The Stacks project

Lemma 89.17.5. Let $\mathcal{F}$ be a deformation category. Let $A' \to A$ be a surjective ring map in $\mathcal{C}_\Lambda $ whose kernel $I$ is annihilated by $\mathfrak m_{A'}$. Let $x \in \mathop{\mathrm{Ob}}\nolimits (\mathcal{F}(A))$. If $\text{Lift}(x, A')$ is nonempty, then there is a free and transitive action of $T\mathcal{F} \otimes _ k I$ on $\text{Lift}(x, A')$.

Proof. Consider the ring map $g : A' \times _ A A' \to k[I]$ defined by the rule $g(a_1, a_2) = \overline{a_1} \oplus a_2 - a_1$ (compare with Lemma 89.10.8). There is an isomorphism

\[ A' \times _ A A' \xrightarrow {\sim } A' \times _ k k[I] \]

given by $(a_1, a_2) \mapsto (a_1, g(a_1, a_2))$. This isomorphism commutes with the projections to $A'$ on the first factor, and hence with the projections of $A' \times _ A A'$ and $A' \times _ k k[I]$ to $A$. Thus there is a bijection
\begin{equation} \label{formal-defos-equation-one} \text{Lift}(x, A' \times _ A A') \longrightarrow \text{Lift}(x, A' \times _ k k[I]) \end{equation}

By Remark 89.17.4 there is a bijection
\begin{equation} \label{formal-defos-equation-two} \text{Lift}(x, A') \times \text{Lift}(x, A') \longrightarrow \text{Lift}(x, A' \times _ A A') \end{equation}

There is a commutative diagram

\[ \xymatrix{ A' \times _ k k[I] \ar[r] \ar[d] & A \times _ k k[I] \ar[r] \ar[d] & k[I] \ar[d] \\ A' \ar[r] & A \ar[r] & k. } \]

Thus if we choose a pushforward $x \to x_0$ of $x$ along $A \to k$, we obtain by the end of Remark 89.17.4 a bijection
\begin{equation} \label{formal-defos-equation-three} \text{Lift}(x, A' \times _ k k[I]) \longrightarrow \text{Lift}(x, A') \times \text{Lift}(x_0, k[I]) \end{equation}

Composing (, (, and ( we get a bijection

\[ \Phi : \text{Lift}(x, A') \times \text{Lift}(x, A') \longrightarrow \text{Lift}(x, A') \times \text{Lift}(x_0, k[I]). \]

This bijection commutes with the projections on the first factors. By Remark 89.17.3 we see that $\text{Lift}(x_0, k[I]) = T\mathcal{F} \otimes _ k I$. If $\text{pr}_2$ is the second projection of $\text{Lift}(x, A') \times \text{Lift}(x, A')$, then we get a map

\[ a = \text{pr}_2 \circ \Phi ^{-1} : \text{Lift}(x, A') \times (T\mathcal{F} \otimes _ k I) \longrightarrow \text{Lift}(x, A'). \]

Unwinding all the above we see that $a(x' \to x, \theta )$ is the unique lift $x'' \to x$ such that $g_*(x', x'') = \theta $ in $\text{Lift}(x_0, k[I]) = T\mathcal{F} \otimes _ k I$. To see this is an action of $T\mathcal{F} \otimes _ k I$ on $\text{Lift}(x, A')$ we have to show the following: if $x', x'', x'''$ are lifts of $x$ and $g_*(x', x'') = \theta $, $g_*(x'', x''') = \theta '$, then $g_*(x', x''') = \theta + \theta '$. This follows from the commutative diagram

\[ \xymatrix{ A' \times _ A A' \times _ A A' \ar[rrrrr]_-{(a_1, a_2, a_3) \mapsto (g(a_1, a_2), g(a_2, a_3))} \ar[rrrrrd]_{(a_1, a_2, a_3) \mapsto g(a_1, a_3)} & & & & & k[I] \times _ k k[I] = k[I \times I] \ar[d]^{+} \\ & & & & & k[I] } \]

The action is free and transitive because $\Phi $ is bijective. $\square$

Comments (0)

Post a comment

Your email address will not be published. Required fields are marked.

In your comment you can use Markdown and LaTeX style mathematics (enclose it like $\pi$). A preview option is available if you wish to see how it works out (just click on the eye in the toolbar).

Unfortunately JavaScript is disabled in your browser, so the comment preview function will not work.

All contributions are licensed under the GNU Free Documentation License.

In order to prevent bots from posting comments, we would like you to prove that you are human. You can do this by filling in the name of the current tag in the following input field. As a reminder, this is tag 06JI. Beware of the difference between the letter 'O' and the digit '0'.