1.3 Methods of Proof

Definition


A proof is a logical argument that establishes the truth of a mathematical statement by a sequence of deductive steps from agreed-upon axioms and previously proven statements.

Direct Proof


To prove a statement of the form P→Q P right arrow Q, assume P P is true and use logical deductions to show Q Q holds.

  1. Assume P P

  2. logical deductions

  3. Therefore, Q Q

Proof by Contrapositive


To prove P→Q P right arrow Q, prove its contrapositive ¬Q→¬P not Q right arrow not P

  1. Assume ¬Q not Q

  2. logical deductions

  3. Therefore, ¬P not P, hence P→Q P right arrow Q

Proof by Contradiction


To prove a statement S S, assume the negation ¬S not S and derive a contradiction e.g., R R and ¬R not R for some R R.

  1. Assume ¬S not S

  2. logical deductions

  3. Derive a contradiction

  4. Therefore, S S

The Mathematical Induction Principle


To prove a statement P⁡(n) P of n is true for all natural numbers n n:

Base Case: Show P⁡(0) P of 0 is true.

Inductive Step: Assume P⁡(k) P of k is true (the inductive hypothesis) for an arbitrary k≧1 k is greater than over equals 1 and prove P⁡(k+1) P of open paren k plus 1 close paren is true.

Conclusion: If both steps are verified, then P⁡(n) P of n holds for all n∈ℕ n is a member of the natural numbers

The Strong Induction Principle


To prove a statement P⁡(n) P of n is true for all natural numbers 𝑛:

Base Case: Prove P⁡(0) P of 0 is true.

Inductive Step: Assume P⁡(1) P of 1, P⁡(2),⋯ P of 2 comma dot dot dot, P⁡(k) P of k are all true (the strong inductive hypothesis) and then prove P⁡(k+1) P of open paren k plus 1 close paren is true.

Note: Strong induction is logically equivalent to ordinary induction but is especially useful when the proof of P⁡(k+1) P of open paren k plus 1 close paren relies on multiple previous cases rather than just the immediate predecessor.

The only difference between regular induction and strong induction is that in strong induction you assume that every number up to k satisfies the condition that you wish to prove whereas in regular induction you only assume that some integer k satisfies this condition.

The Least Number Principle


Every non-empty subset of ℕ the natural numbers has a least element.

Equivalently, in the language of a property P⁡(n) P of n:

  1. Let P⁡(n) P of n be a statement about natural numbers. Suppose, toward a contradiction, that P⁡(n) P of n fails for at least one n n.

    Define S={n∈ℕ:¬P⁡(n)} S equals the set of all n in the natural numbers such that not P of n.

    By hypothesis, S≠∅ S is not equal to empty set.

  2. Application of the LNP

    Then S S has a least element; call it m=min⁡(S) m equals min of S

  3. Derive the contradiction

    If m=1 m equals 1, you show directly that ¬P⁡(1) not P of 1 is impossible (i.e. you show P⁡(1) P of 1 holds.)

    If m>1 m is greater than 1, then by minimality of m m every k<m k is less than m satisfies P⁡(k) P of k.

    Use that fact to prove P⁡(m)→ P of m right arrowagain contradicting ¬P⁡(m) not P of m.

  4. Conclusion

    Since assuming “there is some n n with ¬P⁡(n) not P of n” led to a contradiction, it must be that S S is empty. Hence ¬P⁡(n) not P of n never occurs, and so P⁡(n)∀n∈ℕ P of n for all n is a member of the natural numbers