TITLE: What is known about the "moduli space of morphisms" $X \to Y$?
QUESTION [8 upvotes]: Let $X$ and $Y$ be projective varieties.  I am assuming that there is some construction of a "moduli space" parametrizing the morphisms $X \to Y$. For instance, if one identifies such morphisms with their graphs in $X \times Y$, one can at least hope that these graphs would correspond to a locally closed subscheme of the Chow or Hilbert schemes for $X \times Y$, although this may not be the best way to construct such a space.  In the case of $X = Y = \mathbb{P}^n$, this space should probably be a disjoint union of $PGL_n$ with $\mathbb{P}^n$ [Edit:This is not correct; I know a theorem that the image must be either a point or all $\mathbb{P}^n$, but grossly misapplied it by neglecting maps of degree > 1].  (I am requiring the varieties to be projective because, for instance, if $Y$ is the affine line, then the morphisms $X$ to $Y$ would naturally correspond to the global sections of $X$, which for general $X$ is too big to fit in one nice scheme, being infinite-dimensional.  Although I suppose it might work if we confined ourselves to e.g. morphisms of a fixed degree, and we are already doing something of this sort in looking at Chow or Hilbert schemes, so perhaps my concern here is needless.)
Is there a standard space that represents these morphisms, and if so, what is known about it?  Is there a good reference for this?

REPLY [6 votes]: There is a good non-stacky intro to this in the first chapter of Rational curves on algebraic varieties by János Kollár. You can also find some very nice applications there, such as Mori's Bend & Break.