Scribd
Upload
112 views

Topic 15 Conditional and Indirect Proofs

This document discusses two new proof strategies - conditional proof and indirect proof - that are needed to make the natural deduction system complete. Conditional proof involves assuming the antecedent of a conditional statement and deriving the consequent inside a box, thus proving the conditional. Indirect proof involves assuming the negation of the statement to be proved and deriving a contradiction inside a box, thus proving the statement indirectly. The document provides examples of how to construct conditional and indirect proofs. It explains that both strategies rely on temporarily assuming statements inside boxes in order to derive logical consequences.

Uploaded by

Joshua Janine Lugtu
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
112 views

Topic 15 Conditional and Indirect Proofs

This document discusses two new proof strategies - conditional proof and indirect proof - that are needed to make the natural deduction system complete. Conditional proof involves assuming the antecedent of a conditional statement and deriving the consequent inside a box, thus proving the conditional. Indirect proof involves assuming the negation of the statement to be proved and deriving a contradiction inside a box, thus proving the statement indirectly. The document provides examples of how to construct conditional and indirect proofs. It explains that both strategies rely on temporarily assuming statements inside boxes in order to derive logical consequences.

Uploaded by

Joshua Janine Lugtu
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

Lecture 15

Conditional and
Indirect Proofs

W e have been developing a system of natural deduction. We


want a system that is both sound and complete. A system of
proof is sound if every argument for which there is a proof is valid.
Our system so far is sound. A system of proof is complete if every
valid argument has a proof in the system. Our system is not yet
complete. There are valid arguments that our rules of inference and
our equivalences are not enough to complete a proof. We need two
new proof strategies and a new justification to reach that ultimate
goal. The justification is assumption.

Assumption as a Justification
⊲⊲ At this point in the course, you should be flabbergasted at the
thought of assumption being a justification. How can we enter a
sentence in the second column that is a mere assumption?

⊲⊲ After all the time we spent discussing why it was essential to only
put sentences in the second column that we know to be absolutely
true, do you mean to tell us that now we can just enter any
sentence in the second column and justify it as an assumption?
Surely, there are limits to what can be assumed, and those limits
are based on rational inferences about what we can know.

⊲⊲ No. You are free to assume any sentence you want and enter
it into the second column of the proof. We did say clearly and
explicitly that if even one false sentence shows up in the second
column, logical chaos could result.
⊲⊲ But it is okay. We have protection. We have boxes. When we
introduce an assumption (Assumpt) into a proof, we put it in a
box, and that box is logical quarantine. Nothing inside of the box
is allowed to come out into the general population.

⊲⊲ Anything can be brought into the box, but once a sentence has
been in the box with the assumption, anything that inferred from it
is to be deemed to be contagious in that it could be infected with
the possible falsity of the assumption.

⊲⊲ When we open a box and insert an assumption, we create a new


logical world—one that resembles our logical world in certain
ways, so we can infer true propositions in the real world from
what we observe in the hypothetical box world. One way to draw
such inferences is called conditional proof.

Conditional Proof
⊲⊲ We use conditional proof when we want to prove a conditional.
We use this form of reasoning all the time, especially if we have
children. There are two ways one can learn lessons in life: the
easy way and the hard way. The sentence “You should not put
your hand on a hot stove” can be learned the hard way by putting
your hand on the stove.

⊲⊲ The easy way is conditional proof. We say to the person who


has never yet put a hand on the hot stove, “Before you do, let’s
think about this.” Let’s start by assuming that you put your hand
on the hot stove. Don’t really do it in the real world, but create a
hypothetical world in which you did. What would be the result?

⊲⊲ The temperature of the stove is significantly higher than that of


your hand. By the first law of thermodynamics, heat would flow
rapidly from the stove into your hand. Your hand heats up very
quickly—so quickly that it would cause damage to the skin.

Lecture 15—Conditional and Indirect Proofs 145


You would suffer burns. The nerves in your hand would send
message of those burns to your brain, which would register it as
incredible amounts of pain.

⊲⊲ What do we know from this thought experiment? We know that


if you put your hand on a hot stove, then you will experience
incredible amounts of pain. By assuming the antecedent
and then deriving the truth of the consequent, we can assert
the conditional “If you put your hand on a hot stove, you will
feel incredible amounts of pain.” We learned the truth of the
conditional the easy way.

⊲⊲ This is how conditional proof works. To demonstrate a


conditional by conditional proof, the first step is to enter all
of the premises into the proof. The second step is to open a
box and insert the antecedent into the box as an assumption
justified by “Assumpt.”

⊲⊲ Next, pull your premises into the box as needed. Then, use rules
of inference and equivalences. Proceed as if everything were
normal inside of the box, until such time as the consequent of the
conditional appears as a justified line inside of the box.

⊲⊲ At that point, we know that if the antecedent is true, then


the consequent has to be true. So, close the box, and in the
general proof, write down the conditional justified by CP m,n.,
where m is the first line inside the box and n is the last line
inside the box.

Indirect Proof
⊲⊲ One use of assumptions is in conditional proof. The other is
called indirect proof.

146 An Introduction to Formal Logic


⊲⊲ When we think of proofs in mathematics, we think of Euclidean
geometry, which are proofs of the type we started with, where
you assume the premises and demonstrate the conclusion. This
is called direct proof.

⊲⊲ But most mathematical proofs do not take this form. Most are
indirect proofs, or to use the Latin name, reductio ad absurdum,
or reduce to absurdity.

⊲⊲ The “absurdity” is a contradiction. Recall how much we worry


about contradictions. If even one contradiction is true, then
everything is true, and truth goes away. We fear contradictions.
But, like early humans who learned to harness the power of
otherwise dangerous fire for cooking, we, too, will learn to control
the power of contradictions.

⊲⊲ The fundamental principle behind traditional logic is the law of


the excluded middle—that all sentences are either true or false.
A sentence that isn’t true is false, and a sentence that isn’t false
is true.

⊲⊲ So, if we want to prove that a sentence has to be true, that is the


same thing as proving that it can’t be false. This is the trick. We
want to prove that p cannot be false. But how?

⊲⊲ We know that contradictions are always false, so if we can show


that the negation of p, in conjunction with the premises we are
asserting to be true, necessarily leads to a contradiction, then we
have logical grounds on which to reject the negation of p.

⊲⊲ But if −p is false, then p must be true. If we can show that denying


p leads to absurdity, then we have no choice but to accept p.
That is indirect proof.

Lecture 15—Conditional and Indirect Proofs 147


⊲⊲ It begins when we enter the premises into the proof, then open
a box and insert the negation of what we are trying to prove into
the box. We use our premises, the assumption, and our rules of
inference and equivalences until a contradiction, any sentence of
the form a &−a, appears in the box.

⊲⊲ It does not have to be a contradiction involving the assumption.


Any contradiction will do, because, as you have learned, if you
grant even a single contradiction, all sentences—including all
contradictions—are true.

⊲⊲ At this point, we have shown that the negation of the conclusion,


when added to the premises, necessarily results in contradiction,
so we close the box, and in the general proof, we write down the
conclusion, justified by IP with the line numbers from the opening
to the closing of the box.

⊲⊲ In a conditional proof, the box was a hypothetical logical world in


which we posit something that might not be true in order to see
what else would have to be true.

⊲⊲ In an indirect proof, inside the box is a bizarre logical world in


which we put propositions together that we believe cannot be put
together in order to observe the nonsense that results.

⊲⊲ If you are particularly clever, you might have realized that


it might not be the case that the negation of the conclusion is
responsible for the contradiction. All we have shown is that the
combined set of the premises and the negated conclusion lead
to a contradiction.

⊲⊲ How can we assert that the negation of the conclusion is


necessarily false when it might not be the conclusion that is
to blame for the derived contradiction? If the fault is with the
premises and not the negation of the conclusion, what gives us
the right to assert the conclusion?

148 An Introduction to Formal Logic


⊲⊲ The answer is that if it is not the assumption that is to blame for
the appearance of the contradiction, then it has to arise from the
premises alone. That means that the premises are inconsistent—
that they cannot all be true at the same time.

⊲⊲ An argument with inconsistent premises must be valid, because


it would be impossible for all of its premises to be true and its
conclusion to be false. So, on a technicality, we know that the
argument is valid. So, whether the contradiction comes from the
assumption or not, the derivation of the contradiction guarantees
that the argument is valid.

Readings

Barker, The Elements of Logic, chap. 3.


Hurley, Logic, chap. 7.
Kahane, Logic and Philosophy, chap. 6.

Questions

1.
Use conditional proof to show that the following argument is valid.

s → (y &z). (w ∨q) → (y &x). Therefore, (s ∨q) → (y ∨w).

2.
Use indirect proof to show that the following argument is valid.

(a ∨l) &(a ∨m). a → (t &−l). Therefore, (t ∨m) ∨ −l.

Lecture 15—Conditional and Indirect Proofs 149


3.
Translate the following argument and construct both a conditional
and an indirect proof to demonstrate its validity.

I can’t eat turkey or pasta without overeating. So, if I eat turkey, I will
eat turkey and overeat.

150 An Introduction to Formal Logic

You might also like