I read about this in the book Combinatorics by N Ya. Vilenkin. He presents it like a well-known anecdote, but I haven't been able to find it anywhere else, so as far as I know, he made it up.
The Teakettle Principle
is the idea that the easiest way to solve a problem
is to reduce
it to a problem that you've already solve
d. This seems to make a lot of sense, and I'm now a big fan
of the Teakettle Principle. There is also a good mathematician
that goes along with it:
A mathematician asked a physicist how, given an empty teakettle and an unlit gas stove, water could be boiled. The physicist conjectured that water could be boiled by filling the kettle, lighting the stove, and placing the kettle on the stove. The mathematician agreed. The mathematician then asked how water could be boiled given a filled kettle and a lit stove. The physicist proposed that the kettle should be placed on the stove. The mathematician said, "Wrong! It would be simpler to empty the kettle and turn off the stove. Then, we would have a problem that we already have solved!"