Indirect Proofs and Proofs by Contradiction

Indirect Proof

  • Concept: An indirect proof, also known as proof by contraposition, is a method of proving a statement by assuming the negation of the conclusion and showing that it leads to the negation of the hypothesis.
  • Form: To prove n:(P(n)Q(n))∀n : (P(n) → Q(n)), instead of directly showing that P(n)P(n) implies Q(n)Q(n), we show that ¬Q(n)¬Q(n) implies ¬P(n)¬P(n).
  • Validity: This method is valid because the implication P(n)Q(n)P(n) → Q(n) is logically equivalent to its contrapositive ¬Q(n)¬P(n)¬Q(n) → ¬P(n).

Example: Proof that if n2n^2 is divisible by 3, then nn is divisible by 3

  • Statement: If n2n^2 is divisible by 3, then nn is divisible by 3.
    • P(n)P(n): nn is divisible by 3.
    • Q(n)Q(n): n2n^2 is divisible by 3.
  • Indirect Proof:
    • Assume that nn is not divisible by 3.
    • We want to show that this implies n2n^2 is not divisible by 3.
    • Case 1: If nn is not divisible by 3, it leaves a remainder of 1 when divided by 3.
      • n=3k+1n = 3k + 1 for some integer kk.
      • Then, n2=(3k+1)2=9k2+6k+1=3(3k2+2k)+1n^2 = (3k + 1)^2 = 9k^2 + 6k + 1 = 3(3k^2 + 2k) + 1.
      • Thus, n2n^2 is not divisible by 3, as it leaves a remainder of 1.
    • Case 2: If nn is not divisible by 3, it leaves a remainder of 2 when divided by 3.
      • n=3k+2n = 3k + 2 for some integer kk.
      • Then, n2=(3k+2)2=9k2+12k+4=9k2+12k+3+1=3(3k2+4k+1)+1n^2 = (3k + 2)^2 = 9k^2 + 12k + 4 = 9k^2 + 12k + 3 + 1 = 3(3k^2 + 4k + 1) + 1.
      • Thus, n2n^2 is not divisible by 3, as it leaves a remainder of 1.
    • In both cases, if nn is not divisible by 3, then n2n^2 is not divisible by 3. This completes the indirect proof.

Proof by Contradiction

  • Concept: A proof by contradiction involves assuming the negation of the statement to be proven and showing that this assumption leads to a contradiction.
  • Method:
    • Assume ¬P¬P is true, where PP is the statement to be proven.
    • Show that this assumption leads to a contradiction (a statement that is always false, denoted as FF).
    • Conclude that PP must be true.
  • Validity:
    • The logical equivalence ¬PF\neg P → F implies PP is true.
    • Truth table verification:
      | P | ¬P | F | ¬P → F | P |
      |---|----|---|--------|---|
      | T | F | F | T | T |
      | F | T | F | F | F |
    • Since the columns for PP and ¬PF¬P → F are the same, the equivalence is verified.

Theorem: 2\sqrt{2} is irrational

  • Proof (by contradiction):
    • Assume that 2\sqrt{2} is rational. This is the negation of what we want to prove.
    • If 2\sqrt{2} is rational, then it can be written as 2=ab\sqrt{2} = \frac{a}{b}, where aa and bb are integers with no common factors, and aa and bb are positive.
    • It follows that a2=2b2a^2 = 2b^2.
    • Since bb is an integer, b2b^2 is also an integer, so a2a^2 is even.
    • If a2a^2 is even, then aa must be even. If aa were odd, then a2a^2 would also be odd.
    • Since aa is even, we can write a=2ka = 2k for some positive integer kk.
    • By substitution, (2k)2=2b2(2k)^2 = 2b^2, which gives 4k2=2b24k^2 = 2b^2, and thus b2=2k2b^2 = 2k^2.
    • This tells us that b2b^2 is even. Using the same reasoning as before, bb must also be even.
    • Now we have that both aa and bb are even, which contradicts our initial assumption that aa and bb have no common factors.
    • Therefore, our assumption that 2\sqrt{2} is rational must be false, and 2\sqrt{2} is irrational.

Theorem: There are infinitely many primes

  • Proof (by contradiction):
    • Assume that there are finitely many primes: p<em>1,p</em>2,p<em>3,,p</em>np<em>1, p</em>2, p<em>3, …, p</em>n (where pnp_n is the largest prime).
    • Consider the number q=p<em>1p</em>2p<em>3p</em>n+1q = p<em>1 * p</em>2 * p<em>3 * … * p</em>n + 1.
    • qq is greater than pnp_n (the largest prime).
    • Note that if we divide qq by any of the primes pip_i, we get a remainder of 1.
    • qq cannot be prime because it's greater than pnp_n (the supposed largest prime).
    • However, qq is not divisible by any of the primes p<em>1,p</em>2,p<em>3,,p</em>np<em>1, p</em>2, p<em>3, …, p</em>n, meaning it has no prime factors from that set.
    • This contradicts the Fundamental Theorem of Arithmetic, which states that every integer greater than 1 can be represented uniquely as a product of prime numbers.
    • Therefore, our assumption that there are finitely many primes must be false, and there are infinitely many primes.