In the introduction we provided
some examples of ~I. But recall that it and ~E differ from conditional proof in
that you can only apply the rule if the last line of the subderivation is an
explicit contradiction. And what you obtain if you apply ~I is the negation of
the assumption you have made. Here we will study negation introduction in more
detail. Recall that it is also a mode of proof. When you apply it you are
appealing to the subderivation as a whole, not simply to the two lines you
highlight in the builder. The builder indicates this by placing a '-' rather
than a ',' between the numbers of the two lines that you highlighted. Once an
assumption has been discharged the lines in the subderivation are inaccessible.
Generally, though not invariably, you will set up a negation introduction when
you the sentence you are trying to derive is something other than a conditional
or biconditional, or when you have need of something other than a conditional
or a biconditional as a step along the way toward your conclusion. Let's start
off with an easy one:
Here are the premises we will use in a more complex derivation:
|
1.
|
~(P v Q)
|
Premise
|
Let us suppose that we are trying to show that ~(P v Q), R - P, R v Q I-
Q. Note that if we were to have ~R as an accessible line then we could obtain
the conclusion we want via an application of DS. We could obtain ~R were we to
have ~P. So we shall obtain ~P via negation introduction. Here is one completed
derivation:
|
1.
2.
|
~(P v Q)
|
Premise
|
This is a fairly long derivation but it is straightforward. But there is
another way. Let us still suppose that we are trying to show that ~(P v Q), R
- P, R v Q I- Q. Since we want Q let us assume ~Q. ~I would give us ~~Q,
but we could get Q via ~~E. Or you could use ~E. Here is one completed
derivation where we use ~I:
|
1.
|
~(P v Q)
|
Premise
|
You should now:
Here is another example. You
wish to show that P - Q, ~(~P v R), Q ~A, W - A
I- ~W. In this case we will proceed by initially assuming W. The derivation
will be long, but nonetheless not difficult.
|
1.
|
P - Q
|
Premise
|
We said that this derivation was long, but not difficult. The justification for
the claim that it is not difficult is simply that, for the most part, after we
made our assumption we completed it by applying rules that we could. You should
now:
~E, negation elimination, works
in fundamentally the same way. But it can only be applied if the assumption is
some negation ~p. Here is a very simple example where we show that ~(~P v A) |-
P.
|
1.
|
~(~P v A)
|
Premise
|
Here is a slightly longer but easy derivation that uses a ~E along the way:
|
1.
|
R - (~P - Q)
|
Premise
|
You should now:
Before turning to the next section let us try one more. You should now: