Representation Theory
Arun Ram
Department of Mathematics and Statistics
University of Melbourne
Parkville, VIC 3010 Australia
aram@unimelb.edu.au
Last updated: 2 October 2014
Lecture 5
Review
The Lie algebra
𝔰 𝔩 2
=
{ ( a b c d ) | a + d = 0 }
with bracket
[ x , y ] = x y - y x ,
for x , y ∈ 𝔰 𝔩 2 is presented by generators
e = ( 0 1 0 0 ) ,
f = ( 0 0 1 0 ) ,
h = ( 1 0 0 - 1 )
and relations
[ e , f ] = h ,
[ h , e ] = 2 e ,
[ h , f ] = - 2 f .
The enveloping algebra U 𝔰 𝔩 2 is generated by
e , f , h with relations
e f = f e + h ,
e h = h e - 2 e ,
h f = f h - 2 f
and has basis
{
f m 1
h m 2
e m 3
|
m 1 , m 2 , m 3 ∈ ℤ ≥ 0
}
.
If M = span { m 1 , … , m r }
and N = { n 1 , … , n s }
are U 𝔰 𝔩 2 -modules, then
M ⊗ N = span
{
m i ⊗ n j |
1 ≤ i ≤ r , 1 ≤ j ≤ s
}
is a U 𝔰 𝔩 2 -module, with
e ( m i ⊗ n j )
=
e m i ⊗ n j +
m i ⊗ e n j ,
f ( m i ⊗ n j )
=
f m i ⊗ n j +
m i ⊗ f n j ,
h ( m i ⊗ n j )
=
h m i ⊗ n j +
m i ⊗ h n j .
The quantum group U q 𝔰 𝔩 2
U q 𝔰 𝔩 2 is generated by
E , F , K ± 1 with relations
K E K - 1 = q 2 E ,
K F K - 1 = q - 2 F ,
E F = F E + K - K - 1 q - q - 1 .
The map Δ : U → U ⊗ U given by
Δ ( E )
=
E ⊗ K + 1 ⊗ E ,
Δ ( F )
=
F ⊗ 1 + K - 1 ⊗ F ,
Δ ( K )
=
K ⊗ K ,
is a coproduct .
U = U q 𝔰 𝔩 2 at
q = 1 is U 𝔰 𝔩 2 .
U = U q 𝔰 𝔩 2 has a
2-dimensional simple module
V = L ( ▫ ) = span
{ v 1 , v - 1 }
with
K v 1
=
q v 1 ,
K v - 1
=
q - 1 v - 1 ,
E v 1
=
0 ,
E v - 1
=
v 1 ,
F v 1
=
v - 1 ,
F v - 1
=
0 .
So ρ ▫ : U → End ( L ( ▫ ) ) has
ρ ▫ ( K ) = ( q 0 0 q - 1 ) ,
ρ ▫ ( E ) = ( 0 1 0 0 ) ,
ρ ▫ ( F ) = ( 0 0 1 0 ) .
Computing V ⊗ V = L ( ▫ ) ⊗ L ( ▫ ) = V ⊗ 2
V ⊗ 2 = V ⊗ V = span
{ v 1 ⊗ v 1 , v 1 ⊗ v - 1 , v - 1 ⊗ v 1 , v - 1 ⊗ v - 1 }
with
E ( v 1 ⊗ v 1 )
=
0 ,
E ( v 1 ⊗ v - 1 )
=
v 1 ⊗ v 1 ,
K ( v 1 ⊗ v 1 )
=
q 2 v 1 ⊗ v 1 ,
K ( v 1 ⊗ v - 1 )
=
v 1 ⊗ v - 1 ,
F ( v 1 ⊗ v 1 )
=
v - 1 ⊗ v 1 +
q - 1 v 1 ⊗ v - 1 ,
F ( v 1 ⊗ v - 1 )
=
v - 1 ⊗ v - 1 ,
E ( v - 1 ⊗ v 1 )
=
q v 1 ⊗ v 1 ,
E ( v - 1 ⊗ v - 1 )
=
q - 1 v 1 ⊗ v - 1
+ v - 1 ⊗ v 1 ,
K ( v - 1 ⊗ v - 1 )
=
q - 2 v - 1 ⊗ v - 1 ,
K ( v - 1 ⊗ v 1 )
=
v - 1 ⊗ v 1 ,
F ( v - 1 ⊗ v 1 )
=
q v - 1 ⊗ v - 1 ,
F ( v - 1 ⊗ v - 1 )
=
0 ,
or, equivalently,
( ρ ▫ ⊗ ρ ▫ ) ( E )
=
ρ ▫ ( E ) ⊗
ρ ▫ ( K ) +
ρ ▫ ( 1 ) ⊗
ρ ▫ ( E )
=
( 0 1 0 0 ) ⊗
( q 0 0 q - 1 ) +
( 1 0 0 1 ) ⊗
( 0 1 0 0 )
=
(
0 · ( q q - 1 )
1 · ( q q - 1 )
0 · ( q q - 1 )
0 · ( q q - 1 )
)
+
(
1 · ( 0 1 0 0 )
0 · ( 0 1 0 0 )
0 · ( 0 1 0 0 )
1 · ( 0 1 0 0 )
)
=
(
q
q - 1
q - 1
a a
)
+
(
0 1
0 0
0 1
0 0
)
=
(
0 1 q 0
0 0 0 q - 1
0 0 0 1
0 0 0 0
)
.
In general, if
A =
(
a 11 ⋯ a 1 r
⋮
a r 1 ⋯ a r r
)
and B =
(
b 11 ⋯ b 1 s
⋮
b s 1 ⋯ b s s
)
acting on M = span { m 1 , … , m r }
and N = span { n 1 , … , n s }
respectively then, if
( A ⊗ B ) ( m i ⊗ n j ) =
A m i ⊗ B n j ,
then the matrix of A ⊗ B in the basis
m 1 ⊗ n 1 ,
… ,
m 1 ⊗ n s ,
m 2 ⊗ n 1 ,
… ,
m 2 ⊗ n s ,
… ,
m r ⊗ m 1 ,
… ,
m r ⊗ n s
is
A ⊗ B =
(
a 11 B a 12 B ⋯ a 1 r B
⋮ ⋮
a r 1 B ⋯ a r r B
)
.
Decomposing V ⊗ 2
v 1 ⊗ v 1
↓ F
v - 1 ⊗ v 1 + q - 1 v 1 ⊗ v - 1
↓ F
[ 2 ] v - 1 ⊗ v - 1
F ( v 1 ⊗ v 1 )
=
v - 1 ⊗ v 1 +
q - 1 v 1 ⊗ v - 1 ,
F ( v - 1 ⊗ v 1 + q - 1 v 1 ⊗ v - 1 )
=
[ 2 ] v - 1 ⊗ v - 1 .
v - 1 ⊗ v 1 - q v 1 ⊗ v - 1
E ( v - 1 ⊗ v 1 - q v 1 ⊗ v - 1 )
=
0 ,
F ( v - 1 ⊗ v 1 - q v 1 ⊗ v - 1 )
=
0 .
Let
b 1 = v 1 ⊗ v 1 ,
b 2 = v - 1 ⊗ v 1 + q - 1 v 1 ⊗ v - 1 ,
b 3 = v - 1 ⊗ v - 1 ,
b 4 = v - 1 ⊗ v 1 - q v 1 ⊗ v - 1 .
Then
V ⊗ 2 = L ( ▫ )
⊗ L ( ▫ ) = L
(
)
⊕ L ( ∅ )
where
L
(
)
= span { b 1 , b 2 , b 3 }
and L ( ∅ ) =
span { b 4 } .
In the basis b 1 , b 2 , b 3 , b 4
the matrices for the action of E , F , K on
V ⊗ V are
ρ ⊕ ∅
( E )
=
(
0 [ 2 ]
0 1
0
0
)
,
ρ ⊕ ∅
( F )
=
(
0
1 0
[ 2 ] 0
0
)
,
ρ ⊕ ∅
( K )
=
(
q 2
q 0
q - 2
1
)
.
Decomposing V ⊗ 3 = L ( ▫ ) ⊗ L ( ▫ ) ⊗ L ( ▫ ) .
L ( ▫ ) ⊗
L ( ▫ ) ⊗
L ( ▫ ) = span
{
v 1 ⊗ v 1 ⊗ v 1 ,
v 1 ⊗ v 1 ⊗ v - 1 ,
v 1 ⊗ v - 1 ⊗ v 1 ,
v 1 ⊗ v - 1 ⊗ v - 1 ,
v - 1 ⊗ v 1 ⊗ v 1 ,
v - 1 ⊗ v 1 ⊗ v - 1 ,
v - 1 ⊗ v - 1 ⊗ v 1 ,
v - 1 ⊗ v - 1 ⊗ v - 1
}
.
Another basis of V ⊗ 3 is
{
b 1 ⊗ v 1 ,
b 1 ⊗ v - 1 ,
v 2 ⊗ v 1 ,
v 2 ⊗ v - 1 ,
v 3 ⊗ v 1 ,
v 3 ⊗ v - 1 ,
b 4 ⊗ v 1 ,
b 4 ⊗ v - 1
}
,
i.e.
V ⊗ 3 =
(
L ( )
⊕
L ( ∅ )
)
⊗ V =
( L ( ) ⊗ V ) ⊕
( L ( ∅ ) ⊗ V )
L ( ∅ ) ⊗ V = span
{ b 4 ⊗ v 1 , b 4 ⊗ v - 1 }
with
E ( b 4 ⊗ v 1 )
=
0 ,
E ( b 4 ⊗ v - 1 )
=
b 4 ⊗ v 1 ,
F 4 ( b 4 ⊗ v 1 )
=
b 4 ⊗ v - 1 ,
F ( b 4 ⊗ v - 1 )
=
0 ,
K ( b 4 ⊗ v 1 )
=
q b 4 ⊗ v 1 ,
K ( b 4 ⊗ v - 1 )
=
q - 1 b 4 ⊗ v - 1 .
So
L ( ∅ ) ⊗ V
≃
L ( ▫ )
b 4 ⊗ v 1
⟼
v 1
b 4 ⊗ v - 1
⟼
v - 1
Then
L ( ) ⊗ V = span
{
b 1 ⊗ v 1 ,
b 1 ⊗ v - 1 ,
b 2 ⊗ v 1 ,
b 2 ⊗ v - 1 ,
b 3 ⊗ v 1 ,
b 3 ⊗ v - 1
}
with
F ( b 1 ⊗ v 1 ) = b 2 ⊗ v 1 + q - 2 b 1 ⊗ v - 1 ,
b 1 ⊗ v 1
↓ F
b 2 ⊗ v 1 + q - 2 b 1 ⊗ v - 1
↓ F
[ 2 ] b 3 ⊗ v 1 + b 2 ⊗ v - 1 + q - 2 b 2 ⊗ v - 1 + q - 4 b 2 ⊗ 0 = [ 2 ] ( b 3 ⊗ v 1 + q - 1 b 2 ⊗ v - 1 )
↓ F
0 + [ 2 ] q 2 b 3 ⊗ v - 1 + [ 2 ] b 3 ⊗ v - 1 + 0 + q - 2 [ 2 ] b 3 ⊗ v - 1 + 0 = [ 2 ] [ 3 ] b 3 ⊗ v - 1
So, if
c 1
=
b 1 ⊗ v 1 ,
c 2
=
b 2 ⊗ v 1 + q - 2 b 1 ⊗ v - 1 ,
c 3
=
b 3 ⊗ v 1 +
q - 1 b 2 ⊗ v - 1 ,
c 4
=
b 3 ⊗ v - 1
then the action of F on
L ( ) ⊗ V = span
{ c 1 , c 2 , c 3 , c 4 }
is given by
ρ
( F ) =
(
0
1 0
[ 2 ] 0
[ 3 ] 0
)
and
ρ
( E )
=
(
0 1
0 [ 2 ]
0 [ 3 ]
0
)
,
ρ
( K )
=
(
q 3
q 1
q - 1
q - 3
)
.
Note that
0
↑ E
b 2 ⊗ v 1 - q b 1 ⊗ v - 1
F ↓
[ 2 ] b 3 ⊗ v 1 + b 2 ⊗ v - 1 - q [ 2 ] b 2 ⊗ v - 1 = [ 2 ] b 3 ⊗ v - 1 - q 2 b 2 ⊗ v - 1
F ↓
[ 2 ] q 2 b 3 ⊗ v - 1 - q 2 [ 2 ] b 3 ⊗ v - 1 = 0
So that, if
c 5
=
b 2 ⊗ v 1 - q b 1 ⊗ v - 1 ,
c 6
=
[ 2 ] b 3 ⊗ v - 1 - q 2
b 2 ⊗ v - 1
and
L ∼ ( ▫ ) =
span { c 5 , c 6 }
then
L ∼ ( ▫ ) ≃
L ( ▫ ) .
So
V ⊗ 3
=
L ( ▫ ) ⊗
L ( ▫ ) ⊗
L ( ▫ )
≅
(
L ( ) ⊗
L ( ∅ )
)
⊗ L ( ▫ )
=
( L ( ) ⊗ V ) ⊕
( L ( ∅ ) ⊗ V )
≃
L ( ) ⊕
L ( ▫ ) ⊕
L ( ▫ )
∅
∅
∅
1
1
1
2
1
2
3
1
5
4
1
1 · 2
=
2
1 · 1 + 1 · 3
=
4
2 · 2 + 1 · 4
=
8
2 · 1 + 3 · 3 + 1 · 5
=
16
5 · 2 + 4 · 4 + 2 · 6
=
32
What is the connection between T L k and V ⊗ k for U q 𝔰 𝔩 2 ?
Define an action of T L 2 = span { , }
on
V ⊗ 2 = span
{
v 1 ⊗ v 1 ,
v 1 ⊗ v - 1 ,
v - 1 ⊗ v 1 ,
v - 1 ⊗ v - 1
}
by
( v 1 ⊗ v 1 )
=
0 ,
( v - 1 ⊗ v - 1 )
=
0 ,
( v 1 ⊗ v - 1 )
=
q v 1 ⊗ v - 1 - v - 1 ⊗ v 1 ,
( v - 1 ⊗ v 1 )
=
q - 1 v - 1 ⊗ v 1
- v 1 ⊗ v - 1 .
In matrices we have
ρ ⊗ 2 ( ) =
(
0 0 0 0
0 q - 1 0
0 - 1 q - 1 0
0 0 0 0
)
.
Note that
( ρ ⊗ 2 ( ) ) 2 =
(
0 0 0 0
0 q 2 + 1 - ( q + q - 1 ) 0
0 - ( q + q - 1 ) 1 - q - 2 0
0 0 0 0
)
= [ 2 ]
ρ ⊗ 2 ( )
so that this is an action of T L 2 on V ⊗ 2 .
The T L 2 action commutes with the
U q 𝔰 𝔩 2 action on
V ⊗ 2 , i.e.
ρ ⊗ 2 ( T L 2 )
⊆ End U ( V ⊗ 2 ) .
The Temperley-Lieb algebra T L k is generated by
e j =
1
2
⋯
j
j + 1
⋯
k
,
1 ≤ j ≤ k - 1 .
Define an action of T L k on
V ⊗ k = span
{
v i 1 ⊗ ⋯ ⊗ v i k
|
i 1 , … , i k
∈ { ± 1 }
}
by
e j ( v i 1 ⊗ ⋯ ⊗ v i k ) =
v i 1 ⊗ ⋯ ⊗ v i j - 1 ⊗
( v i j ⊗ v i j + 1 ) ⊗
v i j + 2 ⊗ ⋯ ⊗ v i k .
Claim
(a)
This defines a T L k -action on V ⊗ k .
(b)
This T L k -action commutes with the
U q 𝔰 𝔩 2 -action on
V ⊗ k .
Let A be an algebra and let M be a semisimple A module,
M = ⨁ λ ∈ A ˆ
( A λ ) ⊕ m λ .
Let 𝒵 = End A ( M ) . Then
𝒵 = ⨁ λ ∈ M ˆ
M m λ ( ℂ ) , and
M ≃ ⨁ λ ∈ M ˆ
A λ ⊗ 𝒵 λ
as an ( A , 𝒵 ) bimodule, where M ˆ ⊆ A ˆ
is an index set for the simple A -modules appearing in M .
Proof.
𝒵
=
Hom A ( M , M )
=
Hom A
(
⨁ λ ∈ M ˆ
⨁ i = 1 m λ
A i λ ,
⨁ μ ∈ M ˆ
⨁ j = 1 m λ
A j μ
)
=
⨁ λ ∈ M ˆ
⨁ i , j = 1 m λ
Hom A
( A i λ , A j λ ) ,
by Schur's Lemma. Hence
𝒵 = { e i j λ | λ ∈ M ˆ , 1 ≤ i , j ≤ m λ }
where e i j λ : A i λ → A j λ
(choose e i i λ so that
( e i i λ ) 2 = e i i λ
and e i j λ and e j i λ
so that e i j λ e j i λ = e i i λ ).
□
Notes and References
These are a typed copy of Lecture 5 from a series of handwritten lecture notes for the class Representation Theory given on August 26, 2008.
page history