A Lie algebra weight is a linear functional arising from a simultaneous eigenspace in a Lie
algebra representation. To illustrate, consider a collection of diagonal
matrices , which span
a subspace
. Then the
th eigenvalue, i.e., the
th entry along the diagonal,
is a linear functional on
, and is called a weight.
The general setting for weights occurs in a finite-dimensional Lie
algebra representation of a finite-dimensional
complex semisimple Lie algebra , in which case the Cartan
subalgebra
is abelian and its representing matrices
can be simultaneously diagonalized. For
example, consider the standard Lie algebra
representation of the special linear
Lie algebra
on
. Then
|
(1)
|
and
|
(2)
|
span the Cartan subalgebra . There are three weights,
|
(3)
|
|
(4)
|
and
|
(5)
|
corresponding to the decomposition of
|
(6)
|
into its eigenspaces. Note that , because the matrices
have zero matrix trace.
The eigenvectors
are called weight vectors, and the corresponding
eigenspaces are called weight spaces.
In the important special case of the adjoint representation of , the nonzero
weights are called Lie algebra roots and the weight
space is called the root space. The roots generate a discrete lattice, called the
root lattice, in the dual
vector space
. The set
of all possible weights forms a weight lattice, which contains the root
lattice. The finite-dimensional Lie algebra representations of
can be classified using the weight lattice.
For a real semisimple Lie algebra, this description is applied to its complexification.