5
$\begingroup$

Say we know that $P(k) \implies P(k+3)$. Then if we know $P(1)$ is true, we know $P(4), P(8) \dots$ are also true.

However if we know $P(1)$ is false, does that mean $P(4), P(8) \dots$ are also false?

I'm a little confused on if we can trust an implication which is of the form, false $\implies P(x)$.

$\endgroup$
  • $\begingroup$ What do you mean by "when FALSE$\implies P(x)$, is $P(x)$ false?"? I think that your question is "If (FALSE$\implies P(x)$) is TRUE, is $P(x)$ false?" $\endgroup$ – Gaston Burrull Jul 1 '13 at 0:02
  • $\begingroup$ I suppose you mean $P(4), P(7), \dotsc$, not $P(4), P(8), \dotsc$ $\endgroup$ – dbanet Jun 19 '15 at 18:38
5
$\begingroup$

No, in your example, knowing $P(1)$ is false tells us absolutely nothing about the truth value of $P(n)$, $n> 1$.

Even if we know in advance that the entire implication is true, so that it is true that $P(k) \implies P(k+1)$, and we also learn that $P(k)$ is false, then we still cannot conclude anything about the truth or falsity of $P(k + 1)$, even though the implication is true.

We have defined implication (aka the material conditional), as it is used in logic and in math, such that $$\text{FALSE} \implies P(x)\quad \text{ is always true}.$$

Indeed, by definition: $$P \implies Q \quad \text{is FALSE if and only if}\;\; P \;\;\text{is true}\;\;{\bf and}\;\; Q \;\;\text{is false}.$$

Recall the truth-table for the implication $P\implies Q$ which gives all possible truth-value assignments to P and Q, and the resultant truth value of $P\implies Q$:

enter image description here

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Does this make sense? It's a good idea to keep remember the one and only condition that makes an implication false: we simply can't have TRUE $\implies$ False. Any other truth-values for P, Q are fair-game, and make an entire implication true. $\endgroup$ – amWhy Jul 1 '13 at 18:37
8
$\begingroup$

Here's a simple example to drive the point home in this specific case: let $P(n)$ be the predicate '$n$ is greater than $2$'. Then if $P(k)$ is true, we have $k+3\gt k\gt 2$, so $P(k+3)$ is true; thus $P(k)\implies P(k+3)$. What's more, $P(1)$ is manifestly false; $1$ is certainly not greater than $2$. But in fact, $P(4)$, $P(7)$, $\ldots$ are all true.

| cite | improve this answer | |
$\endgroup$
1
$\begingroup$

No. $P(x)$ can either be false or true, because the only implication statement that is false is True implies False.

| cite | improve this answer | |
$\endgroup$
0
$\begingroup$

No, a false statement implies a true statement. What I mean is that if $A$ is false and $B$ is true, then the statement $A \implies B$ is true. One way to maybe think about this is to note that $A\implies B$ is equivalent to $\neg A \lor B$ that is not $A$ or $B$. But if $A$ is false, then $\neg A$ is true. And if $B$ is true then true or true is true.

Also if $A$ is false and $B$ is false, then the statement $A\implies B$ is true.

So from this you can take that a false statement implies any statement.

I would recommend reading through this question and answers: True or false? $x^2\ne x\implies x\ne 1$

Or take a look at this: http://www.jstor.org/stable/27961699.

| cite | improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.