Science Puzzle
The Dowser’s Challenge
A dowser claims his forked rod dips over buried water. A scientist proposes a test: ten pipes, some filled, some empty, locations concealed. The dowser agrees.
He scores no better than chance. He then explains that the presence of sceptics disturbs the energy field, which is why the rod failed. Any future failure will be explained the same way.
What has gone wrong with the claim?
The Answer
The claim has been rescued into meaninglessness. This move has a name: an ad hoc rescue. When a prediction fails, a new condition is bolted on for the sole purpose of explaining away the failure.
Karl Popper argued that what separates science from pseudoscience is not confidence, complexity, or even being right. It is falsifiability: a scientific claim must forbid something. It must be possible to describe, in advance, a result that would show it to be false.
The original claim was perfectly scientific. "The rod dips over water" forbids the rod failing to dip over water, and that is exactly what the test looked for. It failed the test. The revised claim forbids nothing at all, because every conceivable failure now has a built-in excuse waiting for it.
This is the trap: the revision feels like it strengthens the claim, because it now survives all evidence. In fact it destroys it. A theory that is compatible with every possible observation tells you nothing about which observation you will actually make, and a theory that predicts nothing explains nothing.
The tell to watch for is the direction of travel. Adding a condition because your theory predicted it in advance is progress. Adding a condition only after a failure, purely to absorb that failure, is retreat.
The principle: Falsifiability and ad hoc rescues. A scientific claim must forbid some outcome. Amending a claim after each failure so that no result could ever refute it empties it of content.