Is there a standard reference for the fact that, in an appropriate algebraic-geometrical context, the tangent space at the point $[E]$ to the moduli space $\mathcal M$ is something like $\operatorname{Ext}^1(E, E)$? I am primary interested in two situations:

  • $\mathcal M$ is the moduli space of semi-stable coherent torsion-free sheaves on $\mathbb CP^2$ with given Hilbert polynomial,
  • or $\mathcal M$ is the moduli space of semi-stable finite-dimensional complex representations of finite acyclic quiver with given dimension vector,

and in both situation, if the coefficients of the Hilbert polynomial/dimension vector are coprime, there is a smooth projective variety that is a fine moduli space $\mathcal M$.

Unfortunately, I do not know any deformation theory, so a study book on it that also proves something like $T_{[E]} \mathcal M=\operatorname{Ext}^1(E, E)$ would be welcome.

  • 5
    Have a look at The Geometry of moduli space of sheaves by Huybrechts and Lehn -- see e.g. Corollary 4.5.2. – abx Oct 31 '16 at 20:44
up vote 3 down vote accepted

Theorem 2.6 (page 9) from Hartshorne Lectures on Deformation Theory (it seems that Hartshorne uses the same notation both for an affine scheme $D$ and its function algebra):

Let $X$ be a scheme over $k$, and let $\mathcal F$ be a coherent sheaf on $X$. We define a deformation of $\mathcal F$ over $D=\operatorname{Spec} (k[t]/t^2)$ to be a coherent sheaf $\mathcal F_0$ on $X_0 = X \times D$, flat over $D$, together with a homomorphism $\mathcal F_0 \to \mathcal F$ such that the induced map $\mathcal F_0 \otimes_{k[t]/t^2} k \to \mathcal F$ is an isomorphism. Then the deformations of $\mathcal F$ over $D$ are in natural one-to-one correspondence with the elements of $Ext^1 (\mathcal F, \mathcal F)$, with the zero-element corresponding to the trivial deformation.

Now by the universal property of the moduli space $\mathcal M$ of coherent torsion-free sheaves on $\mathbb{CP}^2$ its tangent space at $[\mathcal F]$ consists of the deformations of $\mathcal F$ over $D$. Moreover, I believe that the proof should work as well over "a non-commutative affine scheme" whose function algebra is the path algebra of a finite acyclic quiver.

Your Answer


By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.