2

I'm beginning to study propositional logic from a book called Elementary Logic by Brian Garrett.

In the chapter on the disjunctive connective, I've come across an exercise that seems to require the Principle of Explosion, but this rule is never introduced anywhere in the book. I only know about this principle because others have solved this very problem using it, but I want to solve it using the rules provided in the book, which are the introduction and elimination of the five connectives ($\land\lor\neg\implies\iff$) and assumptions.

Here is the problem: $\neg A \lor \neg B \models (A \implies C) \lor (B \implies C)$

My understanding is that I must prove the following

$\neg A \implies ((A \implies C) \lor (B \implies C))$

$\neg B \implies ((A \implies C) \lor (B \implies C))$

I don't understand how to introduce the term C into this proof without the Principle of Explosion; I'm assuming you need a term C to introduce any kind of implication involving C as its consequent. I understand that with the explosion principle, I can introduce anything after a contradiction, which is how C enters this problem and how I can then deduce the conditionals $A \implies C$ and $B \implies C$

With explosion, the following is what I have:

$ \neg A \lor \neg B\\ \quad\mid\neg A\\ \quad\quad\mid A\\ \quad\quad\mid A \land \neg A\\ \quad\quad\mid C \qquad\qquad Principle \, of \, Explosion\\ \quad\mid A \implies C\\ \quad\mid (A \implies c) \lor (B \implies C)\\ \neg A \implies ((A \implies C) \lor (B \implies C))\\ \quad\mid\neg B\\ \quad\quad\mid B\\ \quad\quad\mid B \land \neg B\\ \quad\quad\mid C \qquad\qquad Principle \, of \, Explosion\\ \quad\mid B \implies C\\ \quad\mid (A \implies c) \lor (B \implies C)\\ \neg B \implies ((A \implies C) \lor (B \implies C))\\ (A \implies C) \lor (B \implies C) $

Given that this text never discusses this principle, and restricts all solutions to using the primary connectives and basic assumptions, which seems to follow minimal logic, can this problem actually be solved? If not, I wonder why this problem presents itself in the book.

Any and all help will be greatly appreciated. Thank you.

sovpariah
  • 31
  • 3
  • Can you use the fact that $A\rightarrow C$ is equivalent to $\neg A\lor C$? – eyeballfrog Jul 06 '22 at 20:34
  • What rules do you have for working with $\lnot$ statements? (For example, if you have an axiom $(\lnot Q \rightarrow \lnot P) \rightarrow (P \rightarrow Q)$, and you have $A$ and $\lnot A$, then from the latter you can show $\lnot C \rightarrow \lnot A$; then use the axiom and elimination of $\rightarrow$ to conclude $A \rightarrow C$; and then use another elimination of $\rightarrow$ with the assumption of $A$ to conclude $C$.) – Daniel Schepler Jul 06 '22 at 20:50
  • The principle of explosion is not needed at all to solve the exercise. But, by the way, this principle is actually a consequence of the usual logic rules, so it does not need to be explicitely stated in the book. – Jean-Armand Moroni Jul 06 '22 at 20:59

1 Answers1

1

The rules, as I know them from this book, for implication is that if $P$ was used to derive $Q$, then we can derive $P \implies Q$. The negation rule is that if from $P$ we derive $Q \land \neg Q$, then we can derive $\neg P$.

At this point in the book, it is suggested that there is no need to use equivalent propositions, or any derived rules to solve the exercise sequences. I'm new to all this, but it seems that using the explosion principle is a derived rule that allows me to deduce $C$, whereas I believe its necessary to assume $C$ or $\neg C$ given the rules above.

I think the following is correct and consistent with the primary rules stated above without using the explosion principle. Please, let me know otherwise. Greatly appreciated.

$ \neg A \lor \neg B\\ \quad\mid\neg A\\ \quad\quad\mid A\\ \quad\quad\quad\mid \neg C\\ \quad\quad\quad\mid A \land \neg A\\ \quad\quad\quad\mid \neg C \land (A \land \neg A)\\ \quad\quad\quad\mid A \land \neg A\\ \quad\quad\mid C\\ \quad\mid A \implies C\\ \quad\mid (A \implies C) \lor (B \implies C)\\ \neg A \implies ((A \implies C) \lor (B \implies C))\\ \quad\mid\neg B\\ \quad\quad\mid B\\ \quad\quad\quad\mid \neg C\\ \quad\quad\quad\mid B \land \neg B\\ \quad\quad\quad\mid \neg C \land (B \land \neg B)\\ \quad\quad\quad\mid B \land \neg B\\ \quad\quad\mid C\\ \quad\mid B \implies C\\ \quad\mid (A \implies C) \lor (B \implies C)\\ \neg B \implies ((A \implies C) \lor (B \implies C))\\ (A \implies C) \lor (B \implies C)\\ $

sovpariah
  • 31
  • 3
  • You are being way too imprecise with everything for anyone to really know whether you got it right. You state the negation rule as ( ( P ⊢ Q∧¬Q ) ⊢ ¬P ) but you seem to use a different rule ( ( ¬P ⊢ Q∧¬Q ) ⊢ P ). They are not the same. – user21820 Jul 06 '22 at 23:02
  • @user21820 You are correct. I should be more clear. The book does state both rules for negation. I applied the second rule ( ( ¬P ⊢ Q∧¬Q ) ⊢ P ). I believe this proof does the job of proving the conclusion without invoking the explosion principle. – sovpariah Jul 06 '22 at 23:51
  • You can easily derive the explosion principle in a few lines if you have those rules for ¬ and the standard rules for ∧. I guess that's why your book did not mention explosion as it is not only redundant but also actually not very useful. – user21820 Jul 07 '22 at 12:20
  • As for your attempt, I don't understand why you had useless steps "¬C∧(A∧¬A)" and "¬C∧(B∧¬B)" along with a pointless repeated line after each of them, but other than that it is correct. Honestly though, I don't like reading this style, and prefer this style, which is readily understood by any proper mathematician, unlike the style you used, which only people familiar with that style can read. – user21820 Jul 07 '22 at 12:25
  • @user21820 As the author himself said, these steps are redundant once you introduce derived rules and equivalent statements. Unfortunately, I'm relegated to using only the very basic rules presented at this point in the book. The method I used is the method suggested by the author, even though it appears redundant. In order to negate ¬C, I have to derive a contradiction from it so I had to include a conjunction statement of ¬C∧(A∧¬A) then derive (A∧¬A) to satisfy the negation rule and derive C. These "tricks" as the author calls them, allows for the rules to be followed with consistency. – sovpariah Jul 07 '22 at 18:39
  • I know you can only use those rules, but you're not paying attention to what you wrote. You have duplicate lines in your proof, and any sane deductive system will not require that line "¬C∧(A∧¬A)" to deduce "A∧¬A" because you had already gotten "A∧¬A" earlier! – user21820 Jul 08 '22 at 14:47
  • @user21820 Please, remember I'm new to this. I don't think you understand what I'm asking. I care to know how introduce term C into the problem with or without assuming. I see a lot of people solve the problem deducing C using the explosion principle. There are ten rules available to me and explosion isn't one of them. I have to assume either C or ¬C, and find a way to discharge the assumption. I'm not concerned about A∧¬A. And you certainly can have duplicate lines in a proof. I'll post you a problem solved by the author in the negation chapter. – sovpariah Jul 11 '22 at 12:59
  • @user21820 ~(A → B) => A 1 (1) ~(A → B) 2 (2) ~A 3 (3) A 4 (4) ~B 2,3 (5) A & ~A 3,2 &I 2,3,4 (6) ~B & (A & ~A) 4,5 &I 2,3,4 (7) A & ~A 6, &O 2,3 (8) B 4–7 ~O 2 (9) A → B 3–8 →I 1,2 (10) (A → B) & ~(A → B) 9,1 &I 1 (11) A 2–10 ~O – sovpariah Jul 11 '22 at 13:00
  • @user21820 I know its difficult to read, but just copy it somewhere and check out lines 5 and 7. The author repeats lines so he can derive a contradiction A & ~A from ~B to derive B, which is what I did in the problem above. If you have a better solution, please post it as I came here for help and assistance. Thank you. – sovpariah Jul 11 '22 at 13:04
  • You are not reading what I wrote at all. Please stop assuming that others do not know what you are asking. There is absolutely no reason to have lines 6,7; they can be deleted and the proof would still be correct. Similarly in the other case. – user21820 Jul 11 '22 at 15:19
  • 1
    Except for the useless steps (as per comments above) the proof seems correct to me. Having said that, it is only a trick to avoid Explosion: if we have a proof that ends with a contradiction $A \land \lnot A$, we can always add an extra assumption $\lnot P$ whatever and using $(\lnot \text O)$ rule (page 74) we get $P$. This is Explosion. – Mauro ALLEGRANZA Jul 12 '22 at 14:05
  • The $(\lnot \text O)$ rule is simply Double Negation; thus Garrett's proof system is not minimal but classical. – Mauro ALLEGRANZA Jul 12 '22 at 14:07