{"path":"research/catching-a-proposal-made-unnecessary/cases-negatives/AX4.md","content":"*Spoken name: a long leap folds into a route.* *About: How the map shows two claims many links apart: the links between them folded into one marked route, never into a link.* **Two claims many links apart can be shown with the route between them folded into one band on their line. The fold is a view: never stored, never a link, and it shows no total of its own. Along the band it shows each folded step's strength, one dot per folded inference up to a cap, with the weakest step marked (a yellow dot, or that stretch of the band at half opacity); a hover gives the number of links, how many have nothing written under them or are contested, and the claims it shares with sibling routes; it opens a level at a time; and it stays whole when an objection or a person's input attaches inside it, since an objection shows through its step's lowered strength.** *Reworded 2026-10-26 from his direction of 24 October (`folding-joining-and-zoom.md` § 8); the first wording, kept in git, also split the band where a side attacker attached, which he declined, and gave as its reason that support does not carry along a chain, where in Deliberus each link states its relevance given its middle claim (`folded-routes-and-shared-premises.md` § 8 item 2). No total, because a route's share of the end claim's posterior depends on how it overlaps the other routes, which the record does not fix (Q48).*  2,003 pairs of claims sit six or more support links apart today; the longest route (eleven links) runs from instrumental drives to a report of what some people believe about AI descendants, and its two ends barely bear on each other (§ 5). The first part is built on his word of 2026-10-24, *\"Great, go ahead\"*: on the wheel a second-ring claim that only relays one support (one in, one out, no other link) folds, its supporter in its place and three dots with the count on the link (`wheelMap.ts`, spec § 3a); the row as a whole stays unruled. His position the same night against the last clause: *\"The zoom/fold should not be broken up/made messy just because it has human input at some point in the chain/subgraph.\"* and *\"It should visually indicate the strength/truthiness flowing through the folded inference chain.\"* (`folding-joining-and-zoom.md` § 8). **Meets:** applies AX8 (ruled: the map folds along a chain and shows the weakest links); applies Q47 (ruled: the weakest link is what a fold shows a reader, never the arithmetic) and Q48 (ruled: an unrecorded dependence shows its range, so a route shows no single total); the inverse of the chain folded inside a link (AG14, ruled; `the-chain-inside-a-link.md` § 4.5 on opening levels); builds on AG1 (open). Reverses nothing settled. **Outside reader:** depends on AG14 (\"a hover gives the number of links, how many have nothing written under them or are contested\" against \"What a machine writes under a link is written as the smallest steps, and a step counts as smallest by two tests\"); extends AX8 (\"Along the band it shows each folded step's strength, one dot per folded inference up to a cap, with the weakest step marked\" against \"the map, the truth graph and every other surface zoom both ways: sideways... and along a chain, links folded into a route or opened into their steps\"); extends AX9 (\"it opens a level at a time\" against \"A click opens the members as their own claims, refocusing the map on them where room is short\"); extends AX9 (\"No total, because a route's share of the end claim's posterior depends on how it overlaps the other routes, which the record does not fix\" against \"the card has no strength of its own, each member keeping its own fill and band\")."}