@Winna
So are you saying that existence is not a physical predicate or something? Or is it that even purely conceptual subjects would also be regarded as existing according to Kant's definition of existence? That is, a purely conceptual entity does actually exist but it simply does so in a non-physical way, and is completely dependent upon the person thinking about it for its existence?
No, according to Kant's definition of "existence" you can only apply the predicate with meaning for objects in the sphere of our possible experience - i.e. those for which we can have empirical data.
A non-contradictory concept created in our mind is simply
theoretically possible (as opposed to logically impossible = auto-contradictory), but as long as we have no experience of it we cannot apply meaningfully the predicate of existence.
Whatever is outside the possible experience is outside our means to make verifiable statements. Pure constructs of the mind (like the idea of God) are outside our means to make verifiable statements. We can make non-contradictory statements, but they are nothing but that.
Since existence is not a real predicate (it doesn't add something to the object, it merely poses our relation to it), existence cannot be deduced from a concept. i.e. if I can have the concept of God as omnipotent etc, and perfect (better than anything else I can imagine) and I try to state from it that it has to exist (because, lets say, "something that exists is better than something that doesn't") I committed an error: it's an illusion that existence would add something to my concept; to state existence I need an empirical component - i.e. an experience or potential experience of the object - otherwise my concept is still just a theoretical possibility (logical non-contradiction, like an unicorn), not an object of my experience.
I can construct a concept anyway you want (as long as it doesn't contradict logic), but I can never state or deduce existence of an object for it, unless I
can have it in the sphere of my possible experience - which, for God, he denies it.
There is something special about the concept of God (as compared to unicorn), though. It's that our reason
needs (among others) to see a cause in everything, thus it
tends towards posing a cause of a cause of a cause etc, and to see an end of the chain. But the end of the chain is outside our possible experience, thus we get fooled thinking that we can say it "exists" since: we can only have knowledge about what's in the realm of our (potential) experience, and the primal cause, as needed by our reason as it is, is not within it.
i.e. I can think whatever I want (logically non-contradictory, I can't think of a squared triangle), but in order to state existence of an object for my concept, I need an empirical test.
Basically, if I may put it this way, the entire development of criticizing religion by asking for a proof (= a proof submitted to the criteria of science, meaning empirically verifiable at least in principle), stems from it.
(though, Kant quite denies it's useful to keep asking for that, since we already constructed the concept as outside the possible experience
)