<?xml version="1.0" encoding="UTF-8"?>

<record version="4" id="2965">
 <title>adjoint representation</title>
 <name>AdjointRepresentation</name>
 <created>2002-05-29 09:14:59</created>
 <modified>2007-01-25 16:59:20</modified>
 <type>Definition</type>
 <creator id="146" name="rmilson"/>
 <author id="12884" name="Algeboy"/>
 <author id="146" name="rmilson"/>
 <classification>
	<category scheme="msc" code="17B10"/>
 </classification>
 <defines>
	<concept>adjoint action</concept>
	<concept>gl</concept>
	<concept>general linear Lie algebra</concept>
 </defines>
 <related>
	<object name="IsotropyRepresentation"/>
 </related>
 <preamble>\usepackage{amsmath}
\usepackage{amsfonts}
\usepackage{amssymb}
\newcommand{\reals}{\mathbb{R}}
\newcommand{\natnums}{\mathbb{N}}
\newcommand{\cnums}{\mathbb{C}}
\newcommand{\znums}{\mathbb{Z}}
\newcommand{\lp}{\left(}
\newcommand{\rp}{\right)}
\newcommand{\lb}{\left[}
\newcommand{\rb}{\right]}
\newcommand{\supth}{^{\text{th}}}
\newtheorem{proposition}{Proposition}
\newtheorem{definition}[proposition]{Definition}
\newcommand{\nl}[1]{\PMlinkescapetext{{#1}}}
\newcommand{\pln}[2]{\PMlinkname{#1}{#2}}
\newcommand{\lag}{\mathfrak{g}}
\newcommand{\ad}{\mathop{\mathrm{ad}}\nolimits}
\newcommand{\End}{\mathop{\mathrm{End}}\nolimits}</preamble>
 <content>Let $\lag$ be a Lie algebra.  For every $a\in\lag$ we define the
\PMlinkescapetext{{\em  adjoint endomorphism}}, a.k.a. the {\em adjoint action},
$$\ad(a):\lag\rightarrow\lag$$
to be the linear transformation with
action
$$\ad(a): b\mapsto [a,b],\quad b\in\lag.$$

For any vector space $V$, we use $\mathfrak{gl}(V)$ to denote the Lie algebra 
of $\End V$ determined by the commutator bracket.  So 
$\mathfrak{gl}(V)=\End V$ as vector spaces, only the multiplications are different.  

In this notation, treating $\mathfrak{g}$ as a vector space, the linear mapping $\ad:\lag\rightarrow \mathfrak{gl}(\lag)$ with action $$a\mapsto \ad(a),\quad a\in\lag$$
is called the {\em adjoint representation} of $\lag$.  The fact that
$\ad$ defines a representation is a straight-forward consequence of
the Jacobi identity axiom.  Indeed, let $a,b\in \lag$ be given.  We
wish to show that
$$\ad([a,b]) = [\ad(a),\ad(b)],$$
where the bracket on the left is the
$\lag$ multiplication structure, and the bracket on the right is the
commutator bracket.  For all $c\in\lag$ the left hand side maps $c$ to
$$[[a,b],c],$$
while the right hand side maps $c$ to
$$[a,[b,c]]+[b,[a,c]].$$
Taking skew-symmetry of the bracket as a
given, the equality of these two expressions is logically equivalent
to the Jacobi identity:
$$[a,[b,c]] +[b,[c,a]] + [c,[a,b]] = 0.$$</content>
</record>
