arXiv subjects
Publications and source records attributed to Bertram Kostant.
I connect an old result of mine on a Lie algebra generalization of the Amitsur-Levitski theorem with equations for sheets in a reductive Lie algebra and with recent results of Kostant-Wallach on the variety of singular elements in a reductive Lie algebra.
Let $G$ be a complex simply-connected semisimple Lie group and let $\frak{g}= Lie G$. Let $\frak{g} = \frak{n}_- +\frak{h} + \frak{n}$ be a triangular decomposition of $\frak{g}$. One readily has that $Cent\,U({\frak n})$ is isomorphic to the ring $S({\frak n})^{\frak\n}$ of symmetric invariants. Using the cascade ${\cal B}$ of strongly orthogonal roots, some time ago we proved that $S({\frak n})^{{\frak n}$ is a polynomial ring $\Bbb C [ξ_1,...,ξ_m]$ where $m$ is the cardinality of ${\cal B}$. Using this result we establish that the maximal coadjoint of $N = exp \frak {n}$ has codimension $m$. Let $\frak {b}= \frak {h} + \frak {n}$ so that the corresponding subgroup $B$ is a Borel subgroup of $G$. Let $\ell = rank \frak{g}$. Then in this paper we prove the theorem that the maximal coadjoint orbit of $B$ has codimension $\ell - m$ so that the following statements (1) and (2) are equivalent: (1) -1 is in the Weyl group of $G$ (i.e., $\ell = m$), and (2), B has a nonempty open coadjoint orbit. We remark that a nilpotent or a semisimple group cannot have a nonempty open coadjoint orbit. Celebrated examples where a solvable Lie group has a nonempty coadjoint orbit are due to Piatetski--Shapiro in his counterexample construction of a bounded complex homogeneous domain which is not of Cartan type.
Let $G$ be a complex simply-connected semisimple Lie group and let $\frak{g}= Lie G$. Let $\frak{g} = \frak{n}_- +\frak{h} + \frak{n}$ be a triangular decomposition of $\frak{g}$. One readily has that $Cent\,U({\frak n})$ is isomorphic to the ring $S({\frak n})^{\frak\n}$ of symmetric invariants. Using the cascade ${\cal B}$ of strongly orthogonal roots, some time ago we proved that $S({\frak n})^{{\frak n}$ is a polynomial ring $\Bbb C [ξ_1,...,ξ_m]$ where $m$ is the cardinality of ${\cal B}$. Using this result we establish that the maximal coadjoint of $N = exp \frak {n}$ has codimension $m$. Let $\frak {b}= \frak {h} + \frak {n}$ so that the corresponding subgroup $B$ is a Borel subgroup of $G$. Let $\ell = rank \frak{g}$. Then in this paper we prove the theorem that the maximal coadjoint orbit of $B$ has codimension $\ell - m$ so that the following statements (1) and (2) are equivalent: (1) -1 is in the Weyl group of $G$ (i.e., $\ell = m$), and (2), B has a nonempty open coadjoint orbit. We remark that a nilpotent or a semisimple group cannot have a nonempty open coadjoint orbit. Celebrated examples where a solvable Lie group has a nonempty open coadjoint orbit are due to Piatetski--Shapiro in his counterexample construction of a bounded complex homogeneous domain which is not of Cartan type.
Let $G$ be a complex simply-connected semisimple Lie group and let $\g=\hbox{\rm Lie}\,G$. Let $\g = \n_- +\hh + \n$ be a triangular decomposition of $\g$. One readily has that $\hbox{\rm Cent}\,U(\n)$ is isomorphic to the ring $S(\n)^{\n}$ of symmetric invariants. Using the cascade ${\cal B}$ of strongly orthogonal roots, some time ago we proved (see [K]) that $S(\n)^{\n}$ is a polynomial ring $\Bbb C[ξ_1,...,ξ_m]$ where $m$ is the cardinality of ${\cal B}$. The authors in [LW] introduce a very nice representation-theoretic method for the construction of certain elements in $S(\n)^{\n}$. A key lemma in [LW] is incorrect but the idea is in fact valid. In our paper here we modify the construction so as to yield these elements in $S(\n)^{\n}$ and use the [LW] result to prove a theorem of Tony Joseph.
The representation of the conformal group (PSU(2,2)) on the space of solutions to Maxwell's equations on the conformal compactification of Minkowski space is shown to break up into four irreducible unitarizable smooth Fréchet representations of moderate growth. An explicit inner product is defined on each representation. The frequency spectrum of each of these representations is analyzed. These representations have notable properties; in particular they have positive or negative energy, they are of type $A_{\frak q}(λ)$ and are quaternionic. Physical implications of the results are explained.
Let $G$ be a semisimple Lie group and let $\g =\n_- +\hh +\n$ be a triangular decomposition of $\g= \hbox{Lie}\,G$. Let $\b =\hh +\n$ and let $H,N,B$ be Lie subgroups of $G$ corresponding respectively to $\hh,\n$ and $\b$. We may identify $\n_-$ with the dual space to $\n$. The coadjoint action of $N$ on $\n_-$ extends to an action of $B$ on $\n_-$. There exists a unique nonempty Zariski open orbit $X$ of $B$ on $\n_-$. Any $N$-orbit in $X$ is a maximal coadjoint orbit of $N$ in $\n_-$. The cascade of orthogonal roots defines a cross-section $\r_-^{\times}$ of the set of such orbits leading to a decomposition $$X = N/R\times \r_-^{\times}.$$ This decomposition, among other things, establishes the structure of $S(\n)^{\n}$ as a polynomial ring generated by the prime polynomials of $H$-weight vectors in $S(\n)^{\n}$. It also leads tothe multiplicity 1 of $H$ weights in $S(\n)^{\n}$.
Let $G$ be a complex simply-connected semisimple Lie group and let $\g= \hbox{\rm Lie}\,G$. Let $\g = \n_- +\hh + \n$ be a triangular decomposition of $\g$. The authors in [LW] introduce a very nice representation theory idea for the construction of certain elements in $\hbox{\rm cent}\,U(n)$. A key lemma in [LW] is incorrect but the idea is in fact valid. In our paper here we modify the construction so as to yield the desired elements in $\hbox{\rm cent}\,U(\n)$.
Let $G$ be a complex simple Lie group and let $\g = \hbox{\rm Lie}\,G$. Let $S(\g)$ be the $G$-module of polynomial functions on $\g$ and let $\hbox{\rm Sing}\,\g$ be the closed algebraic cone of singular elements in $\g$. Let ${\cal L}\s S(\g)$ be the (graded) ideal defining $\hbox{\rm Sing}\,\g$ and let $2r$ be the dimension of a $G$-orbit of a regular element in $\g$. Then ${\cal L}^k = 0$ for any $k<r$. On the other hand, there exists a remarkable $G$-module $M\s {\cal L}^r$ which already defines $\hbox{\rm Sing}\,\g$. The main results of this paper are a determination of the structure of $M$.
A recent experimental discovery involving the spin structure of electrons in a cold one-dimensional magnet points to a validation of a Zamolodchikov model involving the exceptional Lie group $E_8$. The model predicts 8 particles and predicts the ratio of their masses. I.e., the vertices of the 8-dimensional Gosset polytope identifies with the 240 roots of $E_8$. Under the 2-D (Peter McMullen) projection of the polytope, the image of the vertices are arranged in 8 concentric circles, here referred to as the Gosset circles. The Gosset circles are understood to correspond to the 8 masses in the model, and it is understood that the ratio of their radii is the same as the ratio of the corres-ponding conjectural masses. A ratio of the two smallest circles (read 2 smallest masses) is the golden number. The conjectures have been now validated experimentally, at least for the first five masses. The McMullen projection generalizes to any complex simple Lie algebra whose rank is greater than 1. The Gosset circles also generalize, using orbits of the Coxeter element. Using results in a 1959 paper of mine, I found some time ago a very easily defined operator $A$ whose spectrum is exactly the squares of the radii $r_i$ of these generalized Gosset circles. As a confirmation, in the $E_8$ case, using only the eigenvalues of a suitable multiple of $A$, Vogan computed the ratio of the $r_i$. Happily these agree with the corresponding ratio of the Zamolodchikov masses. The operator $A$ is written as a sum of $\ell +1$ rank 1 operators, parameterized by the points in the extended Dynkin diagram. Involved in this expansion are the coefficients $n_i$ of the highest root. Recalling the McKay correspondence, in the $E_8$ case, the $n_i$, together with 1, are the dimensions of the irreducible representations of the binary icosahedral group.
The symmetric algebra g (denoted S(\g)) over a Lie algebra \g (frak g) has the structure of a Poisson algebra. Assume \g is complex semi-simple. Then results of Fomenko- Mischenko (translation of invariants) and A.Tarasev construct a polynomial subalgebra \cal H = \bf C[q_1,...,q_b] of S(\g) which is maximally Poisson commutative. Here b is the dimension of a Borel subalgebra of \g. Let G be the adjoint group of \g and let \ell = rank \g. Identify \g with its dual so that any G-orbit O in \g has the structure (KKS) of a symplectic manifold and S(\g) can be identified with the affine algebra of \g. An element x \in \g is strongly regular if \{(dq_i)_x\}, i=1,...,b, are linearly independent. Then the set \g^{sreg} of all strongly regular elements is Zariski open and dense in \g, and also \g^{sreg \subset \g^{reg} where \g^{reg} is the set of all regular elements in \g. A Hessenberg variety is the b-dimensional affine plane in \g, obtained by translating a Borel subalgebra by a suitable principal nilpotent element. This variety was introduced in [K2]. Defining Hess to be a particular Hessenberg variety, Tarasev has shown that Hess \subset \g^sreg. Let R be the set of all regular G-orbits in \g. Thus if O \in R, then O is a symplectic manifold of dim 2n where n= b-\ell. For any O\in R let O^{sreg} = \g^{sreg}\cap O. We show that O^{sreg} is Zariski open and dense in O so that O^{sreg} is again a symplectic manifold of dim 2n. For any O \in R let Hess (O) = Hess \cap O. We prove that Hess(O) is a Lagrangian submanifold of O^{sreg} and Hess =\sqcup_{O \in R} Hess(O). The main result here shows that there exists, simultaneously over all O \in R, an explicit polarization (i.e., a "fibration" by Lagrangian submanifolds) of O^{sreg} which makes O^{sreg} simulate, in some sense, the cotangent bundle of Hess(O).
Let $\frak{m}$ be a Levi factor of a proper parabolic subalgebra $\frak{q}$ of a complex semisimple Lie algebra $\frak{g}$. Let $\frak{t} = cent \frak{m}$. A nonzero element $ν\in \frak{t}^*$ is called a $\frak {t}$-root if the corresponding adjoint weight space $\frak{g}_{nu}$ is not zero. If $ν$ is a $\frak{t}$-root, some time ago we proved that $\frak{g}_ν$ is $ad \frak{m}$ irreducible. Based on this result we develop in the present paper a theory of $\frak{t}$-roots which replicates much of the structure of classical root theory (case where $\frak{t}$ is a Cartan subalgebra). The results are applied to obtain new reults about the structure of the nilradical $\frak{n}$ of $\frak{q}$. Also applications in the case where $dim \frak{t}=1$ are used in Borel-de Siebenthal theory to determine irreducibility theorems for certain equal rank subalgebras of $\frak{g}$. In fact the irreducibility results readily yield a proof of the main assertions of the Borel-de Siebenthal theory.
Let $\frak{g} = \frak{k} +\frak{p}$ be a complexified Cartan decomposition of a complex semisimple Lie algebra $\frak{g}$ and let $K$ be the subgroup of the adjoint group of $\frak{g}$ corresponding to $\frak{k} $. If $H$ is an irreducible Harish-Chandra module of $U(\frak{g})$, then $H$ is completely determined by the finite-dimensional action of the centralizer $U(\frak{g})^K$ on any one fixed primary $\k$ component in $H$. This original approach of Harish-Chandra to a determination of all $H$ has largely been abandoned because one knows very little about generators of $U(\frak{g})^K$. Generators of $U(\frak{g})^K$ are given by generators of the symmetric algebra analogue $S(\frak{g})^K$. Let $S_m(\frak{g})^K, m\in {\Bbb Z}_+$, be the subalgebra of $S(\frak{g})^K$ defined by $K$-invariant polynomials of degree at most $m$. Let $Q$ and $Q_m$ be the respective quotient fields of $S(\frak{g})^K$ and $S_m(\frak{g})^K$. We prove that if $n= dim \frak{g}$ one has $Q= Q_{2n}$. We also determine the variety, $Nil_K$, of unstable points with respect to the action $K$ on $\frak{g}$ and show that $Nil_K$ is already defined by $A_{2n}$. As pointed out to us by Hanspeter Kraft, this fact together with a result of Harm Derksen (See [D]) implies, indeed, that $A= A_r$ where $r = {2n\choose 2} dim {\frak p}$.
In this paper, Part II, of a two part paper we apply the results of [KW], Part I, to establish, with an explicit dual coordinate system, a commutative analogue of the Gelfand-Kirillov theorem for M(n), the algebra of $n\times n$ complex matrices. The function field F(n) of M(n) has a natural Poisson structure and an exact analogue would be to show that F(n) is isomorphic to the function field of a $n(n-1)$-dimensional phase space over a Poisson central rational function field in $n$ variables. Instead we show that this the case for a Galois extension, $F(n, {\frak e})$, of F(n). The techniques use a maximal Poisson commutative algebra of functions arising from Gelfand-Zeitlin theory, the algebraic action of a $n(n-1)/2$--dimensional torus on $F(n, {\frak e})$, and the structure of a Zariski open subset of M(n) as a $n(n-1)/2$--dimensional torus bundle over a $n(n+1)/2$--dimensional base space of Hessenberg matrices.
Let $Γ$ be a finite subgroup of SU(2) and let $\widetilde Γ = \{γ_i\mid i\in J\}$ be the unitary dual of $Γ$. The unitary dual of SU(2) may be written $\{π_n\mid n\in \Bbb Z_+\}$ where $dim π_n = n+1$. For $n\in \Bbb Z_+$ and $j\in J$ let $m_{n,j}$ be the multiplicity of $γ_j$ in $π_n|Γ$. Then we collect this branching data in the formal power series, $m(t)_j = \sum_{n=0}^{\infty}m_{n,j} t^n$. One shows that there exists a polynomial $z(t)_j$ and known positive integers $a,b$ (independent of $j$) such that $m(t)_j = {z(t)_j \over (1-t^a)(1-t^b)}$. The problem is the determination of the polynomial $z(t)_j$. If $o\in J$ is such that $γ_o$ is the trivial representation, then it is classical that $z(t)_o = 1 +t^h$ for a known integer $h$. The problem reduces to case where $γ_j$ is nontrivial. The McKay correspondence associates to $Γ$ a complex simple Lie algebra $\g$ of type A-D-E. We explicitly determine $z(t)_j$ for $j\in J-\{o\}$ using the orbits of a Coxeter element on the set of roots of $\frak{g}$. Mysteriously the polynomial $z(t)_j$ has arisen in a completely different context in some papers of Lusztig. Also Rossmann has recently shown that the polynomial $z(t)_j$ yields the character of $γ_j$.
A commutative Poisson subalgebra of the Poisson algebra of polynomials on the Lie algebra of n x n matrices over ${\Bbb C}$ is introduced which is the Poisson analogue of the Gelfand-Zeitlin subalgebra of the universal enveloping algebra. As a commutative algebra it is a polynomial ring in $n(n+1)/2$ generators, $n$ of which can be taken to be basic generators of the polynomial invariants. Any choice of the next $n(n-1)/2$ generators yields a Lie algebra of vector fields that generates a global holomorphic action of the additive group ${\Bbb C}^{n(n -1)/2}$. This paper proves several remarkable properties of this group action and relates it to the theory of orthogonal polynomials.
Let $(X,ω)$ be an integral symplectic manifold and let $(L,\nabla)$ be a quantum line bundle, with connection, over $X$ having $ω$ as curvature. With this data one can define an induced symplectic manifold $(\widetilde {X},ω_{\widetilde {X}})$ where $dim \widetilde {X} = 2 + dim X$. It is then shown that prequantization on $X$ becomes classical Poisson bracket on $\widetilde {X}$. We consider the possibility that if $X$ is the coadjoint orbit of a Lie group $K$ then $\widetilde {X}$ is the coadjoint orbit of some larger Lie group $G$. We show that this is the case if $G$ is a non-compact simple Lie group with a finite center and $K$ is the maximal compact subgroup of $G$. The coadjoint orbit $X$ arises (Borel-Weil) from the action of $K$ on $\p$ where $\g= \k +\p$ is a Cartan decomposition. Using the Kostant-Sekiguchi correspondence and a diffeomorphism result of M. Vergne we establish a symplectic isomorphism $(\widetilde {X},ω_{\widetilde {X}})\cong (Z,ω_Z)$ where $Z$ is a non-zero minimal "nilpotent" coadjoint orbit of $G$. This is applied to show that the split forms of the 5 exceptional Lie groups arise symplectically from the symplectic induction of coadjoint orbits of certain classical groups.
If $\frak g$ is a complex simple Lie algebra, and $k$ does not exceed the dual Coxeter number of $\frak g$, then the k$^{th}$ coefficient of the $dim \frak g$ power of the Euler product may be given by the dimension of a subspace of $\wedge^k\frak g$ defined by all abelian subalgebras of $\frak g$ of dimension $k$. This has implications for all the coefficients of all the powers of the Euler product. Involved in the main results are Dale Peterson's $2^{rank}$ theorem on the number of abelian ideals in a Borel subalgebra of $\frak g$, an element of type $ρ$ and my heat kernel formulation of Macdonald's $η$-function theorem, a set $D_{alcove}$ of special highest weights parameterized by all the alcoves in a Weyl chamber (generalizing Young diagrams of null $m$-core when $\frak g= Lie Sl(m,\Bbb C)$), and the homology and cohomology of the nil radical of the standard maximal parabolic subalgebra of the affine Kac-Moody Lie algebra.
Each infinitesimally faithful representation of a reductive complex connected algebraic group $G$ induces a dominant morphism $Φ$ from the group to its Lie algebra $\g$ by orthogonal projection in the endomorphism ring of the representation space. The map $Φ$ identifies the field $Q(G)$ of rational functions on $G$ with an algebraic extension of the field $Q(\g)$ of rational functions on $\g$. For the spin representation of $\on{Spin}(V)$ the map $Φ$ essentially coincides with the classical Cayley transform. In general, properties of $Φ$ are established and these properties are applied to deal with a separation of variables (Richardson) problem for reductive algebraic groups: Find $\on{Harm}(G)$ so that for the coordinate ring $A(G)$ of $G$ we have $A(G) = A(G)^G\otimes \on{Harm}(G)$. As a consequence of a partial solution to this problem and a complete solution for SL(n) one has in general the equality $[Q(G):Q(\g)] = [Q(G)^G:Q(\g)^G]$ of the degrees of extension fields. Among other results, $Φ$ yields (for the complex case) a generalization, involving generic regular orbits, of the result of Richardson showing that the Cayley map, when $G$ is semisimple, defines an isomorphism from the variety of unipotent elements in $G$ to the variety of nilpotent elements in $\g$. In addition if $G$ is semisimple the Cayley map establishes a diffeomorphism between the real submanifold of hyperbolic elements in $G$ and the space of infinitesimal hyperbolic elements in $\g$. Some examples are computed in detail.