A CONSTRUCTION BY DEFORMATION OF UNITARY IRREDUCIBLE REPRESENTATIONS OF SU(n + 1) BENJAMIN CAHEN

To the memory of my father, Alfred Cahen (1932-2016) Abstract. We derive unitary irreducible representations of SU (n + 1) from a minimal realization of sl(n + 1, C) by using various techniques from deformation theory.

1. Introduction The deformations of Lie algebras were intensively studied in the years 1960-70 [10], [21], [22], [18] and still remain objects of active research, see for instance [7], [8] and [6]. On the other hand, the deformations of Lie algebra representations have not been studied as systematically, with some notable exeptions, see [23], [13], [19] and also [20]. Constructing (formal) deformations of Lie algebra representations is a way to derive a family of representations from a given one and then to get many representations from a few ones. However, the existence and classification problems for deformations depend on some Lie algebra cohomology modules which are not easy to compute in general, see for instance [19] and [3]. The aim of the present note is to use deformation theory in order to recover some unitary irreducible representations of SU(n + 1). More precisely, we construct the unitary irreducible representations of SU(n + 1) considered in [4] by deformation of a so-called minimal realization of sl(n + 1, C) [15]. It is known that such a minimal realization is connected to the minimal (non trivial) nilpotent coadjoint orbit of SL(n + 1, C) [16], [1], [17]. By taking a parametrization of this orbit, we can simplify the computation of the deformations of the minimal realization by using the Moyal star product and the Weyl correspondence as in [1] and [3]. Although the derivation presented here can be considered as a simple exercice in deformation theory, we have not found it explicitly done in details in the literature. Moreover, we could hope for applications of this method to the description of representations of Lie algebras (in particular of unitary dual of Lie groups) in more general situations (some examples can already be found in [20] and [3]). This note is organized as follows. In Section 2, we describe the unitary irreducible representations of SU(n+ 1) introduced in [4] as an analogue to the holomorphic discrete series representations of SU(1, n) and we compute their differentials which can be extended to representations of sl(n + 1, C). In the notation of [11], p. 143, these representations 2000 Mathematics Subject Classification. 17B10; 17B20; 17B56; 22E46; 53D55. Key words and phrases. Deformation of representation; Lie algebra; unitary group; ChevalleyEilenberg cohomology; Moyal star product; Weyl correspondence; minimal realization; minimal coadjoint orbit. 1

2

BENJAMIN CAHEN

have highest weight mǫ1 with m integer ≥ 1. Section 3 is devoted to some generalities on (formal) deformations of Lie algebra homomorphisms and, in Section 4, we recall the Moyal star product and the Weyl correspondence [9], [24]. In Section 5, we show how a symplectic chart of the minimal nilpotent coadjoint orbit of sl(n + 1, C) naturally leads to a minimal realization of sl(n + 1, C) and we expose the method which will be used to recover the representations of SU(n + 1). In Section 6, we compute the first cohomology module corresponding to the deformation of the minimal realization and then we derive the desired representations of SU(n + 1) in Section 7. In particular, by this way we can recover all the irreducible unitary representations of SU(2). 2. Representations of SU(n + 1) Here we consider a family of representations of SU(n + 1) indexed by an integer m ≥ 1. In [4], we showed that this family can be contracted to the unitary irreducible representations of the Heisenberg group of dimension 2n + 1 as the holomorphic discrete series representations of SU(1, n). The group SU(n+1) consists of all complex (n+1)×(n+1) matrices g with determinant 1 such that g ∗ g = In+1 . Here we write the elements of the group SU(n + 1) as block matrices a b g= c d with matrices a(1 × 1), b(1 × n), c(n × 1) and d(n × n). The Lie algebra su(n + 1) of SU(n + 1) consists of all matrices of the form iα b −b∗ A where α ∈ R, b ∈ Cn and A is an anti-Hermitian n × n matrix (that is, A∗ = −A) such that iα + Tr(A) = 0. The group SU(n + 1) acts naturally on the projective space Pn (C) and this action induces an holomorphic action (defined almost everywhere) of SU(n+1) on Cn by fractional linear transformations a b t −1 t t g · z = (a + bz ) (c + dz ) , g= c d where the subscript t denotes transposition. For each integer m ≥ 1, let Pm be the space of all complex polynomial functions on Cn of degree ≤ m. We endow Pm with the Hilbert product Z hf1 , f2 im := f1 (z)f2 (z) dµm (z) Cn

n

where the measure µm on C is defined by (m + 1) . . . (m + n) (1 + kzk2 )−m−n−1 dx1 dy1 . . . dxn dyn . dµm (z) := πn Here we use the notation z = (x1 + iy1 , x2 + iy2 , . . . , xn + iyn ) where xk , yk ∈ R for k = 1, 2, . . . , n. Now, let πm be the representation of SU(n + 1) on Pm defined by a b t m −1 −1 (πm (g) f )(z) = (bz + a) f (g · z), g = . c d

A CONSTRUCTION BY DEFORMATION...

3

We can easily verify that πm is unitary. Moreover, the differential dπm of πm can be extended to a representation of sl(n + 1, C) = su(n + 1)c also denoted by dπm . We have (dπm (X)f )(z) = −m(βz t + α)f (z) + df (z)((α + βz t )z − (γ + δz t )t ) where X=

α β γ δ

with matrices α(1 × 1), β(1 × n), γ(n × 1) and δ(n × n). In order to give more explicit formulas for dπm , let us introduce the following basis of sl(n + 1, C). For 1 ≤ i, j ≤ n + 1, write Eij for the matrix whose ij-th entry is 1 and all of the other entries are 0. Then the matrices Hk = Ek+1k+1 − E11 (1 ≤ k ≤ n) form a basis for the Cartan subalgebra h of sl(n + 1, C) consisting of all diagonal matrices of sl(n + 1, C) and, obviously, the matrices Hk (1 ≤ i ≤ n) and Eij (1 ≤ i 6= j ≤ n + 1) form a basis for sl(n + 1, C). Then we have n

X ∂f ∂f − zj (dπm (Hk )f )(z) =mf (z) − zk ∂zk ∂zj j=1 (dπm (E1k+1 )f )(z) = − mzk f (z) + zk

n X

zj

j=1

∂f ∂zj

∂f ∂zk ∂f (dπm (Ei+1j+1 )f )(z) = − zj ∂zi (dπm (Ek+11 )f )(z) = −

for 1 ≤ k ≤ n and 1 ≤ i 6= j ≤ n. We can easily see that dπm -hence πm - is irreducible. Indeed, let V be a nonzero subspace of Pm which is invariant under dπm (X) for each X ∈ sl(n + 1, C). Then there exists at least one nonzero element f in V. Thus, by applying the operators dπm (Ek+11 ) to f , we get 1 ∈ V and by applying the operators dπm (E1k+1 ) and dπm (Ei+1j+1 ) to 1 we see that V = Pm . Let us denote by ǫk , 1 ≤ k ≤ n, the linear form on h defined by ǫk : Diag(a1 , a2 , . . . , an+1 ) → ak . It is well-known that the root system of sl(n + 1, C) relative to h is ∆ = {ǫi − ǫj : 1 ≤ i, j ≤ n + 1}, see for instance [11]. The ordering on ∆ is usually taken so that the positive roots are ǫi − ǫj (1 ≤ i < j ≤ n + 1). In this context, we can verify that dπm has highest weight mǫ1 and highest weight vector f = znm . 3. Generalities on deformations In this section, we recall some definitions and results of deformation theory. The material of this section is essentially taken from [23], [13], [19], see also [12] and [3].

4

BENJAMIN CAHEN

Let g be a Lie algebra over C and let A be an associative algebra over C with unit element 1. Then A is also a Lie algebra for the commutator [a, b] := ab − ba. Let ϕ : g → A be a Lie algebra homomorphism. P k Definition 3.1. (1) A formal deformation of ϕ is a formal series Φ = k≥0 t Φk where Φ0 = ϕ and Φk is a linear map from g to A for each k ≥ 1, such that (3.1)

Φ([X, Y ]) = [Φ(X), Φ(Y )]

for each X and Y in g. Here we have extended the bracket of A to formal series by bilinearity. (2) Two formal deformations Φ and Ψ of ϕ are said to be equivalent if there exists a series a = 1 + ta1 + t2 a2 + . . . ∈ A[[t]] such that for each X ∈ g, we have a−1 Φ(X)a = Ψ(X).

(3.2)

The study of the formal deformations of ϕ naturally leads us to consider the structure of g-module on A defined by X · a = [ϕ(X), a] for X ∈ g and a ∈ A and the ChevalleyEilenberg cohomology of g with values in the g-module A. Indeed, denoting by ∂ the corresponding cobord operator, we immediately see that Eq. 3.1 is equivalent to the fact that for each n ≥ 0 and each X, Y ∈ g, we have (∂Φn )[X, Y ] :=[ϕ(X), Φn (Y )] + [Φn (X), ϕ(Y )] − Φn ([X, Y ]) =−

n−1 X

[Φk (X), Φn−k (Y )].

k=1

In particular, we see that if such a deformation Φ exists then Φ1 is a 1-cocycle. We have the following result, see for instance [13], Section III and [19], Section I. Proposition 3.2. (1) If we have H 2 (g, A) = (0) then, for each 1-cocycle α : g → A, there exists a formal deformation Φ such that Φ1 = α. (2) If we have H 1 (g, A) = (0) then each formal deformation Φ of ϕ is equivalent to ϕ. In [3], we proved the following result. Proposition 3.3. Assume that H 1 (g, A) is one-dimensional and that there exists a formal deformation Φ of ϕ such the class of Φ1 generates H 1 (g, A). PFor each sequence c = (ck )k≥1 of complex numbers, consider the P formal series Sc (t) := k≥1 ck tk and the formal deformation Φc of ϕ defined by Φc (X) = r≥0 Sc (t)r Φr (X) for each X ∈ g. Then the map c → Φc is a bijection from the set of all sequences c = (ck )k≥1 of C onto the set of all equivalence classes of formal deformations of ϕ. Note that the preceding definitions and results can be applied to the particular case of a representation ϕ of g in a complex vector space V , since ϕ is also a Lie algebra homomorphism from g to End(V ), or, more generally, to a subalgebra A of End(V ). 4. Weyl correspondence and Moyal star product Here we first recall the Moyal star product, see for instance [2]. Take coordinates (p, q) on R2n ∼ = Rn × Rn and let x = (p, q). Then one has xi = pi for 1 ≤ i ≤ n and xi = qi−n

A CONSTRUCTION BY DEFORMATION...

5

for n + 1 ≤ i ≤ 2n. For u, v ∈ C ∞ (R2n ), define P 0 (u, v) := uv, n X X ∂u ∂v ∂u ∂v 1 = Λij ∂xi u∂xj v − P (u, v) := ∂pk ∂qk ∂qk ∂pk 1≤i,j≤n k=1

(the Poisson brackets) and, more generally, for l ≥ 2, X P l (u, v) := Λi1 j1 Λi2 j2 · · · Λil jl ∂xl i1 ...xi u ∂xl j1 ...xj v. l

l

1≤i1 ,...,il ,j1 ,...,jl ≤n

Then the Moyal product ∗M is the following formal deformation of the pointwise multiplication of C ∞ (R2n ) X tl u ∗M v := P l (u, v) l! l≥0

where t is a formal parameter. Moreover, the corresponding Moyal brackets are given by [u, v]∗M :=

X t2l 1 (u ∗M v − v ∗M u) = P 2l+1 (u, v). 2t (2l + 1)! l≥0

Now, we restrict ∗M to polynomials on R2n and take t = −i/2. Then we get an associative product ∗ on polynomials which we denote by ∗. This product corresponds to the composition of operators in the usual Weyl quantization procedure as we will explain below. The Weyl correspondence on R2n is defined as follows, see [5], [9], [14]. For each f in the Schwartz space S(R2n ), we define the operator W (f ) acting on the Hilbert space L2 (Rn ) by Z −n W (f )ϕ(p) = (2π) eisq f (p + (1/2)s, q) ϕ(p + s) ds dq. R2n

As it is well-known, that the Weyl calculus can be extended to much larger classes of symbols (see for instance [14]). In particular, if f (p, q) = u(p)q α where u ∈ C ∞ (Rn ) then we have α ∂ (4.1) W (f )ϕ(p) = i (u(p + (1/2)s) ϕ(p + s)) , ∂s s=0 see [24]. For instance, if f (p, q) = u(p) then W (f )ϕ(p) = u(p) ϕ(p) and if f (p, q) = u(p)qk then (4.2) W (f )ϕ(p) = i (1/2)∂k u(p) ϕ(p) + u(p)∂k ϕ(p) .

Moreover, we have W (f1 ∗ f2 ) = W (f1 )W (f2 ) for each functions f1 , f2 on R2n of the form u(p)q α , in particular for polynomials, see [9], p. 103. Note also that, since the map W and the product ∗ on polynomials can be defined in a purely algebraic way, see Eq. 4.1, we can extended them to the polynomials in complex variables p, q without any modification.

6

BENJAMIN CAHEN

5. Minimal realization In [1], a general method for constructing minimal realizations of semisimple complex Lie algebras from minimal coadjoint orbits was introduced. In the particular case of the Lie algebra g := sl(n + 1, C) of G := SL(n + 1, C), this method goes as follows. First, we can identify the dual g∗ of g with g by means of the bilinear form on g defined by hX, Y i := Tr(XY ). In this identification, the coadjoint action of G corresponds to the adjoint action of G and the coadjoint orbits to the adjoint orbits. This is a simple exercice to show that the minimal (non trivial) nilpotent (co)adjoint orbit O of G consists of all rank one matrices of g. Now, let us consider the map Ψ from C2n to O′ := O ∪ (0) defined by Pn q1 . . . qn − Pj=1 pj qj −p1 n pj qj p1 q1 . . . p1 qn j=1 Ψ(p, q) := . . . . . . . . . . . . . . . . . . . Pn −pn j=1 pj qj pn q1 . . . pn qn Then the image of Ψ is a dense open subset of O′ . ˜ the corresponding coordinate function on C2n : For each X ∈ g, let us denote by X ˜ q) := hΨ(p, q), Xi. X(p,

Proposition 5.1.

(1) For each X, Y ∈ g, we have ˜ Y˜ ]∗ = {X, ˜ Y˜ } = [X,˜Y ]. [X,

˜ is a representation of g in C[q] := C[q1 , q2 , . . . , qn ]. (2) The map ρ0 : X → W (iX) ˜ Y˜ } = [X,˜Y ] can be verified by a direct Proof. (1) Let X and Y in g. The equation {X, ˜ and Y˜ are polynomials of degree ≤ 1 in the computation. On the other hand, since X k ˜ ˜ ˜ Y˜ ]∗ = variables q1 , q2 , . . . , qn , we have P (X, Y ) = 0 for each k ≥ 3, hence we get [X, ˜ ˜ {X, Y }. (2) Let X and Y in g. By (1), we have ˜ ∗ (iY˜ ) − (iY˜ ) ∗ (iX) ˜ = i[X,˜Y ]. (iX) ˜ W (iY˜ )] = W (i[X,˜Y ]) hence Then, by the remark at the end of Section 4, we get [W (iX), the result. The representation ρ0 is a minimal realization of g, that is, a realization of g as Lie algebra of differential operators acting on functions of n variables with n minimal, see [15]. Note that Span{En+12 , . . . , En+1n , E21 , . . . , En+11 } is a Heisenberg Lie algebra of dimension 2n − 1 with central element En+11 , the only non trivial brackets being [En+1k , Ek1 ] = En+11 for k = 2, . . . , n. Then Ψ was chosen so that ˜k1 = qk−1 (for k = 2, . . . , n) and E˜n+11 = qn . In fact, these conditions E˜n+1k = pk−1 qn , E determine Ψ uniquely. Now, we aim to study the deformations of ρ0 . By using the map f → W (if ) this is equivalent to studying the deformations of the Lie algebra homomorphism X → Φ0 (X) := ˜ from g to M := C[p, q] endowed with [·, ·]∗. X

A CONSTRUCTION BY DEFORMATION...

7

As explained in Section 3, we endow M with the g-module structure defined by X · f := ˜ [X, f ]∗ and then consider the corresponding Chevalley-Eilenberg cohomology. 6. Determination of H 1 (g, M) Recall that H 1 (g, M) is the quotient space Z 1 (g, M)/B 1 (g, M) where Z 1 (g, M) consists of all linear maps ϕ : g → M satisfying ˜ ϕ(Y )]∗ + [ϕ(X), Y˜ ]∗ − ϕ[X, Y ] = 0 (6.1) ∂ϕ(X, Y ) := [X, ˜ f ]∗ (the 1-cocycles) and B 1 (g, M) consists of all maps from g to M of the form X → [X, for f ∈ M (the 1-coboundaries). The aim of this section is to compute H 1 (g, M). We begin with the following ’Poincar´e lemma’. Lemma 6.1. Let Fi (q), i = 1, 2, . . . , n be a family of polynomials in the variable q = ∂F i (q1 , q2 , . . . , qn ) such that, for each i, j = 1, 2, . . . , n, one has ∂F = ∂qij . Then there exists ∂qj ∂F = Fi for each i = 1, 2, . . . , n. a polynomial F (q) such that ∂q i Proof. By the usual Poincar´e lemma, the result is true for polynomials in real variables qi which implies that it is also true for polynomials in complex variables qi . Proposition 6.2. The space H 1 (g, M) is one dimensional, generated by the class of the cocycle ϕ1 defined by ϕ1 (E11 − E22 ) = 1, ϕ1 (Ekk − Ek+1k+1 ) = 0 for k = 2, . . . , n, ϕ1 (E1k+1 ) = pk for k = 1, 2, . . . , n and ϕ1 (Eij ) = 0 for i ≥ 2. Proof. We have divided the proof into several steps. The method of the proof is quite elementary and consists in transforming progressively a given 1-cocycle to an equivalent one which is more simple by adding suitable 1-coboundaries. Let us consider a 1-cocycle ϕ : g → M. 1) First we apply Eq. 6.1 to X = Ek+11 and Y = El+11 for k, l = 1, 2, . . . , n. Writing ∂ϕl k ϕk = ϕ(Ek+11 ) for simplicity, we get ∂ϕ = ∂p for each k, l = 1, 2, . . . , n. Then, by k P α ∂pαl decomposing each ϕk as ϕk = α ϕk (p)q with the usual multi-index notation, we have ∂ϕα ∂ϕα k = ∂pkl for each k, l, α. ∂pl α Thus, by Lemma 6.1, for each α there exists a polynomial ϕα (p) such that ∂ϕ = ϕαk ∂pk for each k = 1, 2,P . . . , n. Now, let φ := α ϕα (p)q α . For each k = 1, 2, . . . , n, we have [φ, qk ]∗ =

∂φ = ϕk . ∂pk

Hence, replacing ϕ by the equivalent 1-cocycle ϕ − [φ, ·]∗ , we can always assume that ϕ(Ek+11 ) = 0 for each k = 1, 2, . . . , n. 2) We apply Eq. 6.1 to X = Ekl , k, l ≥ 2, k 6= l and Y = Ej+11 , j = 1, 2, . . . , n. Taking 1) into account, we can immediately see that ϕ(Ekl ) is a polynomial in the variables q1 , q2 , . . . , qn . 3) Similarly, applying Eq. 6.1 to X ∈ h and Y = Ej+11 , we verify that ϕ(X) is a polynomial in the variables q1 , q2 , . . . , qn . Now, we fix k = 1, 2, . . . , n − 1 and we apply Eq. 6.1 to X = En+1k+1 and Y = P4) n−1 j=1 (En+1n+1 − Ej+1j+1 ). Write ϕk := ϕ(En+1k+1 ) for simplicity and recall that ϕk is

8

BENJAMIN CAHEN

a polynomial in q1 , q2 , . . . , qn by 2). Then we see that there exists a polynomial uk (q) in q1 , q2 , . . . , qn such that (6.2)

− nϕk =

n X

qj

j=1

∂ϕk + qn uk (q). ∂qj

m Let ϕk = m ϕm k and uk = m uk be the decompositions of ϕk and uk into homogeneous polynomials of degree m in q1 , q2 , . . . , qn . Then Eq. 6.2 implies that X X −n ϕm mϕm k = k + qn uk−1 (q)

P

P

m

m

and we conclude that, for each k = 1, 2, . . . , n − 1, there exists a polynomial ψk in q1 , q2 , . . . , qn such that ϕk = qn ψk . Taking X = En+1k+1 and Y = En+1l+1 in Eq. 6.1 for k, l = 1, 2, . . . , n − 1, we get ∂ψk l = ∂ψ for each k, l. This implies the existence of a polynomial ψ in q1 , q2 , . . . , qn such ∂ql ∂qk ∂ψ that ψk = ∂q for each k = 1, 2, . . . , n − 1. k Thus, by replacing ϕ by ϕ − [·, ψ]∗ , we are led to the case where ϕ(En+1k+1 ) = 0 for each k = 1, 2, . . . , n − 1 and the condition ϕ(Ek+11 ) = 0 for each k = 1, 2, . . . , n is still satisfied. 5) Let k = 1, 2, . . . , n, l = 1, 2, . . . , n−1 with k 6= l. By applying Eq. 6.1 to X = Ek+1l+1 and Y = En+1j+1 for j = 1, 2, . . . , n−1, we see that ϕ(Ek+1l+1) only depends on qn . Thus, taking into account the equality [Ek+1l+1 , El+1k+1 ] = Ek+1k+1 − El+1l+1 we get ϕ(Ek+1k+1 − El+1l+1 ) = 0. Hence, applying Eq. 6.1 to X = Ek+1k+1 − El+1l+1 and Y = Ek+1l+1 we obtain ϕ(Ek+1l+1 ) = 0. Finally, we apply Eq. 6.1 to X = Ej+1n+1 and Y = Ek+1j+1 and we also obtain ϕ(Ek+1n+1 ) = 0. 6) Now, take X ∈ h and Y = Ek+1j+1 in Eq. 6.1. Then we see that ϕ(X) only depends on qn . Let H0 = E11 − E22 ∈ h. Then we can replace ϕ by ϕ + [·, F (qn )]∗ for a suitable polynomial F (qn ) so that ϕ(H0 ) is a constant which we denote by a. 7) Taking X = E12 and successively Y = Ek+11 , (k = 1, 2, . . . , n) and Y = En+1k , (k = 2, . . . , n − 1) in Eq. 6.1, we see that ϕ(E12 ) = ap1 + f (qn ) where f (qn ) is a polynomial. Moreover, taking also X = E12 and Y = H0 , we get ∂f −2f (qn ) = qn ∂q hence f = 0 and ϕ(E12 ) = ap1 . n 8) Finally, we apply Eq. 6.1 to X = E12 and Y = E2k+1 where k = 2, . . . , n we obtain ϕ(E1k+1 ) = apk . 7. Derivation of the representations πm In this section, we retain the notation of the previous sections. Proposition 6.2 leads ˜ such that Φ1 = aϕ1 for a ∈ C. us to consider the formal deformations Φ of Φ0 : X → X We have the following result.

A CONSTRUCTION BY DEFORMATION...

9

˜+ Proposition 7.1. For each a ∈ C, the map Φa : g → M[[t]] defined by Φa (X) = X taϕ1 (X) is a formal deformation of Φ0 in M. Proof. Taking into account that ϕ1 is a 1-cocycle (see Section 6), the result follows immediately from the equality [ϕ1 (X), ϕ1 (Y )]∗ = 0 for X, Y ∈ g. By using the properties of W (see Section 4), we get the following proposition. Proposition 7.2. For each a ∈ C, let m(a) := −1/2(a + n + 1). Then the map ρa defined by i ˜ ρa (X) = W X − aϕ1 (X) 2 for X ∈ g is a representation of g in C[p] and we have n

X ∂f ∂f (ρa (Hk )f )(z) =m(a)f (z) − zk − zj ∂zk j=1 ∂zj (ρm (E1k+1 )f )(z) = − m(a)zk f (z) + zk

n X j=1

zj

∂f ∂zj

∂f ∂zk ∂f (ρa (Ei+1j+1 )f )(z) = − zj ∂zi (ρa (Ek+11 )f )(z) = −

for 1 ≤ k ≤ n and 1 ≤ i 6= j ≤ n. Proof. The fact that ρa is a representation g follows from Proposition 7.2 and the formulas for ρa can be easily verified by Eq. 4.2. In other words, the formulas for ρa are the same as the formulas for dπm , see Section 2, but note that these two representations don’t act on the same spaces since ρa acts on C[p] and dπm on the finite dimensional space Pm . In order to recover the representations πm of Section 2, we select now the values of a (or, equivalently, of m(a)) for which there exists a non trivial finite dimensional subspace of C[p] that is invariant under ρa . Proposition 7.3. Let a ∈ C. Assume that P is a non trivial finite dimensional subspace of C[p] that is invariant under ρa . Then m(a) is a non negative integer, we have P = Pm(a) and the restriction of ρa to P coincides with dπm(a) . Proof. Let a ∈ C. Let P 6= (0) a finite dimensional subspace of C[p] which is invariant under ρa . Define m := max{deg(f ) : f ∈ P \ (0)}. Let f be an element of P of degree m. P f Let us decompose f as f = m k=0 k where, for each k, fk is an homogeneous polynomial of degree k. Then we have fm 6= 0 and ρa (E1l+1 )f = pl

m X k=0

(k − m(a))fk .

10

BENJAMIN CAHEN

We see that if m 6= m(a), we get a contradiction. Thus we have m(a) = m hence m(a) is a non negative integer and P ⊂ Pm(a) . Since Pm(a) is irreducible under the action of dπm(a) , see Section 2, we can conclude that P = Pm(a) . Then we have recovered the representations dπm of Section 2, hence the representations πm by integration. Note that by taking n = 1 we see that this method gives all the unitary irreducible representations of SU(2). References [1] D. Arnal, H. Benamor and B. Cahen, Minimal realizations of classical simple Lie algebras through deformations, Ann. Fac. Sci. Toulouse VII, 2 (1998), 169-184. [2] D. Arnal and J.-C. Cortet, Repr´esentations ∗ des groupes de Lie exponentiels, J. Funct. Anal. 92, 1 (1990), 103-135. [3] B. Cahen, D´eformations formelles de certaines repr´esentations de l’alg`ebre de Lie d’un groupe de Poincar´e g´en´eralis´e, Ann. Math. Blaise Pascal 8, 1 (2001), 17-37. [4] B. Cahen, Contractions of SU (1, n) and SU (n + 1) via Berezin quantization, J. Anal. Math. 97 (2005) 83-102. [5] M. Combescure and D. Robert, Coherent States and Applications in Mathematical Physics, Springer, 2012. [6] D. Burde, Contractions of Lie algebras and algebraic groups, Arch. Math., Brno 43, 5 (2007), 321-332. [7] A. Fialowski, Deformations in Mathematics and Physics, Intern. Journ. Theor. Physics, 47, 2 (2008), 333-337. [8] A. Fialowski and M. Penkava, Deformations of nilpotent associative algebras of dimension 4, Linear Algebra Appl. 457 (2014), 408-427. [9] B. Folland, Harmonic Analysis in Phase Space, Princeton Univ. Press, 1989. [10] M. Gerstenhaber, On the deformation of rings and algebras, Ann. Math. 79, 1 (1964), 59-103. [11] R. Goodman and N. R. Wallach, Symmetry, Representations and Invariants, Graduate Texts in Mathematics 255, Springer Dordrecht Heidelberg London New-York, 1985. [12] A. Guichardet, Cohomologie des groupes topologiques et des alg`ebre de Lie, Cedic, Paris, 1980. [13] R. Hermann, Analytic Continuation of Group Representations IV, Comm. Math. Phys. 5 (1967), 131-156. [14] L. H¨ ormander, The analysis of linear partial differential operators, Vol. 3, Section 18.5, SpringerVerlag, Berlin, Heidelberg, New-York, 1985. [15] A. Joseph, Minimal Realizations and Spectrum Generating Algebras, Comm. Math. Phys. 36 (1974), 325-338. [16] A. Joseph, The minimal orbit in a simple Lie algebra and its associated maximal ideal, Ann. Sci. Ecole Norm. Sup. 9 (1976), 1-30. [17] D. Kazhdan, B. Pioline, A. Waldron, Minimal representations, spherical vectors and exceptional theta series, Comm. Math. Phys. 226 (2002), 140. [18] M. Levy-Nahas, Deformation and Contraction of lie algebras, J. Math. Phys. 8, 6 (1967), 1211-1222. [19] M. Levy-Nahas, First Order deformations of Lie algebras Representations, E(3) and Poicar´e Examples, Comm. Math. Phys. 9 (1968), 242-266. [20] M. Lesimple and G. Pinczon, Deformations of Lie group and Lie algebra representations, J. Math. Phys. 34, 9 (1993), 4251-4272. [21] A. Nijenhuis and R. W. Richardson, Cohomology and deformations in graded Lie algebras, Bull. Amer. Math. Soc. 72 (1966), 1-29. [22] A. Nijenhuis and R. W. Richardson, Deformations of Lie Algebras Structures, J. Math. Mech. 17 (1967), 89-105. [23] A. Nijenhuis and R. W. Richardson, Deformations of homomorphisms of Lie groups and Lie Algebras, Bull. Amer. Math. Soc. 73 (1967), 175-179. [24] A. Voros, An Algebra of Pseudo differential operators and the Asymptotics of Quantum Mechanics, J. Funct. Anal. 29 (1978), 104–132.

A CONSTRUCTION BY DEFORMATION...

11

´partement de math´ Universit´ e de Lorraine, Site de Metz, UFR-MIM, De ematiques, ˆtiment A, Ile du Saulcy, CS 50128, F-57045, Metz cedex 01, France. Ba E-mail address: [email protected]