Meaning of I(I())

Sample code:

typedef int I;
struct X{ X(int); };

int main()
{
    int(int());
    X(X());
    I(I());
}

The line int(int()); is an expression using functional cast notation - it's a temporary int initialized with a value-initialized int.

The line X(X()); is a declaration of a function named X taking no arguments returning struct X.

My question is: what is the meaning of I(I()) here? And what rules in the standard determine the difference in meaning between these three cases?