Given a symplectic vector space $V$, and a subspace $ \ U \subset V$, if $ \ U^\bot \not\subset \ U$, ($U \ $ is not co-isotropic), is it necessarily the case that $U \subseteq U^\bot \ $ or $ \ U \cap U^\bot = {0}$?

Another way of phrasing it would be, is containment, equality and trivial intersection the only possibilities for a subspace and its (symplectic) complement?

(For the technical definitions: are the cases isotropic, co-isotropic, Lagrangian and symplectic the complete list of possibilities for a subspace and its complement?)


Let $(V,\Omega)$ be a symplectic vector space of $\text{dim} V=2n$, with basis $e_1,...,e_n,f_1,...,f_n$, let us consider the four subspaces.

Isotropic: $\mathscr{L}\{e_1,e_2\}$
Symplectic: $\mathscr{L}\{e_1,f_1\}$
Lagrangian: $\mathscr{L}\{e_1,..,e_n\}$
Co-isotropic: $\mathscr{L}\{e_1,e_2,...,e_n,f_1\}$

By definition, a Lagrangian subspace is both isotropic and co-isotropic. Also if $Y\nsubseteq Y^\Omega$, then either $Y^\Omega\subseteq Y$ or $Y\cap Y^\Omega=\{0\}$. So these are the only possibilities. Another way to check this is to ask, can you come up with a subspace which doesn't have any of the above four subspace properties ?

  • $\begingroup$ What about $W =\mathscr L \{e_1,e_2,f_1\}$? $W^\bot = \mathscr L \{e_2,e_3,\ldots, e_n, f_3,\ldots, f_n\}$, so it is not one of the above. $\endgroup$ – mlainz Jan 16 at 14:25

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.