has to name.

The honest counter-case, stated plainly#

A technical sequence — which team goes first — is a reasonable thing for a delivery lead and a sponsor to work out before spending a large meeting’s time on it, provided the sequence was the only thing pre-narrowed and the Working Group still had a real chance to weigh in on the parts the small group didn’t already decide. Whether that’s what actually happened here is not settled by these two documents alone; Week 6’s Decision Autopsy is where you’ll be asked to make and defend a call like this on a real decision of your own choosing.

A little further reading#

The linked entry is a general legal-dictionary definition, not a source about this pilot or about meetings at all, and it states less than it might seem to promise: it defines circumstantial evidence as indirect evidence — evidence that requires an additional reasonable inference to reach the fact in question — and illustrates that with one example. It does not itself say that circumstantial evidence can support a conclusion beyond reasonable doubt, that a larger stack of it is not automatically stronger than a smaller one, or that each piece must independently exclude the innocent explanation; none of those three claims appear on the page, however natural they might sound next to the word “circumstantial.” Read it for the one distinction it does state plainly — direct evidence versus evidence that needs an inference to connect it to the conclusion — and treat everything else here as this course’s own extension, not the source’s: “stacking,” and the standard that a pattern of documents raises the burden of proof without retiring the “genuinely open” reading, are this week’s own argument, built on top of that one definition, not a claim the Wex entry itself makes. Bring the course’s own standard, not a borrowed authority, to how you weigh the calendar invite and the agenda above.

Before Week 5#

Nothing new to bring. Week 5 moves forward to a later Working Group meeting, on 22 March, and a new document — a risk-register addendum, not the 13 March agenda above — and asks a different question of it: not who met beforehand, but who controlled when the paperwork around a decision actually arrived.