My understanding of the ontological argument is that if you want it not to sound entirely retarded, you build it in an abstract form of modal logic based around “possible worlds” where “possible” and “necessary” are predicates. Then using this system you get people to accept an absurd premise based largely on the fact it sounds reasonable if you use the terms wrong.
The basic rules for this sort of logical system is that a “possible world” is roughly describable as “any world that can possibly be imagined” - which is to say, any world that isn’t itself internally contradictory. Statements may be true in it that aren’t true in our world, because each “possible world” is self-contained and not dependent on realities in the other ones.
There are then two new predicates, as noted above:
“Possible” - if something is “possible” it must be true in one or more “possible worlds”.
“Necessary” - if something is “necessary” it must be true in every “possible world”.
Okay, got all that? You might wonder what this is useful, aside from constructing the modal ontological argument in. The answer is…I’m not entirely sure. I gather that it’s an apparently useful way to visualise and describe whole categories of possibilities, without bogging yourself down with a billion if statements. For example you could say, “in every possible world where it’s raining, it’s necessary that something is getting wet”, or “It’s not possble for both Qui-gon and Luke Skywalker to be alive at the same time - so in every possible world where Qui-gon is alive, it’s necessary that luke skywalker isn’t.”
It’s relevent that I reference fictional characters - possible worlds include every possible world including fictional worlds - anything that you or anyone can imagine (so long as it’s internally consistent). Obviously, this includes reality, because there’s nothing self-contradictory about reality. Any world you can imagine, any world at all, is a possible world of its own. (And any variants of it, however similar, are also their own possible worlds, too.)
Okay, that’s the background. So what’s this have to do with the ontological argument? Let me relate it in (roughly) these terms, numbered for convenience to show where the bullshit gets inserted.
Statement 1: “God exists” is possible.
Statement 2: It’s awesomer to be necessarily existent, than to merely be possibly existent, and God’s totally awesome, so if God exists, his existence is also necessary.
Statement 3: By statement 1, there’s a possible world where “God exists” is true. Therefore, from statement 2, “God exists” is necessarily true.
Statement 4: Any thing that’s got necessary existence exists in all possible worlds, including actual reality. So, “God exists” is true in actual reality.
Statement 5: So, Christ exists.
The part that everyone points out first is the leap from Statement 4 to Statement 5 - "when did “Necessary God” become “Christ”? In a way this is the brilliance of the argument, because this is not a valid objection to the argument, and it distracts from the valid objection.
The reason why the 4/5 leap isn’t a valid point of objection is because whatever God they want, they were talking about it all along. Aside from pesky internal-contradiction disproofs, the specific god they worship does exist in some “possible world” - remember, all that means is that it’s imaginable. So they don’t have to tack on the other attributes they want later; they had them in the argument all along.
So, if that’s not the point of objection, what is? Well, it’s statement 2, obviously; the point where we concede that it’s possible to imagine a necessary god. What’s happening here is abuse of the terminology and the logic system itself. Necessary sounds awesome, and it seems like a reasonable thing to let a person imagine about their god, since we’re still talking about an imaginary god at that point. So what’s the problem?
Well, the problem is that “necessary” means something specific in this logic system - put succinctly, it means “something that cannot be imagined not to be true”. Because, remember, any imaginable world is a possible world. And anything necessary is true in every possible world. So by allowing them to take as axiomatic the innocuous-sounding statement that their God’s existence is “necessary”, you are actually conceding that it’s unimaginable that their God is nonexistent. Once you’ve conceded that, of course, the argument falls out as expected.
As should be clear now, the entire argument is based on abuse of terminology. In reality, nobody should ever accept necessaryness as true based on somebody’s say-so.
In fact, it’s trivial to prove that nothing’s existence is necessary, because it’s trivial to imagine a world that’s entirely empty - one in which nothing exists. It’s not self-contradictory, and it’s easily imaginable, so it’s a possible world. And in it, all “this or that thing exists” statements are false - so it’s impossible for any “this or that thing exists” statements to be necessarily true.
What good is the term “necessary” if nothing necessary can exist, you might demand. The answer of course is it’s useful for things that don’t exist. Remember, when it comes down to it “necessay” describes statments, not objects; it’s about truth, not existence. And there are things that are always true, in every possible universe. Here’s an example: “1 + 1 = 2” (assuming standard definitions of the symbols). It’s always true, because it’s true by definition, regardless of the universe you’re in. Much, but not all, of math is this way. The bits that aren’t are the ones which would define where you’re using euclidian or non-euclidian geomotry, for example. When you get to that point, it’s useful to have symbology to distinguish which bits of the math are always true, and which ones are dependent on the geometry of the world you’re in. And that’s where “necessary” and “possible” come into play. Not in some sleazy tricky god-proof.