↩ ExploRational · A Catalogue of Divergent Rationality
explorational · a divergent move
“A tomato is a fruit.” True — for a botanist. The United States Supreme Court ruled it a vegetable, and both were doing their jobs well. Meet the absolute claim by asking in which sense, for which purpose, under which standard it holds. The question neither swallows the claim nor dismisses it; it indexes it.
In 1893 the Supreme Court decided Nix v. Hedden: tomatoes, botanically fruit, would be vegetables under the Tariff Act, because tariff law tracks how things are eaten, not how they reproduce. Nobody discovered a fact about tomatoes that day. The Court chose which sense of “fruit” the customs house would use, and gave a reason anchored in the purpose at hand. The Court performed the move entire: a category question answered by asking what the category is for.
David Chapman's writing on meta-rationality puts the eggplant through the same door — fruit to the botanist, not to the grocer — and draws the general lesson: “does this count as an X?” usually has no context-free answer, and does not need one. Categories are equipment. They are built for purposes, held in place by contexts, and open to renegotiation when circumstances change — which is why standards bodies, courts, and style guides exist, and why they keep meeting. J. L. Austin made the same point about truth itself: “France is hexagonal” is true enough for a general planning a campaign and false for a cartographer, and there is no purpose-free tribunal to settle which of them owns the word.
The move has a meta-epistemic ceiling, and it is worth using deliberately: ask the question about truth. A courtroom means proof beyond reasonable doubt; an engineering review means within tolerance under the stated loads; a journal means survived the field's current tests. Before an argument starts, the parties can ask what “true” will mean here, and why that standard suits this occasion — a choice that classical treatments of rationality assume has already been made somewhere else. Nobody is exempt from giving reasons at this level; the standard is chosen, and the choosing can be done well or badly.
the signature instrument
Pick a claim, then move it from bench to bench. Watch three things change: the sense the key word takes, the verdict, and the standard that settles it. Nothing about the world changes between clicks — only the purpose the category serves.
The indexing bench
qualitative — the instrument computes nothing; the shift happens in the reader
the claim, as stated
Choose a bench. The claim will not change; its sense will.
what to try
1 · Index a live dispute. Take a disagreement you are currently in. Write each side's claim, then add “— in the sense that …” to both. If the completed sentences are compatible, the dispute was two purposes sharing one word; what remains after indexing is the disagreement worth having.
2 · Ask the truth question out loud. At the start of the next contested discussion, ask: what would count as settling this, for us, today? A number? A precedent? A demonstration? Agreement on the standard first is the cheapest de-escalation on record.
3 · Renegotiate one category. Find a boundary case your team keeps stumbling on — is a config change a “deploy”? is a reorg a “layoff”? Convene the five-minute meeting that decides what it counts as, for which purposes, and write the ruling down with its reason, Nix-style.
the mapping
| in the instrument | at your desk |
|---|---|
| the claim | the absolute sentence somebody just said with total confidence |
| the bench | the purpose-context the word is working for right now |
| the sense | what the key word means once the purpose is named |
| the standard | what “settled” means on this occasion — the meta-epistemic choice |
| the residue | whatever still conflicts after indexing: the live dispute |
where it tears
risk · flattening
The move indexes claims to purposes; it does not dissolve them into preference. Within a named sense and standard, the tomato question has a right answer, and so did Nix. If asking the question leaves everything equally sayable, it was performed as evasion, and the flattening the catalogue warns about has occurred.
risk · the palter door
“In what sense did I promise?” is this move run in reverse, as escape. The question is for opening categories that were closing inquiry — not for reopening commitments that inquiry had already closed. SocioRhetoric files that use under palter, and so should you.
risk · infinite regress
What counts as a good reason for choosing this meaning of “true”? — and so on down. In practice the regress bottoms out in a shared purpose good enough for the occasion; demanding a final floor is how the move becomes a filibuster. Take the floor the occasion offers, and note that you took it.
The claim stays; the sense moves. Index it to its purpose, and keep only the conflict that survives.
can you use it?
RECOGNITION — Which is the move? A: “That's just semantics.” B: “Fruit for the customs form, or fruit for the greenhouse? The form, I think — so: vegetable, and here's why the form's purpose governs.” C: “Both sides have a point.”
B. It names the rival senses, picks one by the purpose in play, and gives the reason. A dismisses the category question; C flattens it — neither indexes anything.
THE NEAREST NEIGHBOR — Moving the point at issue also redirects an argument. What separates asking “in what sense?” from it?
Level. Moving the point at issue changes which question the argument is about. Asking “in what sense?” keeps the question and splits its key term, so both parties may turn out to be right about different senses — and the standard for settling it becomes discussable.
PRODUCTION — Your team is arguing about whether last night's incident “counts as an outage.” Run the move on paper: two senses, their purposes, and the standard you'd propose.
Sense one: customer-facing unavailability (purpose: the SLA and the postmortem obligation). Sense two: any unplanned degradation (purpose: internal learning and the anomaly ledger). Proposed standard: the SLA sense decides the label and the customer email; the ledger sense decides whether a review happens. Yours works if each sense carries a purpose and the standard says which purpose governs which decision.