1
$\begingroup$

Do there exist results towards answering the following question?

Consider the Poisson algebra of regular functions $A=\mathbb R[V]$ on the symplectic vector space $V:=T^* \mathbb R^n$. Using canonical coordinates we can write $A=\mathbb R[q^1,\dots,q^n,p_1,\dots,p_n]$ with Poisson bracket being $\{q^i,p_j\}=\delta_j^i$ (all other brackets between coordinates being zero).

Question: What are the finite dimensional Lie subalgebras of the Poisson algebra $(A,\cdot,\{\:,\:\})$?

Partial answer: 1. There are subalgebras in degree 2 coming from moment maps of cotangent lifted representations. 2. There are a couple of Heisenberg subalgebras. 3. There are plenty of abelian subalgebras.

I wonder if one can give a complete classification using these building blocks.

| cite | improve this question | | | | |
$\endgroup$
  • $\begingroup$ The Poisson bracket is supposed to be a Lie bracket (valued in the vector space), but the formula you gives yields a scalar-valued bracket. What is the Lie bracket? $\endgroup$ – YCor Jul 25 '15 at 8:49
  • 2
    $\begingroup$ This problem is at least as hard as classifying the finite dimensional Lie subalgebras of the vector fields in $\mathbb{R}^n$. Even when one identifies two such algebras when they differ by a diffeomorphism, this latter problem is quite hard for $n=2$, and a complete list for $n=3$ (with various 'regularizing' assumptions) is still not done. As far as I know, no one has seriously attempted $n=4$. $\endgroup$ – Robert Bryant Jul 25 '15 at 10:20
  • $\begingroup$ Hello YCor, a scalar is a constant regular function, i.e. $\in A$. One can use the relation $\{q^i,p_j\}=\delta_j^i$ together with the Leibniz rule $\{a,bc\}=b\{a,c\}+\{a,b\}c$ to define $\{\:,\:\}:A\times A\to A$ since $A$ is generated as an algebra by the coordinate functions. By definition, the bracket of a Lie subalgebra arrises as the restriction of the Lie bracket of the ambient space, in this case $\{\:,\:\}$. $\endgroup$ – HCH Jul 25 '15 at 18:53
2
$\begingroup$

Using the Leibniz rule and the induction on $deg(f) + deg (g)$, we can see that the braket on $A$ is exactely the Poisson bracket defined by him in 1809 :) That is: $$ \{f, \, g\} = \sum_{i=1}^r \, \Bigl({\frac {\partial f} {\partial p_i}} {\frac {\partial g} {\partial q_i}} - {\frac {\partial f} {\partial q_i}} {\frac {\partial g} {\partial p_i}} \Bigl) $$ for all $f$, $g \in A$.

| cite | improve this answer | | | | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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