Idea
Make the explanation cross a boundary
An explanation can fit the facts it was designed around. A more demanding test asks whether it also works somewhere the original fitting exercise did not control. If one account connects distinct observations, the evidence begins to constrain it from more than one direction.
Whewell called attention to this joining of inductions. For a simple illustration, suppose a building inspector infers a concealed leak from pressure loss, then finds moisture exactly where that diagnosis predicts. Two methods have met on one cause. Repeating the pressure reading alone would provide a different kind of support.
Agreement matters more when the routes to it are different.
Ask whether the second clue could disagree with the first.
Where it breaks down
Different-looking reports may share the same source or hidden assumption. Their errors can travel together. Consilience strengthens an explanation without proving it uniquely correct; rival explanations may still account for the combined evidence.
How it connects
An explanation gains reach when a second kind of observation constrains it.
This idea appeared in When different clues meet, the Involves connection for September 18, 2026, which asked: What makes several pieces of evidence stronger than repeated agreement?
Check yourself
What should you check before counting two clues separately?
Whether they share a source or assumption that could produce the same error. Exactly. Apparent diversity can conceal dependence between evidence streams.
What the sources establish
Support from different evidence classes is distinct from repeated reports of one observation.
Sources
William Whewell (Stanford Encyclopedia of Philosophy), Consilience of inductions and confirmation