ahepi/deepreason/precedent-transport

Precedent Transport (new in 0.5, from FR-24)

<!-- PROMPT-CORE-BEGIN --> An argument from precedent is a claim that a rule transports from A to B. It is valid only if what transports is the invariant the rule PROTECTS - not a property A happens to have.

1. Name the precedent's rule AND the invariant it protects, from the

authority's own text - not from what makes the analogy convenient. (The source cycle argued from "faces exist selectively" while the text - "For every face f, the compiler creates..." - shows the rule protects universal targetability. Wrong invariant, invalid transport, reversed conclusion.)

1. Show B has the same invariant, or the transport fails. A property

both A and B have that the rule does not protect is decoration.

1. Check the DIRECTION: an analogy that would create in B something A

structurally lacks (an unblockable rule where no unblockable face can exist) is transporting difference while claiming sameness - automatic failure.

1. A valid transport is still an argument, tagged as the authority

tags new claims (it is [N] unless the text states it); an invalid one is withdrawn, not weakened - and the withdrawal recorded, with the refutation pinned by a test where one is possible.

1. Adversarial check: before relying on any transport, write the best

argument that the precedent cuts the OTHER way. If you cannot, you have not understood the precedent; if you can, weigh it in the open. <!-- PROMPT-CORE-END -->