$$∀x ∃y ∀z p(x,y,z) ≠ ∀x ∀z ∃y p(x,y,z)$$
I need to prove this is true, but am confused since the only difference in the two statements is that the order of the $∀z$ and $∃y$ is reversed.
$$∀x ∃y ∀z p(x,y,z) ≠ ∀x ∀z ∃y p(x,y,z)$$
I need to prove this is true, but am confused since the only difference in the two statements is that the order of the $∀z$ and $∃y$ is reversed.
On
This is because in the LHS, the existing $y$ must only be a function of $x$ NOT $z$ but in the RHS, $y$ can be a function of both $x$ and $z$. This is why the RHS and LHS are not equivalent.
On
It often helps to break these formulas down into quasi-English to see what they're trying to say.
For the left-hand side:
"For every $x$, there exists a $y$, such that for every $z$, $p(x, y, z)$ is true."
Or, alternatively,
"No matter what $x$ I pick, I can always find a $y$ that will make $p(x, y, z)$ true for all values of $z$."
For the right-hand side:
"For every $x$, and every $z$, there exists a $y$, such that $p(x, y, z) is true."
Or:
"No matter what $x$ and $z$ I pick, I can always find a $y$ to make $p(x, y, z)$ true."
Can you see the difference?
To prove that they're not the same, you need to find a case where one side holds but the other doesn't - and the left side is more restrictive than the right (because the right side allows for you to pick a different $y$ for each value of $z$, whereas the left side says that there's a single $y$ that makes the statement true regardless of $z$).
If I remove $x$ from the statements, I could use something like this to show the difference:
$q(y, z) = (y = z + 1)$, i.e. my statement $q$ is that "$y$ is one more than $z$". So $\forall z \exists y q(y, z)$ means "For every $z$, there exists a $y$ such that $q$ is true" which makes sense - if you pick $z$, I pick the $y$ that's one more than it. On the other hand, $\exists y \forall z q(y, z)$ means "There exists a $y$ such that $q$ is true for all $z$", i.e. there's a magical number that is always one more than every number in existence, which is clearly untrue.
On
First, let me point out that the order of quantifier can really change the meaning of statements.
Consider: $\forall x \exists y \ x<y$ is true for the natural numbers, since for every every number you can find a greater number.
However, $\exists y \forall x \ x<y$ is false, since it is not true that there is a number greater than all numbers.
To demonstrate the non-equivalence of your two statements, consider as a domain the integers, and interpret $p(x,y,z)$ as $x+y=z$
Then the left claim says: "for all integers $x$ there is some integer $y$ such that for all integers $z$ it holds that $x+y=z$ .... which is false, since for whatever $x$ and $y$ you have, you can point to $z=x+y+1$ to show that $x+y=z$ does not hold.
The right claim says: "for all integers $x$ and $z$ there is some integer $y$ such that $x+y=z$ ... which is true, since for $y$ you can just pick $z-x$
On
Since all you've done is reorder the quantifiers, let's simplify the situation by reducing the number of variables and quantifiers so the reason is obvious.
Let $X$ be the set of all people, let $Y$ be the set of all mothers, and let the propositional function $P(x,y)$ be the statement "$y$ gave birth to $x$." Now, are the following two statements logically equivalent?
$$\forall x \exists y P(x,y)$$ $$\exists y \forall x P(x,y)$$
The first statement says "For every person $x$, there is at least one mother $y$ such that $y$ gave birth to $x$." In other words, it says "Every person was born to at least one mother."
The second statement says "There is at least one mother $y$ such that, for every person $x$, $y$ gave birth to $x$." In other words, it says "There is at least one mother that has given birth to all people."
No, they are certainly not equivalent! Reordering the quantifiers of a statement changes the meaning of the statement.
Here is an example where they are not equivalent. Interpreting the symbols, Suppose that our variables can take values in $\mathbb R \setminus \{0\}$ and that $p(x,y,z)$ is said to hold if $xyz = 1$. Then the right-hand statement is true: Whenever we have some non-zero $x$ and $z$, then we can let $y = 1/xz$ to ensure that $xyz = 1$. Thus, in these semantics, the formula $(\forall x) (\forall z) (\exists y) p(x,y,z)$ is true.
However, Let $x = 1$ and let $y$ be arbitrary. Then for $z \neq 1/y$. Then $xyz \neq 1$. This proves the negation of $(\forall x) (\exists y) (\forall z) p(x,y,z)$, so this formula is false.
Do you think you could formalize this?