I would like to ask about the setting of the book "Representations of Semisimple Lie Algebras in the BGG Category $\mathcal{O}$" by Humphreys. I would like to know whether the semisimple Lie algebra $\mathfrak{g}$ is finite dimensional or not. I search the book and did not find any assumption about the semisimple Lie algebra $\mathfrak{g}$ being finite dimensional. However, the book mentions root space decomposition of $\mathfrak{g}$, which depends on the finite dimensionality of $\mathfrak{g}$ by Humphreys' other book "Introduction to Lie Algebras and Representation Theory". Also a Cartan subalgebra $\mathfrak{h}$ of $\mathfrak{g}$ is mentioned to be finite dimensional on p.1 of "Representations of Semisimple Lie Algebras in the BGG Category $\mathcal{O}$".

My questions:

  1. Is the semisimple Lie algebra $\mathfrak{g}$ finite dimensional?
  2. If $\mathfrak{g}$ is not assumed to be finite dimensional, do we still have the root space decomposition of complex semisimple Lie algebra with respect to a Cartan subalgebra?
  3. If $\mathfrak{g}$ is not assumed to be finite dimensional, does the root space decomposition imply that any complex semisimple Lie algebra with a Cartan subalgebra is finite dimensional?
  4. If $\mathfrak{g}$ is not assumed to be finite dimensional, how do we prove $\mathfrak{h}$ is finite dimensional?
  • $\begingroup$ Yes, the Lie algebra $\mathfrak{g}$ is finite dimensional. I'm sorry if this traditional setting isn't clear enough, but beyond this is mainly the Kac-Moody case which is not usually viewed as "semisimple". $\endgroup$ May 8, 2019 at 22:17
  • $\begingroup$ Thank you for your reply. $\endgroup$ May 9, 2019 at 6:42
  • $\begingroup$ By the way, I would like to know does $\mathfrak{g}$ is complex semisimple Lie algebra imply $\mathfrak{g}$ is finite dimensional by the root space decomposition or do people just omit the words "finite dimensional" when assuming $\mathfrak{g}$ is finite dimensional complex semisimple since it is a traditional setting? $\endgroup$ May 9, 2019 at 6:51

1 Answer 1

  1. In the book, $\mathfrak{g}$ is always assumed to be finite-dimensional. (I believe that this is explicitly stated in the first chapter of the book.)
  2. Depending on $\mathfrak{g}$ and on the definition of Cartan subalgebras in the infinite-dimensional setting, you may have an infinite-dimensional semisimple Lie algebra $\mathfrak{g}$ and a Cartan subalgebra $\mathfrak{h}$ such that, with respect to $\mathfrak{h}$, the root-space decomposition of $\mathfrak{g}$ exists. You can look at something like root-reductive Lie algebras.
  3. See my answer to 2.
  4. Also see my answer to 2.
  • $\begingroup$ Thank you for your answers. But I really cannot find out the assumption $\mathfrak{g}$ is finite dimensional throughout the book by manually searching the word "finite". I just find out some sentences implicitly suggest that $\mathfrak{g}$ is finite dimensional. e.g. p.36 (the last sentence) and p.259 (the last sentence in Section 13.6). If you can find it out, please let me know. $\endgroup$ May 8, 2019 at 12:15
  • $\begingroup$ Well, isn't a semisimple lie algebra just a (finite) direct sum of simple Lie algebras, each of which is finite dimensional. $\endgroup$
    – Filip
    Jul 27, 2021 at 23:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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