What is the difference between these two logical expressions?

89 Views Asked by At

I'm reading our lecturers recitations, and the lecturer remarked some comment that confuses me.

Consider the expression $\exists !x:P(x)$, where $P$ is some predicate.
This expresses the existence of a unique value for which the predicate $P$ holds.
In logical form, this can also be written as: $\exists x:(P(x)\wedge \neg\exists y:(P(y)\wedge y\ne x))$, and I can see why, but then it says there in a small remark: "Observe the difference between that and $(\exists x:P(x))\wedge (\neg\exists y:(P(y)\wedge y\ne x))$. can you see why the latter form is not valid?"
I tried to think about it but couldn't figure out why is it not valid.
Can someone please clarify this for me?

1

There are 1 best solutions below

2
On BEST ANSWER

In the first expression :

$∃x:[P(x)∧¬∃y:(P(y)∧y≠x)]$

I've slightly modified the parentheses in order to highlight the scope of the first existential quantifier: the whole formula, with the two occurrences of the variable $x$ bounded by $∃x$.

The fact that both occurrences are in the scope of the quantifier means that they are referring to the same object.

In the other expression :

$(∃x:P(x))∧(¬∃y:(P(y)∧y≠x))$

the first existential quantifier bound only the variable $x$ of $P(x)$.

The occurrence of $x$ in the subformula $(¬∃y:(P(y)∧y≠x))$ is free; this means that it can refer to an objcet different from the one for which $P$ holds of, according to the sub-formula $(∃x:P(x))$.


Consider as example the domain $\mathbb N$ of natural numbers and consider as interpreation of the predicate $P(x)$ the relation : $(x=0)$.

With this interpretation the first formula means :

$∃x:[(x=0)∧¬∃y:((y=0)∧y≠x)]$

which is a (trivial but) true statement regarding the natural numbers and it express exactly the fact taht $0$ is the unique number satisfying the formula $(x=0)$.

If we "instantiate" the existential quantifier with $x:=0$ we get :

$(0=0)∧¬∃y:((y=0)∧y≠0)$

that is clearly true ; we cannot find a value for $y$ such that $(y=0)∧(y≠0)$ holds.

Consider now the second expression; with the above interpretation it means :

$(∃x:(x=0))∧(¬∃y:((y=0)∧y≠x))$.

The left-concjunct is clearly true ; it's enough to choose $0$ as value for the variable $x$.

But now, the right-conjunct has a free occurrence of $x$ and we are not "forced" to assign to it the same value used in the left-conjunct.

We can choose $1$ as value for the free $x$ in the right-conjunct and what we get is :

$(0=0))∧(¬∃y:((y=0)∧y≠1))$.

But now the formula is false; it is enough to choose $0$ also as value for $y$ and we have that $(0=0)∧(0≠1)$ holds. Thus, it is not true that $¬∃y:((y=0)∧y≠1)$.