A Method for Weight Multiplicity Computation Based on Berezin Quantization
Let G be a compact semisimple Lie group and T be a maximal torus of G. We describe a method for weight multiplicity computation in unitary irreducible representations of G, based on the theory of Berezin quantization on G/T. Let Γhol(Lλ) be the reproducing kernel Hilbert space of holomorphic section...
Збережено в:
| Дата: | 2009 |
|---|---|
| Автор: | |
| Формат: | Стаття |
| Мова: | English |
| Опубліковано: |
Інститут математики НАН України
2009
|
| Назва видання: | Symmetry, Integrability and Geometry: Methods and Applications |
| Онлайн доступ: | http://dspace.nbuv.gov.ua/handle/123456789/149125 |
| Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Цитувати: | A Method for Weight Multiplicity Computation Based on Berezin Quantization / D. Bar-Moshe // Symmetry, Integrability and Geometry: Methods and Applications. — 2009. — Т. 5. — Бібліогр.: 21 назв. — англ. |
Репозитарії
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-149125 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-1491252019-02-20T01:25:25Z A Method for Weight Multiplicity Computation Based on Berezin Quantization Bar-Moshe, D. Let G be a compact semisimple Lie group and T be a maximal torus of G. We describe a method for weight multiplicity computation in unitary irreducible representations of G, based on the theory of Berezin quantization on G/T. Let Γhol(Lλ) be the reproducing kernel Hilbert space of holomorphic sections of the homogeneous line bundle Lλ over G/T associated with the highest weight λ of the irreducible representation πλ of G. The multiplicity of a weight m in πλ is computed from functional analytical structure of the Berezin symbol of the projector in Γhol(Lλ) onto subspace of weight m. We describe a method of the construction of this symbol and the evaluation of the weight multiplicity as a rank of a Hermitian form. The application of this method is described in a number of examples. 2009 Article A Method for Weight Multiplicity Computation Based on Berezin Quantization / D. Bar-Moshe // Symmetry, Integrability and Geometry: Methods and Applications. — 2009. — Т. 5. — Бібліогр.: 21 назв. — англ. 1815-0659 2000 Mathematics Subject Classification: 22E46; 32M05; 32M10; 53D50; 81Q70 http://dspace.nbuv.gov.ua/handle/123456789/149125 en Symmetry, Integrability and Geometry: Methods and Applications Інститут математики НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| description |
Let G be a compact semisimple Lie group and T be a maximal torus of G. We describe a method for weight multiplicity computation in unitary irreducible representations of G, based on the theory of Berezin quantization on G/T. Let Γhol(Lλ) be the reproducing kernel Hilbert space of holomorphic sections of the homogeneous line bundle Lλ over G/T associated with the highest weight λ of the irreducible representation πλ of G. The multiplicity of a weight m in πλ is computed from functional analytical structure of the Berezin symbol of the projector in Γhol(Lλ) onto subspace of weight m. We describe a method of the construction of this symbol and the evaluation of the weight multiplicity as a rank of a Hermitian form. The application of this method is described in a number of examples. |
| format |
Article |
| author |
Bar-Moshe, D. |
| spellingShingle |
Bar-Moshe, D. A Method for Weight Multiplicity Computation Based on Berezin Quantization Symmetry, Integrability and Geometry: Methods and Applications |
| author_facet |
Bar-Moshe, D. |
| author_sort |
Bar-Moshe, D. |
| title |
A Method for Weight Multiplicity Computation Based on Berezin Quantization |
| title_short |
A Method for Weight Multiplicity Computation Based on Berezin Quantization |
| title_full |
A Method for Weight Multiplicity Computation Based on Berezin Quantization |
| title_fullStr |
A Method for Weight Multiplicity Computation Based on Berezin Quantization |
| title_full_unstemmed |
A Method for Weight Multiplicity Computation Based on Berezin Quantization |
| title_sort |
method for weight multiplicity computation based on berezin quantization |
| publisher |
Інститут математики НАН України |
| publishDate |
2009 |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/149125 |
| citation_txt |
A Method for Weight Multiplicity Computation Based on Berezin Quantization / D. Bar-Moshe // Symmetry, Integrability and Geometry: Methods and Applications. — 2009. — Т. 5. — Бібліогр.: 21 назв. — англ. |
| series |
Symmetry, Integrability and Geometry: Methods and Applications |
| work_keys_str_mv |
AT barmoshed amethodforweightmultiplicitycomputationbasedonberezinquantization AT barmoshed methodforweightmultiplicitycomputationbasedonberezinquantization |
| first_indexed |
2025-07-12T21:26:03Z |
| last_indexed |
2025-07-12T21:26:03Z |
| _version_ |
1837477994004742144 |
| fulltext |
Symmetry, Integrability and Geometry: Methods and Applications SIGMA 5 (2009), 091, 12 pages
A Method for Weight Multiplicity Computation
Based on Berezin Quantization
David BAR-MOSHE
Dune Medical Devices Ltd., P.O. Box 3131, Caesarea Industrial Park, Israel
E-mail: david@dunemedical.com
Received July 26, 2009, in final form September 16, 2009; Published online September 25, 2009
doi:10.3842/SIGMA.2009.091
Abstract. Let G be a compact semisimple Lie group and T be a maximal torus of G. We
describe a method for weight multiplicity computation in unitary irreducible representations
of G, based on the theory of Berezin quantization on G/T . Let Γhol(Lλ) be the reproducing
kernel Hilbert space of holomorphic sections of the homogeneous line bundle Lλ over G/T
associated with the highest weight λ of the irreducible representation πλ of G. The multi-
plicity of a weight m in πλ is computed from functional analytical structure of the Berezin
symbol of the projector in Γhol(Lλ) onto subspace of weight m. We describe a method of
the construction of this symbol and the evaluation of the weight multiplicity as a rank of
a Hermitian form. The application of this method is described in a number of examples.
Key words: Berezin quantization; representation theory
2000 Mathematics Subject Classification: 22E46; 32M05; 32M10; 53D50; 81Q70
Dedicated to the memory of Professor Michael Marinov
1 Introduction
Computation of weight multiplicities is a necessary step to the construction of compact semi-
simple Lie group representations. There are several known formulas and methods for the com-
putation of weight multiplicities, such as Freudenthal recursive formula [17], the Kostant for-
mula [15], the Littelman’s path model [16], and the Billey–Guillemin–Rassart vector partition
function [9]. Also, geometric quantization and the orbit method offer many elegant multiplicity
formulas for group actions, see for example [13, 12, 20, 21]. The combination of the combinatorial
and geometric methods leads to algorithms for multiplicity computation [2].
The present work is also based on quantization methods on coadjoint orbits, namely the
Berezin quantization [7] on G/T , however, the method of multiplicity computation, presented
here, does not involve a direct combinatorial computation but is rather based on the functional
analytical structures of the reproducing kernels on the quantization spaces.
The application of Berezin quantization to the restriction of a unitary irreducible represen-
tation of a compact Lie group to a closed subgroup has been studied before in [1, 10]. The work
in the present paper has been generalized and extended in [11], which is mainly based on [1, 10]
and on the arXiv version of the present work.
Let λ be a dominant weight of the Lie algebra g of G and πλ be the corresponding unitary
irreducible representation of G. Let Lλ be the associated homogeneous holomorphic line bundle
over G/T , then according to the Borel–Weil theorem [14], the reproducing kernel Hilbert space
Γhol(Lλ) of holomorphic sections of Lλ is a G-irreducible carrier space of the representation πλ.
Berezin quantization (in its generalized version on spaces of sections of line bundles [18])
leads to the realization of πλ [5] in terms of the Berezin symbols [8] which act as integration
kernels on Γhol(Lλ).
mailto:david@dunemedical.com
http://dx.doi.org/10.3842/SIGMA.2009.091
2 D. Bar-Moshe
We describe a method for the construction of the Berezin’s symbol of projector in Γhol(Lλ)
onto the subspace of weight m. The functional analytic structure of this projector expressed in
the affine coordinates of the big Schubert cell of G/T enables the computation of the weight
multiplicity of m as the rank of a Hermitian form with rational coefficients.
The main results of the present work are given in the following propositions:
Proposition 1. Let G be compact semisimple Lie group, T a maximal torus of G, and G/T its
fundamental projective space. Let πλ be the irreducible unitary representation of G of highest
weight λ. Then the multiplicity γλ(m) of the vector of weight m in the representation space
of πλ is given by:
γλ(m) =
Nλ
V
∫
G/T
∫
T
(m− λ)(h) exp
(
Kλ(h · z, z)
)
exp
(
−Kλ(z, z)
)
dµ(h)dµ(z, z),
where Nλ is the dimension of the irreducible representation πλ, dµ(z, z) is the Liouville measure
on G/T , V is its total mass, dµ(h) is the Lebesgue measure on T (normalized to unit total
mass), µ(h) is the character representation of h ∈ T corresponding to the integral weight µ and
Kλ(ζ, z) : G/T × G/T −→ C is the analytic continuation of a Kähler potential associated with
the first Chern class of the holomorphic homogeneous line bundle Lλ over G/T .
The application of this formula to weight multiplicity computation can be simplified due to
the following two propositions:
Proposition 2. The Berezin principal symbol of the projector onto the subspace of weight m,
in the representation space of πλ, is given by:
Lλ
m(ζ, z) =
∫
T
(m− λ)(h) exp
(
Kλ(h · ζ, z)
)
dµ(h).
The Berezin principal symbol of the projector onto the subspace of weight m, restricted to
the largest Schubert cell Σs of G/T is polynomial in the affine coordinates of Σs in both of its
arguments [5, 6]. This parametrization is used to compute the multiplicity of the weight m in πλ
as follows:
Proposition 3. Let dα be the polynomial degree of coordinate zα in the restriction of Lλ
m(ζ, z)
to the largest Schubert cell Σs of G/T . Then the monomials:
un =
∏
α∈∆+
(ζα)nα , vn =
∏
α∈∆+
(zα)nα , 0 ≤ nα ≤ dα,
n = (nα1 , nα2 , . . . , nαD), D = dimCG/T,
define a biholomorphic transformation f : Σs × Σs −→ V× V, V ∼= Cd, d =
∑
α∈∆+ dα. (∆+ is
the set of positive roots of g.) Let Lλ
m(u, v) be the unique Hermitian form (linear in the first
argument and antilinear in the second), such that: (f ∗ Lλ
m)(ζ, z) = Lλ
m(ζ, z), then multiplicity
is given by:
γλ(m) = rankV
(
Lλ
m
)
.
In Section 2, a brief review of the application of Berezin quantization on G/T to the con-
struction of unitary irreducible representations of G, is given. In Section 3, the formula of the
projector onto the subspace of a given weight and the integration formula for weight multi-
plicity based the Berezin theory are developed. In Section 4, the method of computation of
a weight multiplicity, based on Proposition 3, is described. In Appendix A, two examples of the
application of the present method are given.
A Method for Weight Multiplicity Computation Based on Berezin Quantization 3
2 Berezin quantization and the Borel–Weil theory
2.1 Notations
G is a compact semisimple Lie group;
T is a maximal torus of G;
Gc is the complexification of G;
G/T is a fundamental projective pace of G;
W is the Weyl group of G;
B is a Borel subgroup of containing T ;
N is the unipotent radical of B;
N− is the subgroup of opposite to N (N− = sNs−1,where s ∈ Norm(T ) is any representa-
tive of the unique maximal length element of W );
g is the Lie algebra of G;
t is the Cartan subalgebra of g corresponding to T ;
gc is the Lie algebra of Gc;
n is the Lie algebra of N ;
n− is the Lie algebra of N−;
g∗ is the dual space of g;
〈·, ·〉 is the duality map g∗ × g −→ C;
∆ ⊂ g∗ is the set of roots of t in gc;
∆+ ⊂ ∆ is the set of positive roots (we choose ∆+ as the set of roots of n−);
Σ ⊂ ∆+ is the set of primitive roots (Σ =
{
γj , j = 1, . . . , rank(g)
}
);
{Eα, α ∈ ∆+} is the set of positive root generators of (generators of n−);
W is the Weight lattice of G (λ ∈ W is an integral weight of T );
C ⊂ W is the positive Weyl chamber W (C = {λ ∈ W | λ · α ≥ 0, ∀ α ∈ ∆+}, where λ · α
denotes the inner product induced on W by the Cartan–Killing form);
{wj ∈ W, j = 1, . . . , rank(g)} is the set primitive weights (wj · γi = δi
j).
2.2 The quantization space in Berezin theory
Let Lλ = (Gc × Cλ)/B be the homogenous holomorphic line bundle over G/T associated with
the dominant weight λ, defined as the set of equivalence classes of elements of Gc×C under the
equivalence relation:
(gb, ψ) ∼ (g, λ(b)ψ).
(The equivalence class of (g, ϕ) is denoted by: [g, ϕ].) Here, g ∈ Gc, b ∈ B, ϕ ∈ C, λ ∈ C,
and λ(b) is the character representation B −→ C×, defined by:
λ(exp(iH)) = exp(i 〈λ,H〉), H ∈ t,
λ(n) = 1, n ∈ N.
These representations (for all λ ∈ W) exhaust all one dimensional representations of B, since B
and T have the same fundamental group and any one dimensional representation of B must be
trivial on N .
4 D. Bar-Moshe
According to the Borel–Weil theorem [14]: if λ is dominant, then the G action on Lλ,
defined by:
g1 · [g, ϕ] = [g1g, ϕ] ,
is irreducible on the space of holomorphic sections Γhol(Lλ), and the corresponding represen-
tation is equivalent to the representation πλ of highest weight λ of G. Berezin quantization
provides a method for realizing πλ on Γhol(Lλ) [5]. The representation space in Berezin theory
is realized as a reproducing kernel Hilbert space. Specifically, let ψ : G/T −→ Lλ be a holomor-
phic section of Lλ, let (z, ψ(z)) be the trivialization of ψ on some open neighborhood U . Then
the inner product on the space of sections in the Berezin theory is defined by:
(ψ1, ψ2) =
Nλ
V
∫
G/T
ψ1(z)ψ2(z) exp
(
−K(λ)(z, z)
)
dµ(z, z),
where the sum over an appropriate open cover of G/T weighted by a partition of unity [18] is
implicit. Here, ψ1, ψ2 ∈ L2(Γhol(Lλ)), Nλ is the dimension of the representation πλ, and V is
the total mass of the Liouville measure dµ(z, z). We note that due to the compatibility of the
connection on Lλ defined by the Kähler potential and the Hermitian metric on the Fibers, all
the integrations over G/T in this work are of global functions.
The representatives of group elements and of elements of the universal enveloping algebra of
its Lie algebra are realized in the Berezin theory by means of symbols, which act as integration
kernels on L2(Γhol(Lλ)). The action of an operator A in the representation space of πλon
L2(Γhol(Lλ)) is realized by means of its symbol Aλ(ζ, z) : G/T ×G/T −→ C as follows:
(πλ(A) ◦ ψ)(ζ) =
Nλ
V
∫
G/T
Aλ(ζ, z) exp
(
K(λ)(ζ, z)
)
ψ(z) exp
(
−K(λ)(z, z)
)
dµ(z, z).
We define the covariant principal Berezin symbol of A by:
Ãλ(ζ, z) = Aλ(ζ, z) exp
(
K(λ)(ζ, z)
)
.
In the parametrization we use, the principal covariant symbols have the property that their re-
striction to the largest Schubert cell Σs ⊂ G/T is polynomial in the affine coordinates of Σs [5, 6].
The reproduction property of is expressed through the relation:
ψ(ζ) =
Nλ
V
∫
G/T
Lλ(ζ, z)ψ(z) exp
(
−K(λ)(z, z)
)
dµ(z, z),
where Lλ(ζ, z) is the reproducing kernel of L2(Γhol(Lλ)). The reproducing kernel can be viewed
as the principal covariant symbol of the unit operator in L2(Γhol(Lλ)). On the other hand it is
the kernel of the orthogonal projector on L2(Γhol(Lλ)) in L2(Γ(Lλ)).
Clearly:
Lλ(ζ, z) =
Nλ∑
j=1
ψ̂j(ζ)ψ̂j(z),
where
{
ψ̂j , j = 1, . . . , Nλ
}
is any orthonormal set of L2(Γhol(Lλ)). We choose to work in
the weight basis in which the basic vectors of the orthonormal set are indexed by their weight
vectors m corresponding to the Cartan generators and the corresponding degeneracy index im,
in terms of which the latter equation can be rephrased as:
Lλ(ζ, z) =
∑
m∈W λ
γλ(m)∑
im=1
ψ̂m,im(ζ)ψ̂m,im(z), (2.1)
A Method for Weight Multiplicity Computation Based on Berezin Quantization 5
where W λ is the weight set of πλ and γλ(m) is the multiplicity of the weight m in πλ. A basic
property of L2(Γhol(Lλ)), as a reproducing kernel Hilbert space, is that its reproducing kernel is
given in terms of the analytic continuation of the corresponding Kähler potential by the relation:
Lλ(ζ, z) = exp
(
Kλ(ζ, z)
)
. (2.2)
In Berezin’s original work [7] the condition (2.2) was imposed in order to establish a quantiza-
tion (correspondence principle). In the case under study of the reproducing kernel Hilbert spaces
of holomorphic sections of line bundles over compact Kähler manifolds, the validity of (2.2) was
established in [19].
We shall refer to the Kähler potentials
{
Kj(ζ, z), j = 1, . . . , rank(g)
}
corresponding the fun-
damental tensor representations, i.e., the representations of highest weights equal to fundamental
weights {wj ∈ W, j = 1, . . . , rank(g)}, as the basic Kähler potentials. The corresponding repro-
ducing kernels will be addressed to as the basic reproducing kernels:
Lj(ζ, z) ≡ Lwj (ζ, z) = exp(Kj(ζ, z)).
Therefore, the Kähler potential corresponding to a general highest weight representation is
given by:
Kλ(ζ, z) =
rank(g)∑
j=1
ljK
j(ζ, z), λ =
rank(g)∑
j=1
ljwj , lj ∈ N∪{0} , j = 1, . . . , rank(g).
2.3 The T action on G/T and L2(Γhol(Lλ))
G/T can be parametrized through the canonical diffeomorphism G/T ∼= Gc/B by a holomorphic
section ξ(z) : Gc/B −→ Gc of the principal bundle B −→ Gc −→ Gc/B . The T action on G/T
is given by:
ξ(h · z) = hξ(z)h−1, h ∈ T.
The induced T action on L2(Γhol(Lλ)) can be obtained from the Lλ property as a homoge-
neous bundle:
(h ◦ ψ)(z) = λ(h)ψ(h−1 · z). (2.3)
2.4 Parametrization of G/T and the construction
of the basic Kähler potentials
In this section and in the computational examples, all group and universal enveloping algebra
elements of Gc will be represented in the basic fundamental representation of G (i.e., operators
in the representation space of the basic fundamental representation of G).
We use the Bando–Kuratomo–Maskawa–Uehera method [3, 4] for the construction of the
basic Kähler potentials. Let {Yj ∈ t, j = 1, . . . , rank(g)} be the set of coweights of t defined by:
〈α, Yj〉 =
rank(g)∑
k=1
Gjk wk · α,
where Gjk = (A−1)ij
γ(i)·γ(j)
2 is metric on the weight space in the weight basis (A is the Cartan
matrix). In the terminology of [3] and [4], the generators Yj are called central charges. Let
{ηj ∈ t, j = 1, . . . , rank(g)} be the set of projectors on the lowest value eigenspace of Yj (in the
6 D. Bar-Moshe
basic fundamental representation). Then according to [3] and [4], the basic Kähler potentials
can be constructed according to:
Kj(ζ, z) = log
(
det
(
ηjξ(z)†ξ(ζ)ηj + 1− ηj
))
, j = 1, . . . , rank(g),
where (·)† denotes Hermitian conjugation.
For the specific computations, we use the parametrization of G/T , obtained from the holo-
morphic diffeomorphism: N− −→ Σs, given by:
ξ(z) = exp
( ∑
α∈∆+
zαEα
)
, zα ∈ C. (2.4)
3 A weight multiplicity formula based on Berezin quantization
The action on L2(Γhol(Lλ)) can be used to project the representation space onto the subspace
spanned by a given weight m, as follows.
On one hand, since ψ̂m,jm is a section corresponding to a vector of weight m, then:(
h ◦ ψ̂m,jm
)
(z) = m(h)ψ̂m,jm(z).
On the other hand, according to (2.3):(
h ◦ ψ̂m,jm
)
(z) = λ(h)ψ̂m,jm(h · z).
Combining the two equations, we obtain:
ψ̂m,jm(h · z) = (λ−m)(h)ψ̂m,jm(z).
Thus, the projection onto the subspace of a given weight m, can be obtained by the following
integration over T :∫
T
(m− λ)(h)ψ̂m′,jm′ (h · z)dµ(h) = δm,m′ψ̂m,jm(z),
where dµ(h) is the Lebesgue measure on T . Applying this projection operation to the reproduc-
ing kernel Lλ(ζ, z), we obtain:
γλ(m)∑
im=1
ψ̂m,im(ζ)ψ̂m,im(z) =
∫
T
(m− λ)(h)
( ∑
m∈W λ
γλ(m)∑
im=1
ψ̂m,im(h · ζ)ψ̂m,im(z)
)
dµ(h),
which, upon using (2.1) and (2.2), is the statement of Proposition 2:
Lλ
m(ζ, z) =
∫
T
(m− λ)(h) exp
(
Kλ(h · ζ, z)
)
dµ(h), (3.1)
where
Lλ
m(ζ, z) =
γλ(m)∑
im=1
ψ̂m,im(ζ)ψ̂m,im(z). (3.2)
The principal symbol Lλ
m(ζ, z) of the projector in L2(Γhol(Lλ)) onto the subspace spanned
by vectors of weight m acts as a reproducing kernel on the same subspace. Now, since the
sections ψ̂m,im are orthonormal, we get:
γλ(m) =
Nλ
V
∫
G/T
Lλ
m(ζ, z) exp
(
−K(λ)(z, z)
)
dµ(z, z). (3.3)
A Method for Weight Multiplicity Computation Based on Berezin Quantization 7
Combining (3.1) with (3.3) we obtain the statement of Proposition 1:
γλ(m) =
Nλ
V
∫
G/T
∫
T
(m− λ)(h) exp
(
Kλ(h · ζ, z)
)
exp
(
−Kλ(z, z)
)
dµ(h)dµ(z, z).
4 A method for weight multiplicity computation
4.1 Computation by direct integration
The weight multiplicity formula given in the previous section is not constructive. While the
construction of the Kähler potentials along the method described in Section 2 and the integra-
tion over the maximal torus are straightforward, no simple rules are known in advance for the
integration of holomorphic sections. Although, the integration can be performed by elementary
integration techniques on the largest Schubert cell, since its complement in G/T is of Liouville
measure zero, but even for spaces as small as Fl(3) ∼= SU(3)/S(U(1)× U(1)× U(1)) and small
representations, direct integration is a quite lengthy task.
4.2 Weight multiplicity computation as a rank of a Hermitian form
The decomposition of Lλ
m(ζ, z) given in (3.2) allows multiplicity computations without direct
integration. In the parametrization of G/T defined in (2.4), the restriction of any principal
Berezin symbol to the largest Schubert cell Σs is polynomial in the affine coordinates (ζα, zα).
Let dα be the polynomial degree of coordinate zα in the restriction of Lλ
m(ζ, z), and define the
monomials:
un(ζ) =
∏
α∈∆+
(ζα)nα , vn(z) =
∏
α∈∆+
(zα)nα , 0 ≤ nα ≤ dα,
n = (nα1 , nα2 , . . . , nαD), D = dimCG/T. (4.1)
Clearly, any section of the orthonormal subset of weight m is a linear combination of the
monomials of (4.1):
ψ̂m,im(ζ) =
∑
n
an
imun(ζ). (4.2)
The monomials in (4.1), define a biholomorphic transformation f : Σs × Σs −→ V × V,
V ∼= Cd, d =
∑
α∈∆+ dα. Consider the Hermitian form on V defined by:
Lλ
m(u, v) =
γλ(m)∑
im=1
(∑
n
an
imun
)(∑
n
an
imvn
)
.
We have: f ∗Lλ
m = Lλ
m. Suppose there exists another Hermitian form L′λ
m such that f ∗L′λ
m = Lλ
m.
This would imply that: f ∗ (Lλ
m − L′λ
m) = 0. But f ∗ (Lλ
m − L′λ
m) is a polynomial function on
Σs × Σs, therefore it can be identically 0 only if all of its coefficients vanish. This would imply
that the coefficients of the Hermitian form Lλ
m − L′λ
m vanish, hence L′λ
m = Lλ
m.
Let A be the matrix of dimension γλ(m)× d whose coefficients are an
im
. Clearly d ≥ γλ(m),
and also the row vectors of A are linearly independent, otherwise, at least one of the sections
in (4.2) is a linear combination of the others, which is impossible, since they are orthonormal.
Therefore,
rank(A) = γλ(m).
The Hermitian matrix of the Hermitian form Lλ
m is A†A, whose rank is identical to A, hence
we obtained the proof of Proposition 3:
γλ(m) = rankV
(
Lλ
m
)
.
8 D. Bar-Moshe
A Examples
A.1 Computation of some weight multiplicities
in the representation π(4,2) of SU(3)
A suitable basis for the generators of sl(3) = su(3)c in the basic fundamental representation is
given by:
Positive root generators:
Eα1 =
0 1 0
0 0 0
0 0 0
, Eα2 =
0 0 0
0 0 1
0 0 0
, Eα3 =
0 0 1
0 0 0
0 0 0
.
Negative root generators:
E−α1 = E†
α1
, E−α2 = E†
−α2
, E−α3 = E†
α3
.
Cartan subalgebra generators:
H1 = [Eα1 , E−α1 ] = diag(1,−1, 0) H2 = [Eα2 , E−α2 ] = diag(0, 1,−1).
Central charges:
Y1 = G11H1 +G12H2 = 1
3diag(2,−1,−1), Y2 = G21H1 +G22H2 = 1
3diag(1, 1,−2).
The projectors required:
η1 = diag(0, 1, 1), η2 = diag(0, 0, 1).
The coset representative:
ξ(z) = exp
(
3∑
i=1
ziEαi
)
=
1 z1 z3 − z1z2/2
0 1 z2
0 0 1
.
The action of the element h = exp(iθ1H1 + iθ2H2) ∈ T on the affine coordinates:
(h · z)1 = exp(2iθ1 − iθ2)z1, (h · z)2 = exp(−iθ1 + 2iθ2)z2,
(h · z)3 = exp(iθ1 + iθ2)z3.
The fundamental reproducing kernels:
L1(ζ, z) = det
(
η1ξ(z)†ξ(ζ)η1 + 1− η1
)
= 1 + ζ1z1 + ζ+
3 z
+
3 ,
L2(ζ, z) = det
(
η2ξ(z)†ξ(ζ)η2 + 1− η2
)
= 1 + ζ2z2 + ζ−3 z
−
3 ,
where ζ±3 = ζ3 ± ζ1ζ2/2, and similarly for the antiholomorphic coordinates.
The reproducing kernel of L2(Γhol(L(4,2))):
L(4,2)(ζ, z) = (1 + ζ1z1 + ζ+
3 z
+
3 )4(1 + ζ2z2 + ζ−3 z
−
3 )2.
A Method for Weight Multiplicity Computation Based on Berezin Quantization 9
A.1.1 Computation of the multiplicity of the weight (0, 1) in π(4,2)
Computation of the projector onto the weight space (−2,−1) in L2(Γhol(L(4,2))):
L
(4,2)
(−2,−1)(ζ, z) =
1
(2π)2
2π∫
0
dθ1
2π∫
0
dθ2 exp(−6iθ1 − 4iθ2)
× L(4,2)(exp(2iθ1 − iθ2)ζ1, exp(−iθ1 + 2iθ2)ζ2, exp(iθ1 + iθ2)ζ3, z1, z2, z3).
The integration result:
L
(4,2)
(−2,−1)(ζ, z) = 17
64ζ
5
1ζ
4
2z
5
1z
4
2 + 6ζ3
1ζ
2
2ζ
2
3z
3
1z
2
2z
2
3 + 20ζ1ζ4
3z1z
4
3 + 5
4ζ
4
1ζ
3
2ζ3z
4
1z
3
2z3
+ 20ζ2
1ζ2ζ
3
3z
2
1z2z
3
3 + 3
2ζ
4
1ζ
3
2ζ3z
3
1z
2
2z
2
3 + 3
2ζ
3
1ζ
2
2ζ
2
3z
4
1z
3
2z3 + 2ζ2
1ζ2ζ
3
3z
5
1z
4
2
+ 2ζ5
1ζ
4
2z
2
1z2z
3
3 − 1
2ζ
4
1ζ
3
2ζ3z
5
1z
4
2 − 1
2ζ
5
1ζ
4
2z
4
1z
3
2z3 − 5ζ4
1ζ
3
2ζ3z
2
1z2z
3
3
− 5ζ2
1ζ2ζ
3
3z
4
1z
3
2z3 − 6ζ2
1ζ2ζ
3
3z
3
1z
2
2z
2
3 − 6ζ3
1ζ
2
2ζ
2
3z
2
1z2z
3
3 − 6ζ1ζ4
3z
3
1z
2
2z
2
3
− 6ζ3
1ζ
2
2ζ
2
3z1z
4
3 + 2ζ1ζ4
3z
4
1z
3
2z3 + 2ζ4
1ζ
3
2ζ3z1z
4
3 + 1
4ζ1ζ
4
3z
5
1z
4
2 + 1
4ζ
5
1ζ
4
2z1z
4
3
− 8ζ2
1ζ2ζ
3
3z1z
4
3 − 8ζ1ζ4
3z
2
1z2z
3
3 − 9
8ζ
3
1ζ
2
2ζ
2
3z
5
1z
4
2 − 9
8ζ
5
1ζ
4
2z
3
1z
2
2z
2
3
The monomial set of L(4,2)
(−2,−1):
u1 = ζ1ζ
4
3 , u2 = ζ2
1ζ2ζ
3
3 , u3 = ζ3
1ζ
2
2ζ
2
3 , u4 = ζ4
1ζ
3
2ζ3, u5 = ζ5
1ζ
4
2 ,
v1 = z1z
4
3 , v2 = z2
1z2z
3
3 , v3 = z3
1z
2
2z
2
3 , v4 = z4
1z
3
2z3, v5 = z5
1z
4
2 .
The Hermitian form L(4,2)
(−2,−1)(u, v):
L(4,2)
(−2,−1)(u, v) =
v1
v2
v3
v4
v5
†
20 −8 −6 2 1
4
−8 20 −6 −5 2
−6 −6 6 3
2
−9
8
2 −5 3
2
5
4
−1
2
1
4 2 −9
8
−1
2
17
64
u1
u2
u3
u4
u5
.
Computation of the multiplicity:
γ(4,2)((−2,−1)) = rank
(
L(4,2)
(−2,−1)(u, v)
)
= 2.
A.1.2 Computation of the multiplicity of the weight (−6, 4) in π(4,2)
This weight lies on the Weyl group orbit of the highest weight, therefore its multiplicity should
be 1. Repeating the same type of computation as in the previous case, we obtain:
L
(4,2)
(−6,4)(ζ, z) = 1
256ζ
4
1ζ
2
3z
4
1z
2
3 + 1
64ζ
5
1ζ2ζ3z
5
1z2z3 + 1
256ζ
6
1ζ
2
2z
6
1z
2
2 + 1
128ζ
5
1ζ2ζ3z
4
1z
2
3
+ 1
256ζ
6
1ζ
2
2z
4
1z
2
3 + 1
128ζ
6
1ζ
2
2z
5
1z2z3 + 1
128ζ
4
1ζ
2
3z
5
1z2z3 + 1
256ζ
4
1ζ
2
3z
6
1z
2
2
+ 1
128ζ
5
1ζ2ζ3z
6
1z
2
2.
The monomial set of L(4,2)
(−6,4):
u1 = ζ4
1ζ
2
3 , u2 = ζ5
1ζ2ζ3, u3 = ζ6
1ζ
2
2 ,
v1 = z4
1z
2
3 , v2 = z5
1z2z3, v3 = z6
1z
2
2 .
10 D. Bar-Moshe
The Hermitian form L(4,2)
(−6,−4)(u, v):
L(4,2)
(−2,−1)(u, v) =
v1
v2
v3
†
1
256
1
128
1
256
1
128
1
64
1
128
1
256
1
128
1
256
u1
u2
u3
.
Computation of the multiplicity:
γ(4,2)((−6, 4)) = rank
(
L(4,2)
(−6,4)(u, v)
)
= 1.
A.2 Computation of some weight multiplicities
in the representation π(1,1) of SO(5)
We choose to work in the four dimensional basic fundamental representation of sp(2)∼= so(5).
We use a quaternionic basis for the generators, with:
1 =
(
1 0
0 1
)
, σ+ =
(
0 1
0 0
)
, σ− =
(
0 0
1 0
)
, σ0 =
(
1 0
0 −1
)
.
A suitable basis for the generators of sp(2)c in the basic fundamental representation is given
by:
Positive root generators:
Eα1 = 1⊗ σ+, Eα2 = σ+ ⊗ σ−, Eα3 = σ+ ⊗ σ0, Eα4 = σ+ ⊗ σ+.
Cartan subalgebra generators:
H1 = [Eα1 , E−α1 ] = 1⊗ σ0, H2 = [Eα2 , E−α2 ] = 1
2(σ0 ⊗ 1− 1⊗ σ0).
Central charges:
Y1 = G11H1 +G12H2 = 1
2(σ0 ⊗ 1 + 1⊗ σ0), Y2 = G21H1 +G22H2 = σ0 ⊗ 1.
The action of the element h = exp(iθ1H1 + iθ2H2) ∈ T on the affine coordinates:
(h · z)1 = exp(2iθ1 − iθ2)z1, (h · z)2 = exp(−2iθ1 + 2iθ2)z2,
(h · z)3 = exp(iθ2)z3, (h · z)4 = exp(2iθ1)z4.
The basic reproducing kernels:
L1(ζ, z) = 1 + ζ1z1 + ζ+
4 z
+
4 + ζ−3 z
−
3 ,
L2(ζ, z) = 1 + ζ2z2 + 2ζ+
3 z
+
3 + (ζ+
4 − ζ1ζ
+
3 )(z+
4 − z1z
+
3 ) + (ζ2ζ+
4 − ζ−3 ζ
+
3 )(z2z
+
4 − z−3 z
+
3 ),
where ζ±3 = ζ3 ± ζ1ζ2/2, ζ+
4 = ζ4 + ζ2
1ζ2/6 and similarly for the antiholomorphic coordinates.
The reproducing kernel of L2(Γhol(L(1,1))):
L(1,1)(ζ, z) = (1 + ζ1z1 + ζ+
4 z
+
4 + ζ−3 z
−
3 )
×
(
1 + ζ2z2 + 2ζ+
3 z
+
3 + (ζ+
4 − ζ1ζ
+
3 )(z+
4 − z1z
+
3 ) + (ζ2ζ+
4 − ζ−3 ζ
+
3 )(z2z
+
4 − z−3 z
+
3 )
)
.
The principal symbol of the projector onto the weight space (−1, 0) in L2(Γhol(L(1,1))):
L
(1,1)
(−1,0)(ζ, z) = 7
144ζ
3
1ζ
2
2z
3
1z
2
2 + 1
12ζ
2
1ζ2ζ3z
2
1z2z3 + 3ζ3ζ4z3z4 + 2ζ1ζ2
3z1z
2
3 + 7
4ζ1ζ2ζ4z1z2z4
+ 3
2ζ1ζ
2
3z1z2z4 + 1
3ζ
3
1ζ
2
2z3z4 − ζ1ζ
2
3z3z4 + 1
2ζ1ζ2ζ4z3z4 − 1
12ζ
3
1ζ
2
2z1z2z4
+ 1
12ζ
2
1ζ2ζ3z1z2z4 + 1
2ζ3ζ4z1z2z4 + 1
2ζ
2
1ζ2ζ3z3z4 − ζ3ζ4z1z
2
3 − 1
4ζ
3
1ζ
2
2z1z
2
3
+ 1
18ζ
3
1ζ
2
2z
2
1z2z3 − 1
6ζ1ζ
2
3z
2
1z2z3 + 1
2ζ3ζ4z
2
1z2z3 − 1
6ζ
2
1ζ2ζ3z1z
2
3 − 3
2ζ1ζ2ζ4z1z
2
3
− 1
4ζ1ζ
2
3z
3
1z
2
2 + 1
18ζ
2
1ζ2ζ3z
3
1z
2
2 − 1
12ζ1ζ2ζ4z
3
1z
2
2 + 1
12ζ1ζ2ζ4z
2
1z2z3 + 1
3ζ3ζ4z
3
1z
2
2.
A Method for Weight Multiplicity Computation Based on Berezin Quantization 11
The monomial set:
u1 = ζ3
1ζ
2
2 , v1 = z3
1z
2
2 ,
u2 = ζ2
1ζ2ζ3, v2 = z2
1z2z3,
u3 = ζ3ζ4, v3 = z2
1z2z3,
u4 = ζ1ζ
2
3 , v4 = z2
1z2z3,
u5 = ζ1ζ2ζ4, v5 = z2
1z2z3.
The Hermitian form L(1,1)
(−1,0)(u, v):
L(1,1)
(−1,0)(u, v) =
v1
v2
v3
v4
v5
†
7
144
1
18
1
3 −1
4
−1
12
1
18
1
12
1
2 −1
6
1
12
1
3
1
2 3 −1 1
2
−1
4 −1
6 −1 2 3
2
−1
12
1
12
1
2
3
2
7
4
u1
u2
u3
u4
u5
.
Computation of the multiplicity:
γ(1,1)((−1, 0)) = rank
(
L(1,1)
(−1,0)(u, v)
)
= 2.
Acknowledgements
I would like to express my sincere gratitude to R. Pnini for his kind help and support.
References
[1] Arnal D., Ben Ammar M., Selmi M., Le problème de la réduction à un sous-groupe dans la quantification
par déformation, Ann. Fac. Sci. Toulouse Math. (5) 12 (1991), 7–27.
[2] Baldoni M.W., Beck M., Cochet C., Vergne M., Volume computation for polytopes and partition functions
of classical root systems, Discrete Comput. Geom. 35 (2006), 551–595, math.CO/0504231.
[3] Bando M., Kuratomo T., Maskawa T., Uehera S., Structure of nonlinear realization in supersymmetric
theories, Phys. Lett. B 138 (1984), 94–98.
[4] Bando M., Kuratomo T., Maskawa T., Uehera S., Nonlinear realizations in supersymmetric theories, Progr.
Theoret. Phys. 72 (1984), 313–349.
Bando M., Kuratomo T., Maskawa T., Uehera S., Nonlinear realizations in supersymmetric theories. II,
Progr. Theoret. Phys. 72 (1984), 1207–1213.
[5] Bar-Moshe D., Marinov M.S., Realization of compact Lie algebras in Kähler manifolds, J. Phys. A: Math.
Gen. 27 (1994), 6287–6298, hep-th/9407092.
[6] Bar-Moshe D., Marinov M.S., Berezin quantization and unitary representations of Lie groups, in Topics in
Statistical and Theoretical Physics, Amer. Math. Soc. Transl. Ser. 2, Vol. 177, Amer. Math. Soc., Providence,
RI, 1996, 1–21, hep-th/9407093.
[7] Berezin F.A., Quantization, Izv. Akad. Nauk SSSR Ser. Mat. 38 (1974), 1116–1175.
[8] Berezin F.A., Covariant and contravariant symbols of operators, Izv. Akad. Nauk SSSR Ser. Mat. 36 (1972),
1134–1167.
[9] Billey S., Guillemin V., Rassart E., A vector partition function for the multiplicities of slk(C), J. Algebra
278 (2004), 251–293, math.CO/0307227.
[10] Baoua O.B., Gelfand pairs with coherent states, Lett. Math. Phys. 46 (1998), 247–263.
[11] Cahen B., Multiplicities of compact Lie group representations via Berezin quantization, Mat. Vesnik 60
(2008), 295–309.
http://arxiv.org/abs/math.CO/0504231
http://arxiv.org/abs/hep-th/9407092
http://arxiv.org/abs/hep-th/9407093
http://arxiv.org/abs/math.CO/0307227
12 D. Bar-Moshe
[12] Guillemin V., Sternberg S., Geometric quantization and multiplicities of group representations, Invent.
Math. 67 (1982), 515–538.
[13] Heckman G.J., Projections of orbits and asymptotic behavior of multiplicities for compact connected Lie
groups, Invent. Math. 67 (1983), 333–356.
[14] Hurt N.E., Geometric quantization in action, Mathematics and Its Applications (East European Series),
Vol. 8, D. Reidel Publishing Co., Dordrecht-Boston, Mass., 1983.
[15] Kostant B., A formula for the multiplicity of a weight, Trans. Amer. Math. Soc. 93 (1959), 53–73.
[16] Littelmann P., Paths and root operators in representation theory, Ann of Math. (2) 142 (1995), 499–525.
[17] O’Raifeartaigh L., Group structures of gauge theories, Cambridge Monographs on Mathematical Physics,
Cambridge University Press, Cambridge, 1986.
[18] Peetre J., The Berezin transform and Ha-plitz operators, J. Operator Theory 24 (1990), 165–168.
[19] Rawnsley J., Cahen M., Gutt S., Quantization of Kähler manifolds. I. Geometric interpretation of Berezin’s
quantization, J. Geom. Phys. 7 (1990), 45–62.
[20] Vergne M., Multiplicities formula for geometric quantization. I, Duke Math. J. 82 (1996), 143–179.
[21] Vergne M., Multiplicities formula for geometric quantization. II, Duke Math. J. 82 (1996), 181–194.
1 Introduction
2 Berezin quantization and the Borel-Weil theory
2.1 Notations
2.2 The quantization space in Berezin theory
2.3 The T action on G/T and L2( hol( L) )
2.4 Parametrization of G/T and the construction of the basic Kähler potentials
3 A weight multiplicity formula based on Berezin quantization
4 A method for weight multiplicity computation
4.1 Computation by direct integration
4.2 Weight multiplicity computation as a rank of a Hermitian form
A Examples
A.1 Computation of some weight multiplicities in the representation (4,2) of SU(3)
A.1.1 Computation of the multiplicity of the weight ( 0,1) in (4,2)
A.1.2 Computation of the multiplicity of the weight (-6,4) in (4,2)
A.2 Computation of some weight multiplicities in the representation (1,1) of SO(5)
References
|