NOT.5:5.1 - Translate a dependency drawing without inventing geometric meaning

A drawing has named nodes A, B, C and D. An arrow X→Y means that Y takes X’s result as its direct input. Box positions carry no meaning. In one drawing the arrows are A→B and B→C; D is isolated. The receiver must determine C’s direct input using ordinary text.

Construct this textual notation:

nodes: A, B, C, D
direct-input pairs (supplier, receiver): (A, B), (B, C)

The reading rule selects the pair whose receiver is C and returns its supplier, B. Translating back draws one node per declared name and one arrow per pair. This recovers the named directed structure, including isolated D. It need not recover the original layout.

A list of node names alone fails. The second drawing with A→B and A→C would give the same list but a different answer for C. Retaining the pairs repairs that loss. Retaining only pairs would create a different loss: isolated D would disappear from a reconstructed drawing.

Now suppose horizontal position additionally records a planned start time. The earlier translation still answers the direct-input question, but it cannot reconstruct that timing. Add a start-time value for each node if timing is now needed. That extension follows a changed interpretation or question; it is not evidence that the original drawing already specified times.