Monthly Archives: March 2017

Formal Proof Challenge!

Intermediate Logic

Several years ago I was teaching a logic course, and we were learning about formal proofs of validity. I enjoy proofs, and to keep myself sharp I was working through a practice quiz in David Kelley’s The Art of Reasoning, when I came across this argument:

D ⊃ (E ⊃ F)
D ⊃ (F ⊃ G)
∴ D ⊃ (E ⊃ G)

I was in a quiet library with plenty of time, but despite all my efforts I could not solve this (without using the Conditional Proof). The next day in class some students were finishing their assignment early, so I  challenged them with this proof, thinking to myself, “That ought to keep them busy,” but not really expecting anyone to succeed. Before the end of class, Caroline Jones came forward and said, “I solved it, Mr. Nance.” I scoffed inwardly at first, only to be pleasantly surprised by her correct solution.

Since that time I have called this “The Caroline Jones” proof, and have challenged my logic students to solve it using only the regular rules of inference and replacement. The most elegant proof I have seen requires twelve total steps.

Anyone up to the challenge?

Reductio Challenge

Intermediate Logic,

In formal proofs of validity, the reductio ad absurdum method can be used to make some proofs easier, and even some shorter. For example, consider this argument:

(~P ⊃ R) • (~Q ⊃ S)    ~(R ∨ S)    ∴ P • Q

The proof for this valid argument is 14 steps without the reductio (which I will let you try to solve on your own), but only 7 steps with the reductio, as shown here:

  1. (~P ⊃ R) • (~Q ⊃ S)
  2. ~(R ∨ S)   /  ∴  P • Q
  3. ~(P • Q)                     R.A.A.
  4. ~P ∨ ~Q                    3 De M.
  5. R ∨ S                         1, 4 C.D.
  6. (R ∨ S) • ~(R ∨ S)   5, 2 Conj.
  7. P • Q                          3-6 R.A.

The reasoning behind the reductio method is this: If assuming that a proposition is false leads to a self-contradiction, then the proposition must be true. This reasoning can itself be written as a propositional argument:

~P ⊃ (Q • ~Q)   ∴  P

This is a valid argument, as a shorter truth table will show. But the proof for this argument (if you are not allowed to use reductio) requires 13 steps, and it is rather difficult to solve. Any takers?

Two Strange Proofs

Mr. Nance,

Could you give real-world examples of the arguments to prove in Intermediate Logic Lesson 18, number 7) U / ∴ W ⊃ W, and number 8) X / ∴ Y ⊃ X, showing how they would be used, or explain them a bit? Thank you.

Thanks for the great question! These two arguments are unusual, so I am not surprised that you are asking about them.

A real-world example for #7 might be Esther 4:16, “I will go to the king which is against the law; if I perish, then I perish!” This argument form basically shows that any proposition implies a tautology.

An example for #8 could be, “God created all things. So even if evolution can be used to explain some fossils, it’s still true that God created all things.” The form of this argument shows that if a proposition is given, any other proposition implies it.

To be honest, my purposes for including those two problems were: 1) to show how very strange the conditional proof is, and 2) to show how this method can be used to simplify otherwise difficult proofs.

Blessings!

Conditional Proof Assumption

Intermediate Logic, Logic,

With the nine rules of inference and the ten rules of replacement taught in Lessons 13-17 of Intermediate Logic, any valid propositional argument can be proven. But for the benefit of the logic student, I introduce an additional rule in Lesson 18: the conditional proof. The conditional proof will often simplify a proof, especially one that has a conditional in the conclusion, making the proof shorter or easier to solve. Conditional proof starts with making an assumption. I want to clarify what happens with that assumption.

To use conditional proof, you start by assuming the antecedent of a conditional. If by using that assumption along with the other premises you are able to deduce the consequent, you can conclude the entire conditional using conditional proof. More briefly, if an assumed proposition p implies the proposition q, we can conclude if p then q.

One misconception new logic students often make is thinking that the assumption actually “comes from” some previous step in the proof. They think that the assumption must appear somewhere else in order to make it. This is not the case. The assumed antecedent doesn’t come from anywhere; it is quite simply assumed. I tell my students we get the antecedent from our imagination; from Narnia, Middle Earth, Badon Hill. With conditional proof, you are allowed to assume any antecedent you wish, as long as you use conditional proof correctly from that point on.

 

May Proofs Use the Same Line Twice?

Mr. Nance,

In the answer to Exercise 17a, problem #12, is there a typo? It has row 5 twice.

There is no mistake there. A given line may be used more than once in a proof, as I say at the end of Lesson 15, “Usually, though by no means always, every step in a proof is used and used once.” Line 5 is used twice, once to simplify to get ~L, and once to commute and simplify to get ~M. 

Blessings!