Name the alibi as a property of the layout

[tschm]

Two sentences into the alibi paragraph in The Handover Problem, before "The checkerboard structure is partly a response to this."

The alibi is not a failure of character; it is a property of the layout. Each party can see only their own station, so each account of events is locally true and collectively useless.

Why

The paragraph is the most original argument in the document — it explains why the handover model persists, which the efficiency case cannot. But it describes the behaviour without saying what produces it. The mechanism is that each party sees only their own station, and that is worth naming, because it is what makes the alibi structural rather than a matter of individual bad faith. Nobody has to be evasive for accountability to diffuse; the layout does it for them.

Naming it also earns the sentence that follows. "The checkerboard structure is partly a response to this" currently answers a problem that has been shown but not diagnosed. With the diagnosis in place, the checkerboard is answering something specific: it removes the condition that made the alibi available.

What this replaces

This came out of a longer discussion about cross-communication between stations in a modern factory, and about what a team should do when it cannot escape the handover model. That material was heading towards a full subsection on running the factory well — stop authority, executable interfaces, upstream feedback channels, shared vocabulary, station rotation.

We dropped it. The document's argument is that the factory should be avoided, and a well-developed manual for running one is exactly what a reader under institutional constraints would take as the practical conclusion, at the expense of the argument it sits inside. The existing treatment is the right size: the mitigation list in The Handover Problem steelmans the position before rejecting it, and the Toyota paragraph in A New Direction concedes the minimum to readers who have no choice. Both are immediately subordinated, and neither needs expanding.

What survived is this: the insight that lateral visibility is what the handover destroys, pointed back at the checkerboard, where it supports the thesis instead of qualifying it.

Base branch

Stacked on #4, which is stacked on #3. Merge in order and GitHub retargets each automatically.

Verification

Built with CI's command and flags (xelatex -halt-on-error -interaction=nonstopmode, twice): compiles clean, 15 pages, unchanged from #4. Checked the rendered page to confirm the sentences read correctly in the paragraph rather than only that they compiled.

🤖 Generated with Claude Code

[Copilot]

Pull request overview

This PR refines the argument in The Handover Problem by explicitly naming the “alibi” as a structural consequence of the handover layout, strengthening the causal link to the checkerboard mitigation.

Changes:

• Inserts two sentences that explain the mechanism behind the “alibi” (local visibility → locally true but collectively useless accounts).
• Reflows the surrounding paragraph so the “checkerboard structure” sentence now follows an explicit diagnosis.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.