Proof by exhaustion

Proof by exhaustion, also known as proof by cases, perfect induction, or the brute force method, is a method of mathematical proof in which the statement to be proved is split into a finite number of cases and each case is checked to see if the proposition in question holds[1]. A proof by exhaustion contains two stages:

  1. A proof that the cases are exhaustive; i.e., that each instance of the statement to be proved matches the conditions of (at least) one of the cases.
  2. A proof of each of the cases.

In the Curry–Howard isomorphism, proof by exhaustion and case analysis are related to ML-style pattern matching.

Contents

Example

To prove that every integer that is a perfect cube is either a multiple of 9, or 1 more, or 1 less than a multiple of 9.

Proof:
Each cube number is the cube of some integer n. This integer n is either a multiple of 3, or 1 more or 1 less than a multiple of 3. So these 3 cases are exhaustive:

How many cases?

There is no upper limit to the number of cases allowed in a proof by exhaustion. Sometimes there are only two or three cases. Sometimes there may be thousands or even millions. For example, rigorously solving an endgame puzzle in chess might involve considering a very large number of possible positions in the game tree of that problem.

The first proof of the four colour theorem was a proof by exhaustion with 1,936 cases. This proof was controversial because the majority of the cases were checked by a computer program, not by hand. The shortest known proof of the four colour theorem today still has over 600 cases.

Mathematicians prefer to avoid proofs with large numbers of cases. Such proofs feel inelegant to them. A proof with a large number of cases leaves an impression that the theorem is only true by coincidence, and not because of some underlying principle or connection. Other types of proofs—such as proof by induction (mathematical induction)—are considered more elegant. However, there are some important theorems for which no other method of proof has been found, such as

See also

Notes

  1. ^ Reid & Knipping 2010, p. 133.

References