Page 182 - HOW TO PROVE IT: A Structured Approach, Second Edition
P. 182

P1: PIG/  P2: OYK/
                   0521861241c04  CB996/Velleman  October 20, 2005  2:54  0 521 86124 1  Char Count= 0






                                   168                        Relations
                                   so by one of our strategies for disjunctions from Chapter 3 we can assume
                                   ¬(A = ∅∨B = ∅) and prove A = B. Note that by one of DeMorgan’s laws,
                                   ¬(A = ∅∨B = ∅) is equivalent to A  = ∅∧B  = ∅, so we treat this as two
                                   assumptions, A  = ∅ and B  = ∅. Of course we could also have proceeded dif-
                                   ferently, for example by assuming A  = B and B  =∅ and then proving A =∅.
                                   But recall from the commentary on part 5 of Theorem 4.1.3 that A = ∅ and
                                   B = ∅ are actually negative statements, so because it is generally better to
                                   work with positive than negative statements, we’re better off negating both of
                                   them to get the assumptions A  = ∅ and B  = ∅ and then proving the pos-
                                   itive statement A = B. The assumptions A  = ∅ and B  = ∅ are existential
                                   statements, so they are used in the proof to justify the introduction of y and z.
                                   The proof that A = B proceeds in the obvious way, by introducing an arbitrary
                                   object x and then proving x ∈ A ↔ x ∈ B.
                                     For the ← direction of the proof, we have A = ∅∨B = ∅∨A = B as a
                                   given, so it is natural to use proof by cases. In each case, the goal is easy to
                                   prove.
                                     This theorem is a better illustration of how mathematics is really done than
                                   most of the examples we’ve seen so far. Usually when you’re trying to find
                                   the answer to a mathematical question you won’t know in advance what the
                                   answer is going to be. You might be able to take a guess at the answer and you
                                   might have an idea for how the proof might go, but your guess might be wrong
                                   and your idea for the proof might be flawed. It is only by turning your idea into
                                   a formal proof, according to the rules in Chapter 3, that you can be sure your
                                   answer is right. Often in the course of trying to construct a formal proof you will
                                   discover a flaw in your reasoning, as we did earlier, and you may have to revise
                                   your ideas to overcome the flaw. The final theorem and proof are often the result
                                   of repeated mistakes and corrections. Of course, when mathematicians write
                                   up their theorems and proofs, they follow our rule that proofs are for justifying
                                   theorems, not for explaining thought processes, and so they don’t describe
                                   all the mistakes they made. But just because mathematicians don’t explain
                                   their mistakes in their proofs, you shouldn’t be fooled into thinking they don’t
                                   make any!
                                     Now that we know how to use ordered pairs and Cartesian products to talk
                                   about assigning values to free variables, we’re ready to define truth sets for
                                   statements containing two free variables.


                                   Definition 4.1.5. Suppose P(x, y) is a statement with two free variables in
                                   which x ranges over a set A and y ranges over another set B. Then A × B is the
                                   set of all assignments to x and y that make sense in the statement P(x, y). The
   177   178   179   180   181   182   183   184   185   186   187