When a cross-domain analogy breaks down, investigate…

When This Fires

• When a cross-domain analogy partially succeeds but breaks down at a specific point
• During knowledge graph bridge-building (A bridge node must generate insights in BOTH connected…) when testing bidirectional mappings
• When explaining a concept from one domain using another and the explanation stops working
• When a mental model borrowed from one field produces wrong predictions in another

Common Failure Mode

Forcing the analogy past its breaking point: stretching the mapping until it technically "works" but has lost all precision and usefulness. "Software forking is like... renovating a building while people live in it!" This preserves the analogy at the cost of accuracy and, critically, at the cost of learning what the failure was trying to teach you.

The Protocol

(1) When a cross-domain mapping breaks down, stop trying to fix it. (2) Ask: "Where exactly does the mapping fail? What specific concept or relationship in Domain A has no meaningful analog in Domain B?" (3) Investigate the mismatch: "What structural feature of Domain B makes this concept impossible/meaningless here?" (4) Document the finding: "Domain B has property X that Domain A lacks, which is why the mapping breaks at this point." (5) This finding is often more valuable than the successful parts of the mapping — it reveals domain-specific structure that you wouldn't have noticed without the cross-domain comparison.