Geometric rep theory and universal centralizers
Abstract.
The goal of this course is to introduce many geometric objects and tools that are in the heart of GRT. Below is the tentative plan of the course:
- •
Part I. I will talk about the geometry of complex semisimple Lie algebras and groups. Specific topics include
- (1)
The geometry of flag variety, nilpotent cone, Grothendieck-Springer resolution, Steinberg variety and an introduction to Springer theory
- (2)
The wonderful compactification of semisimple groups and more general toroidal compactifications of reductive groups.
We will follow some standard text/notes (and some additional references):
- –
Chriss-Ginzburg, Representation theory and complex geometry, mostly chapter 3 and part of chapter 2
- –
Evens-Jones, On the wonderful compactification
- (1)
- •
Part 2. We will focus on a class of completely integrable systems, called the universal centralizers (or under other names: Toda systems, regular centralizer group schemes, or bi-Whittaker reductions), associated with a semisimple (or reductive) group . This class of completely integrable systems lies at the crossroads of many different subjects in differential geometry, GRT (including geometric Langlands), (homological and enumerative) mirror symmetry, and mathematical physics (e.g. the class of Coulomb branches). Selected topics include:
- (1)
A partial log-compactification of , by Balibanu (as an application of (2) in part 1)
- (2)
Symplectic geometry of and sketch of its homological mirror symmetry (based on some of my recent work)
- (3)
Depending on interest, I will sketch the proof of a hyperKahler metric on it constructed from gauge theory (following Donaldson, Kronheimer and Bielawski).
- (1)
- •
Prerequisite: Basic knowledge of differential topology and complex geometry is assumed. Familiarity with semisimple Lie algebras and algebraic groups, as well as their representations, will be helpful but is not strictly required. We will follow the approach of Chriss–Ginzburg in recalling and supplementing the necessary background as needed.
1 Introduction
Representation theory studies some group with some structure by considering representations where is a vector space and is a group homomorphism with suitable properties. This seems like a purely algebraic question, but it is already very hard even when is a finite group (in the nonabelian case). We may now bring in various geometric methods to construct representations of :
- (1)
If acts on an algebraic variety , then there is an action of on the space of functions on ;
- (2)
If acts on an algebraic variety , then we may consider a -equivariant line bundle on (which is a line bundle on ), we may consider the cohomology groups which are representations of . If is projective, then these cohomology groups are finite-dimensional;
- (3)
If acts on , then it also acts on the cohomology for any generalized cohomology theory;
- (4)
There are also more hidden and subtle symmetries, where instead of actual group actions, we consider correspondences acting on cohomology groups. We will see examples of this later when we discuss Springer theory. In addition, we can lift this to the level of derived categories and consider Fourier-Mukai transforms;
1.1 Overview of the course.
The first part of the course will focus on the basics of geometric representation theory, largely focusing on Springer theory, the Borel-Weil-Bott theorem, and the geometry of the wonderful compactification .
The second part of the course will focus on a completely integrable system called the Toda system and its homological mirror symmetry. One motivation for this is the notion of categorification, for example going from the Hecke algebra to the Hecke category, which will turn out to be a category of constructible sheaves.
One origin of categorification in algebraic geometry is Grothendieck’s function-sheaf correspondence, which we will adapt to our setting (in particular, we will not need -adic sheaves). One particularly simple example of this is upgrading the integers to the derived category of vector spaces over a field , where the original integer is recovered as the Euler characteristic.
For a more sophisticated example, we may consider a smooth manifold with a stratification by locally closed submanifolds. We may consider the space of constructible functions (taking values in integers) with respect to , which are locally constant on each stratum . The natural categorification of this is the derived category of constructible sheaves on , where all cohomology sheaves are locally constant on each stratum . The original constructible function is recovered by taking the Euler characteristic of the stalk at each point. In addition, we may recover the functions as
1.2 Homological mirror symmetry.
We will now discuss a connection to symplectic geometry. If we have a stratification , then we may consider the union
of conormal bundles to the strata and consider the microlocal sheaf theory. In particular, each constructible sheaf has a microlocal stalk at each as in Figure 1. For example, if we have a restriction map , then the microlocal stalk is
where the shift is not canonical. One upshot of this approach is that
The following theorem inspired many subsequent works.
Theorem 1.1 (Nadler-Zaslow).
The dg-category of constructible sheaves on equivalent the Fukaya category of the cotangent bundle of .
We will now discuss a geometric interpretation. We may consider the characteristic cycle of a constructible sheaf, which is a conic Lagrangian cycle in . However, the objects of the Fukaya category are smooth Lagrangian submanifolds with the cochain complexes of morphisms being given by counting interesection points and holomorphic polygons. See Figure 2 for a visualization. It is usually very difficult to calculate the Fukaya category directly, but the theorem allows us to translate these questions into questions about constructible sheaves, which are more manageable.
Sketch of proof.
We will begin by discussing why there is a fully faithful functor . The first step is to find some nice generators of the constructible sheaf category. For this, we consider open embeddings and extension by zero of the constant sheaf on . These have characteristic cycles given by the conormal bundle to the boundary (together with the zero section at ). We will send this sheaf to a smoothing of its characteristic cycle, as depicted in Figure 3.
To prove essential surjectivity, all of the Hom complexes in the Fukaya category can be calculated using Morse theory. For more details, see the original paper by Nadler-Zaslow. □
One reason to consider the Fukaya category is homological mirror symmetry, which was originally proposed by Kontsevich. For certain symplectic manifolds , there is a mirror partner , which is a complex algebraic variety, and Kontsevich conjectured that there is an equivalence of categories
The simplest example is when , where we need to consider the fully wrapped Fukaya category. We have an equivalence
(modulo technical details), but the upshot is that all characteristic cycles are supported inside the zero section. Local systems are on are simply representations of , which are simply modules over the group algebra . Ignoring finiteness conditions, these are simply quasicoherent sheaves on .
Remark 1.2.
For the purposes of representation theory, we will treat as cocharacters, which will be identified with the characters on the mirror side. This picture generalizes if we consider algebraic tori of any rank and is a version of Langlands duality in the abelian case.
Remark 1.3.
Homological mirror symemtry is closely related to the geometric Langlands correspondence, where we send the Hitchin system for to the character variety for the Langlands dual group . There are some technical issues, but the Betti geometric Langlands correspondence is essentially homological mirror symmetry here.
One modification of this is to complexify the Fukaya side into , and in this case the (appropriate version of the) wrapped Fukaya category still reduces to local systems on . Traditionally, we considered (the cotangent bundle of the compact form) because of the SYZ mirror symmetry philosophy, which predicts that a manifold and its mirror have dual Lagrangian torus fibrations. In general, it is very hard to use SYZ to construct mirrors because there are usually singular fibers, which are hard to control.
However, one way to produce a mirror is to believe in homological mirror symmetry and then try to find a variety such that
If the fibration has a Lagrangian section, then each smooth fiber becomes a Lie group, so is expected to pick up a “convolution” symmetric monoidal structure, which should correspond to the tensor product of coherent sheaves in some kind of non-linear Fourier transform.
In the second part of the course, we will generalize the duality to the universal centralizer (or Toda system) , where is a semisimple complex Lie group. This is related to representation theory for two reasons.
- •
is also called the bi-Whittaker reduction and represents something like the generic part of the representation of , namely the “commutative part” of the Hecke category;
- •
If we add a loop to , then we obtain the affine Toda system;
- •
There is a partial (log) compactification over the base constucted by Balibanu whose fibers are a kind of regular Hessenberg variety (a generalization of Springer fibers). The central fiber is the Peterson variety , which is a closed subvariety of the flag variety and is related to the quantum cohomology of via the isomorphism
1 Geometry of complex semisimple Lie algebras and groups
2 Review of semisimple Lie algebras/groups
Eventually, the groups we are interested in are , , and with Lie algebras , , and . However, along the way we will encounter groups like Borels and unipotent groups.
2.1 Generalities.
Every Lie group has both a left and right action of on itself given by the multiplication map
We may consider the tangent space at the identity, and left translation of vector fields gives an isomorphism
The right action gives another trivialization
These two are related by the diagram
where is the map sending an element of to its inverse. If we fix , then applying gives us a left-invariant vector field on , and similarly applying gives a right-invariant vector field .
Lemma 2.1.
- (1)
Left (resp. right) invariant vector fields on are closed under the Lie bracket;
- (2)
The left and right-invariant vector fields induce the same Lie bracket on .
The main idea of the proof is that the Lie bracket is preserved by diffeomorphisms and in particular left (or right) translation.
Remark 2.2.
If we have any action , then taking the derivative
gives an infinitesimal action of on (as in a map from to vector fields on ). For example, is simply
and a similar statement is true for .
One important action of on itself is the adjoint action, where acts on itself by conjugation. On the tangent space, we have . By definition, we now have
which induces a map
Therefore, another equivalent definition of the Lie bracket is to simply differentiate the adjoint action. There is also an exponential map , and the Lie bracket is defined as
Remark 2.3.
Any finite-dimensional abstract Lie algebra is isomorphic to a subalgebra of by Ado’s theorem, which says that every finite-dimensional Lie algebra has a finite-dimensional faithful representation. In particular, every finite-dimensional Lie group over is locally an immersed subgroup of for some .
We will give some principles of Lie theory:
- •
The Lie group is (up to taking its universal cover) determined by its Lie algebra. Given a connected Lie group , then any discrete covering inherits a unique group structure whose kernel lies in the center of (this works because the fundamental group of is abelian due to the group structure). Conversely, any discrete central quotient of is a Lie group with the same Lie algebra. If we work analytically, has a universal cover, which is the unique simply connected Lie group with the same Lie algebra as .
- •
If and are Lie groups such that is connected, then any Lie group homomorphism
is uniquely determined by its derivative , which is a Lie algebra homomorphism.
- •
For and any embedding , the subgroup of generated by is an immersed Lie subgroup of with Lie algebra . For example, if and is a line with irrational slope, then is immersed rather than embedded.
- •
Let and be finite dimensional Lie groups with connected and simply connected. Then a map is the differential of a Lie group homomorphism if and only if it is a Lie algebra homomorphism. The standard trick to prove this is to consider the graph
which is a Lie algebra embedding. Then we get an immersed subgroup of (this receives a map from because is simply connected), which we project down to .
By Ado’s theorem, we see that the embedding gives an immersed subgroup . In addition, the third item actually implies that there is a unique simply connected Lie group for any Lie algebra .
Example 2.4.
Let be a projective variety. Then studying the identity component is similar to studying
which is a standard result in deformation theory. For example, if is a partial flag variety, then in almost all cases, but if (the full flag variety), then we really do have
In particular, we have
and when is semisimple, the center is discrete.
We will call a Lie algebra homomorphism
a representation of . For example, we previously considered the adjoint representation
We can also take the direct sums and tensor products of representations. It should be noted that the action on should take the form
in analogy with differentiating the action of on , which takes the form
2.2 Rough classification of Lie algebras.
Given a Lie algebra , there is a lower central series
and the smaller derived series
Definition 2.5.
A Lie algebra is nilpotent if the lower central series becomes zero at some point.
One example of a nilpotent Lie algbera is if we consider upper-triangular matrices with zeroes on the diagonal (usually denoted ).
Definition 2.6.
A Lie algebra is solvable if the derived series eventually vanishes.
An example of a solvable Lie algebra is the algebra of all upper-triangular matrices. The reason for this is that . If we write
then we see that , , and .
Theorem 2.7 (Engel).
Let be a Lie subalgebra such that every is a nilpotent endomorphism of . Then there exists a nonzero such that for all .
Engel’s theorem implies that all can be made (strictly) upper-triangular simultaneously.
Theorem 2.8 (Lie).
Let be a solvable Lie subalgebra over . Then there is a common eigenvector for all .
In particular, Lie’s theorem implies that all can be simultaneously made upper-triangular.
We will now consider all solvable ideals in . In fact, if we have a finite collection of solvable ideals, then the sum is still solvable, so there exists a maximal solvable ideal of , called the radical of and denoted . This fits into the exact sequence
2.3 Semisimple and reductive Lie algebras.
Definition 2.9.
A Lie algebra is semisimple if .
Example 2.10.
The Lie algebra is semisimple.
Definition 2.11.
A Lie algebra is reductive if .
Remark 2.12.
There are other characterizations of reductive Lie algebras, for example that the nilradical (maximal nilpotent ideal) , which fits into the exact sequence
vanishes.
We now give some important properties of semisimple Lie algebras.
Theorem 2.13.
Let be a semisimple Lie algebra. Then every finite-dimensional representation of decomposes uniquely as a direct sum of irreducible representations.
Sketch of proof.
There are two ways to do this, either algebraically (which is tedious) or by using Weyl’s unitary trick. The key fact using the unitary trick is that every complex semisimple Lie algebra has a unique compact real form satisfying (being a real form) and being the Lie algebra of a simply connected compact Lie group.
For example, the compact real form of is (traceles anti-Hermitian matrices), which corresponds to . The unitary trick is to restrict to the action of the compact form on and then try to find a -invariant Hermitian metric on , which will split for free. The way to find the inner product is to pick any Hermitian metric and a Haar measure on such that
and define a new inner product by
(which is the same trick we use for finite groups). Finally, the -representations lift to , so we are done. □
Theorem 2.14.
In fact, a Lie algebra is semisimple if and only if all of its finite-dimensional representations are semisimple.
Proof.
Consider the adjoint representation . Because it is semisimple, we have a decomposition
In particular, we must have , where .
If , then by Lie’s theorem says that any representation must have an eigenvector, so all irreducible representations must be -dimensional. This implies that
An abelian Lie algebra has indecomposable representations which are not semisimple, for example representing a line by matrices of the form , so the radical must have vanished to begin with. □
Remark 2.15.
Levi’s theorem actually states that .
Theorem 2.16 (Preservation of Jordan decomposition).
Let be semisimple. For any , there exist such that and for any finite-dimensional representation
the Jordan normal form of has and .
Remark 2.17.
If we consider the adjoint representation of (which is faithful), then in fact the condition on the Jordan decomposition in any representation actually implies that .
Example 2.18.
If we consider , then the representation
has and not in the image of in general.
Definition 2.19.
If is semisimple, then we say is semisimple if and nilpotent if .
We will now consider reductive Lie algebras. In fact, by Levi’s theorem and the fact that , is reductive if and only if it is the product of a semisimple and an abelian Lie algebra, which is equivalent to .
Definition 2.20.
A Lie group is semisimple if its Lie algebra is semisimple.
We will now consider all complex semisimple Lie groups, which are actually affine algebraic groups. By Borel’s theorem, they are actually all linear, meaning closed subgroups of some . If is semisimple, then there is a simply connected form and a tower down to the adjoint form .
Example 2.21.
For the Lie algebra , then the simply connected form is and the adjoint form is .
Example 2.22.
If we consider , then the simply connected form is and the adjoint form is .
Example 2.23.
For , there is a form , which for has fundamental group , so the simply connected form is the universal cover and the adjoint form is .
Remark 2.24.
We may also consider simple Lie algebras, which are semisimple Lie algebras with no nontrivial ideals. Then the classical simple Lie algebras are the ones we have already listed and are the only infinite families.
There is also the Cartan-Killing criterion of semisimple Lie algebras.
Definition 2.25.
An invariant bilinear form on is given by
satisfying the condition that
Definition 2.26.
The Killing form is the form
Because the trace has cyclic invariance and the trace is invariant under conjugation, this is invariant and symmetric.
Theorem 2.27 (Cartan-Killing criterion).
A Lie algebra is semisimple if and only if the Killing form is nondegenerate.
In particular, the means that the adjoint and coadjoint representations are isomorphic.
Remark 2.28.
If , then for all , so for the Killing form.
We now briefly pause to consider issues of analytic vs algebraic categories. If we are considering semisimple Lie groups over , then they are in fact also (linear) algebraic groups. However, if we leave the semisimple setting, then there are differences. For example, if we consider , then the algebraic options are , , and compact -dimensional tori, namely elliptic curves. Usually, in algebraic groups theory, we will only consider affine algebraic groups.
Theorem 2.29 (Borel).
Every affine algebraic group is linear, and has a closed immersion to some .
In the analytic world, there is the map
which is not algebraic. In particular, and are totally different and have no nontrivial morphisms between them (and we have to consider their formal completions to recover the exponential map). In fact, the category is semisimple (prove this using the unitary trick for ), while the same is not true for (consider the representation by matrices of the form ). However, any analytic Lie group over not containing a compact torus is algebraic.
We now proceed to defining reductive groups. Recall that we essentially need to figure out what to do with the center of a reductive Lie algebra . In the algebraic group world, we will consider tori, which are groups isomorphic to products of copies of .
Definition 2.30.
A reductive group is a finite quotient of the product of an (affine) algebraic torus and a semisimple algebraic group.
In particular, we have
where is semisimple and might be disconnected. For example, , which is clearly not connected. Alternatively, there is a quotient morphism
where again . Here, we see that is always connected. Using Langlands duality, we can produce a cover by something with connected center. A third alternative is to rewrite the product as
where .
Example 2.31.
In the case where , we have and , so we have
Reductive groups are characterized by complete reducibility of their finite-dimensional representations (in characteristic zero). In addition, algebraic and analytic morphisms between reductive groups agree, so we no longer need to worry about the difference between algebraic and analytic categories. Representations of reductive groups also satisfy the preservation of Jordan decompositions, namely that for any , there exist such that and for any finite-dimensional representation
the Jordan decomposition of satisfies and .
Lemma 2.32.
If is semisimple, then any invariant bilinear form is symmetric. In addition, any invariant non-degenerate bilinear form is a scalar multiple of the Killing form.
Proof.
Any invariant bilinear form induces a map
as morphisms of -representations. Recall that decomposes into a product
of simple Lie algebras, so we have a morphism
Using the Killing form, we have an isomorphism , so by Schur’s lemma, the map is zero if and a scalar multiple of the identity if . Therefore, any invariant bilinear form is a sum of scalar multiples of Killing forms on each simple factor, which are all symmetric. The second statement follows similarly. □
2.4 Cartan subalgebras.
We will begin by considering general and specialize to semisimple ones later.
Definition 2.33.
If is a Lie subalgebra, then the normalizer of in is given by
Definition 2.34.
A subalgebra is a Cartan subalgebra if
- (1)
is nilpotent;
- (2)
(as in it is equal to its own normalizer).
We will see later that every Lie algebra has a Cartan subalgebra.
For any , let
denote the characteristic polynomial of . If we choose coordinates on , then each is a homogeneous polynomial in of degree .
Definition 2.35.
The rank of is the least such that is not identically zero.
Example 2.36.
If is abelian, then for all , so has rank .
Remark 2.37.
Note that in the nonabelian case because is never zero and because for all and therefore for all .
From now on, we will denote the rank of by .
Definition 2.38.
We say that is strongly regular if .
Remark 2.39.
We will see another notion of regularity later, so we call this strongly regular to distinguish it.
We will denote the set of strongly regular elements by . The definition of a strongly regular element implies the following.
Proposition 2.40.
The set is a connected open dense subset of (in the Zariski topology and hence also the analytic topology).
Theorem 2.41.
If , then the generalized eigenspace of for , denoted , is a Cartan subalgebra of .
Note that by definition, we have , so the dimension of this Cartan equals the rank of . In fact, we have already described all Cartan subalgebras.
Theorem 2.42.
Let be a Lie algebra over . Then all Cartan subalgebras of are of the form for some .
These Cartan subalgebras are all conjugate under the action of
so in particular acts transitively on the set of Cartan subalgebras.
Remark 2.43.
If is reductive, then .
2.5 Root systems.
We will now specialize to the semisimple case, which is much more explicit.
Theorem 2.44.
Let be a Cartan subalgebra of a semisimple Lie algebra . Then
- (1)
is abelian;
- (2)
is its own centralizer, i.e. ;
- (3)
Every element of is semisimple.
- (4)
The restriction of the Killing form to is nondegenerate.
Example 2.45.
If , then the Cartan subalgebras are precisely the conjugates of the diagonal matrices. This has dimension (given the constraint that the trace vanishes), so the rank of is .
For a semisimple Lie algebra, if , then in fact for a general choice of . The necessary condition on is that it is regular semisimple, which we can take to simply mean that is as small as possible. Because we can simultaneously diagonalize all for , we have a decomposition
where
and is the set of nonzero such that .
Example 2.46.
Returning to , if we let
then twe see that
If we refer to the coordinates on as , then the roots are given by
where . We refer to the set as the roots of with respect to and as the root space corresponding to .
Because the Killing form is nondegenerate when restricted to , we may identify with . In particular, we may consider as a subset of itself.
Corollary 2.47.
- (1)
is a maximal abelian subalgebra of ;
- (2)
Every strongly regular element is semisimple;
- (3)
acts transitively on the set of all Cartan subalgebras of .
Warning 2.48.
Not all maximal abelian subalgebras are Cartan subalgebras, for example the nilpotent matrices in .
Lemma 2.49.
Assume that is semisimple. Let (usually, it is not connected). Then is a maximal torus in with .
Sketch of proof.
Choose a compact real form . We now consider the set
Now let be strongly regular. Then is a Cartan subalgebra of . Next, the normalizer
is a maximal torus of of real dimension , so we can conclude the proof by complexifying everything. □
Example 2.50.
In the case of , if is strongly regular, then we see that , and so , so must be semisimple. In particular, the root space decomposition is
Lemma 2.51.
Choose a root decomposition
Then is strongly regular if and only if there exists such that for all .
Proof.
Note that is strongly regular if and only if , which holds if and only if is nonzero for all , which holds if and only if for all . □
Remark 2.52.
From now on, we will refer to strongly regular elements as regular semisimple elements in the semisimple case and refer to the set of such elements as . We will also denote
which is given by the equation
Example 2.53.
The Lie algebra is generated by the matrices
with the relations
Then we see that the Cartan is and the root spaces are and , where . Therefore, the root system is and is shown in Figure 4. In this case, a choice of is the same as choosing a set of positive roots. In this case, is also a simple root, but in general positive and simple roots are not the same.
Now set , which is a simple coroot. Then we have . We may now consider the maximal torus
The adjoint action of on is given by
Note that , which matches the exponent. We may now consider the cocharacter lattice
If we project down to the adjoint form we see in fact that
Note that this cocharacter lattice is dual to the root lattice and is called the coweight lattice. We can now identify the character lattice of with the weight lattice , whereas is the root lattice.
Now consider the exact sequence
This induces the exact sequence
In particular, we see once again that . If we instead consider maximal tori, we have
which induces the exact sequence
on fundamental groups. This gives us the identification
and relates the coroot and coweight lattices.
Example 2.54.
In the case of , the root system is drawn in Figure 5. In particular, the roots are given by
In order to treat the coroots, note that
for any root . Therefore, we may choose and such that
For example, if we consider , then we may choose
We now have the coroot such that .
To choose a basis, we cut by a general linear hyperplane and choose a positive direction. This gives a set of positive roots. For example, we will choose for . Inside this, there is a unique set of simple roots satisfying
- (1)
It is a basis of ;
- (2)
Any is a nonnegative integer combination of elements of .
For example, in this case, we have . The simple coroots are then simply . Then the Cartan matrix is given by
We should also explain why the picture has the angles as shown. If we recall that the root lattice is identified with the character lattice , this is proved by considering the kernel . In particular, anything killed by must actually act trivially on the root spaces and acts trivially on by definition, so it lives in the center of , which is trivial. Continuing as before, we consider the coweight lattice, which is spanned by the matrices
Remark 2.55.
The Cartan matrix is symmetric if and only if the Dynkin diagram is simply laced, i.e. has no multiple edges.
Unfortunately, we will not be able to discuss root systems in detail for time reasons. In type A (where ), the picture generalizes that of , where we can choose and . As a remark, we will also refer to the weight lattice as and the coweight lattice as . In general, we will interpret the fundamental weights as the dual basis to the simple coroots .
In the group picture, if we have an intermediate form with maximal torus , then we have the inclusions
induced by
To find the center of , we simply have the quotient , which comes from the long exact sequence on homotopy groups coming from
To find the fundamental group of , we apply this to the exact sequence
which gives us . In particular, this specializes to
For example, if , then we have
Because this group has prime order, there are no intermediate forms between and .
Using the Killing form, there is the formula
and this actually implies that
Because this needs to be an integer, we see that for all roots, the angle between them must satisfy . This restricts the possible angles to , or unless .
Remark 2.56.
Root systems are classified by Dynkin diagrams, which is equivalent to classifying Cartan matrices. The entries of the Cartan matrix satisfy and . Non-symmetry can be explained by the fact that in some root systems, and may have different lengths. This gives arise to the notions of long roots and short roots. In fact, there are only two possible lengths for all roots (under the Weyl group action, any root can become simple).
For an example of a root system with two lengths, consider the root system, whose root and coroot systems are shown in Figure 7.
2.6 Weyl groups.
Recall that the maximal torus has Lie algebra . Then we define the Weyl group
Here, we know that , so .
Example 2.57.
If , then because we can permute the diagonal entries.
The Weyl group acts on the set of roots. Each root corresponds to a reflection given by
This satisfies and fixes the hyperplane pointwise.
In the example, the root corresponds to the reflection that swaps the -th and -th coordinates. In particular, its kernel is the set of all matrices satisfying .
The Weyl group is generated by the simple reflections, which correspond to the simple roots . We will refer to the simple roots as and the corresponding reflections as . Then, the Weyl group has the presentation
and is a finite reflection group.
The reflection hyperplanes cut the space into a set of chambers, which are fundamental domains for the Weyl action and are in bijection with the choices of positive roots . In particular, the set corresponds to the chamber
Note also that the choice of a set of positive roots corresponds to a choice of Borel . In particular, there is the direct sum decomposition
In addition, for any we have
2.7 Regular elements.
Let . Note that acts freely on . Finally, let .
Proposition 2.58.
Let . Then the centralizaer contains an -dimensional abelian subalgebra.
Proof.
Since the regular semisimple locus in dense, there exists a sequence of regular semisimple elements. For each , its centralizer is -dimensional ane abelian.
Finally, the compactness of the Grassmannian implies that there is some subsequence which converges to some limit . We need to check that , but we know that if , there is some sequence such that . Because the Lie bracket is continuous on both arguments, we see that
We now need to check that is abelian. But this follows similarly: if , there are sequences and such that . Then
Corollary 2.59.
For all , then and . In addition, every maximal abelian subalgebra has dimension at least .
Definition 2.60.
An element is called regular if . Equivalently, the adjoint orbit has maximal dimension.
Example 2.61.
An example of a regular element which is not semisimple is
Its centralizer is given by
and is -dimensional.
We will now discuss a general criterion for regularity. Recall the Jordan decomposition
where is semisimple and is nilpotent. Because , we have
Proposition 2.62.
The element is regular if and only if is regular in the derived algebra of .
Remark 2.63.
The centralizer is a reductive subalgebra of , called a Levi subalgebra. Here, if for some Cartan and
is the root decomposition, then we have
We now consider the root system
This is the root system for (note we need to take the derived subalgebra to eliminate the center). Finally, we note that
so the rank of the root system increases as we move to smaller strata of (with respect to the stratification by root hyperplanes). The structure of the Levi subalgebra is in fact
where the center is and the semisimple part has root system .
Example 2.64.
In , the matrix
lies in the root hyperplane and its Levi subalgebra is given by traceless block-diagonal matrices of the form
Taking the derived subalgebra, we obtain a copy of in the upper-left block. The center is simply matrices of the form
In , we can group the blocks by shared eigenvalues and proceed similarly. If the blocks have sizes , then the derived subalgebra of the Levi is given by and the center has dimension .
Proof of Proposition 2.62.
Note first that
Set-theoretically, we have
If , then , where the orthogonal complement is taken in . Therefore, we know that is regular if and only if , which is equivalent to being regular in . □
Example 2.65.
We may assume that is already in Jordan normal form. Then in fact it is regular if and only if all Jordan blocks have distinct eigenvalues. Here, the Levi subalgebra is given by block-diagonal matrices where the blocks are grouped by shared eigenvalues. However, any regular nilpotent element in has a single Jordan block, so we see that each eigenvalue of must correspond to a single Jordan block.
2.8 Flag varieties and the Bruhat decomposition.
We will now discuss Borel subalgebras .
Definition 2.66.
A Borel subalgebra in is a maximal solvable subalgebra. A Borel subgroup in is a maximal connected solvable subgroup.
Note that if we have a Borel subalgebra, we can integrate it to get a Borel subgroup . Here are some properties:
- (1)
All Borel subalgebras are conjugate under the adjoint representation;
- (2)
If is a Borel subalgebra, then and ;
- (3)
Every Borel subalgebra contains a Cartan subalgebra. On the other hand, a fixed Cartan is contained in -many Borel subalgebras, which form a -torsor. In fact, they are in bijection with the set of choices of positive roots via
For example, we can let be the diagonal matrices in , and we can generate all Borels by taking the upper-triangular matrices as a base point and conjugating by permutation matrices.
- (4)
If is regular semisimple, then in fact .
Now let be the set of Borel subalgebras of . We can give it the structure of an algebraic variety as follows. First, consider the Grassmannian
which is naturally an algebraic variety which carries a -action induced from the adjoint representation. Then sits inside as a subset which is in fact the orbit of any Borel under this action. This gives
It is easy to check that is in fact closed, so it is a projective variety called the flag variety of (and the quotient description means it is a homogeneous space).
Remark 2.67.
Because the Borel subgroup always contains the center , the flag variety is insensitive to the choice of intermediate form of . If is reductive, then this construction kills the center, so we only see the semisimple part of in the flag variety.
Example 2.68.
If (or ), then in fact
is the usual flag variety of . This clearly admits a -equivariant surjection from the variety of frames
If we consider the standard flag
then the stabilizer of this flag is the standard Borel subgroup of upper-triangular matrices, so we see that .
Now we may restrict the -action on to the Borel , where we choose a particular Borel as a base point. Then the set of -orbits is denoted by
We may also consider the -action on and restrict to the diagonal subgroup . Then the set of -orbits is given by
To check that this is well-defined, suppose that . This follows from the fact that
To check surjectivity, we look at the orbit of , but we can bring to , and surjectivity follows. To check injectivity, we see that if , then for some . But if , then and thus .
We will now consider a more refined bijection which will give the Bruhat decomposition. Fix a maximal torus , or equivalently a Cartan subalgebra . Then we have a map
where we choose a representative of a coset in . Because , this map is well-defined. Taking stock, we now have maps
Our goal is to show that the final arrow is a bijection.
Theorem 2.69 (Bruhat decomposition).
The map is a bijection.
Remark 2.70.
This implies the decomposition
Example 2.71.
If and we choose is the upper-triangular matrices with the diagonal matrices, then we can present as a permutation matrix, and the Bruhat decomposition says that any invertible matrix can be written as a product of an upper-triangular matrix, a permutation matrix, and another upper-triangular matrix.
The proof of this theorem is based on the Bialynicki-Birula decomposition, which is an analog of Morse theory. If we have an action of on a projective variety , then we may consider the fixed point locus . We will assume for now that is discrete (hence finite).
Example 2.72.
If , then we consider the action with fixed points and . For any other point, we consider the limit
In this case, the attracting locus of is , while the attracting locus of is just the point .
For each fixed point , we consider the tangent space , which is a representation of . We can decompose it into weight spaces
where is the subspace on which acts by multiplication by . By assumption, there are no zero weights, so we have the decomposition
where the first summand is the sum of positive weight spaces and the second summand is the sum of negative weight spaces.
Theorem 2.73 (Bialynicki-Birula decomposition).
The variety decomposes into a disjoint union
where the attracting manifold is the locally closed subvariety
and is also given by
Proof of Bruhat decomposition.
We will establish bijections
Note that the third bijection follows directly from the Bialynicki-Birula decomposition, so we need to establish the other bijections.
First note that there is a bijection between -fixed points in and the set of all Borels . This is simply the infinitesimal version of the fact that if , then . Now this is in bijection with the set of for , which establishes the first bijection.
Now choose a cocharacter which is regular and a primitive nonzero cocharacter of . For this -action, the fixed points are in bijection to the set of all Borel subgroups containing . Taking the derivative at the identity, this is in bijection with
Because is regular, then , so we have recovered the set of all .
It remains to identify -orbits with attracting sets of the -fixed points. If is a Borel, then we consider its attracting set
The attracting sets are the positive parts of the tangent spaces at the fixed points, and so we must have because . Note that the action of on decomposes into eigenspaces
where will disappear when we quotient by . Our goal is to have
Considering the root decomposition
considering -weights gives us
There is a Weyl chamber (see Figure 6 for the example) which pairs with the positive roots positively, so we are free to choose in this chamber. Once we choose , then the inclusion factors through .
Remark 2.74.
One convenient choice for a regular cocharacter is , where
This implies that
which lies in the coroot lattice . By definition, we have for all simple roots, so it pairs positively with all positive roots.
Now let such that . Then if we consider the orbit , we have
Dimension counting shows that is open and dense in , so our goal is to show that it is also closed, i.e. that .
Lemma 2.75.
Suppose that is a unipotent group acting on an affine variety . Then any -orbit in is closed.
Proof.
Let be a -orbit and be its closure. Suppose that is nonempty (and hence is a closed subvariety of ) and has defining ideal . Because is -stable, so is .
Lemma 2.76.
For all , the span is finite-dimensional.
Proof.
The action of on gives a coaction
Then we see that
and thus
which is finite-dimensional. □
Now consider . Choose . Then is -invariant, so it must be constant on , and hence constant on . But we know that vanishes on , so it must vanish on , which contradicts the fact that is nonzero. Therefore, must be empty and is closed. □
The two lemmas together emply that . □
Example 2.77.
In the case when , then the flag variety if , and if we choose to be the upper-triangular matrices with the diagonal matrices, then the Weyl group is
and the nontrivial element swaps the upper-triangular and lower-triangular Borel subgroups. See Figure 8.
Note that the Bruhat decomposition even refines to
and gives a cell decomposition of into affine spaces (which are even dimensional). This allows us to easily compute the homology and cohomology of , which is given by
We can compute the dimension of each cell as follows. Note that
where is the unipotent radical of the opposite Borel. Now we compute
which is called the length of relative to . For example, the identity has length and the longest element has length .
Example 2.78.
In the case, the flag variety has dimension
The cohomology has dimension
Later, we will see that this is actually a representation of the Weyl group and describe its algebra structure.
Remark 2.79.
In fact, a Borel subgroup is a minimal subgroup of such that the quotient is projective.
Theorem 2.80 (Fixed point theorem of Borel).
A connected solvable linear algebraic group acting algebraically on a non-empty complete variety has a fixed point.
2.9 Parabolic subgroups.
If is a homogeneous space which is compact, we can consider the action of a Borel subgroup on . By Borel’s fixed point theorem, this has a fixed point, say . But then the stabilizer of this point contains , and therefore
so contains a Borel. Conversely, if contains a Borel, then admits a surjection from the flag variety and is hence projective.
Definition 2.81.
A parabolic subgroup of is a (necessarily connected) subgroup which contains a Borel subgroup.
Remark 2.82.
We can drop connectedness from the definition because is in fact always connected.
Proposition 2.83.
A subgroup is parabolic if and only if is projective.
We can in fact “classify” parabolic subgroups. If we fix a Borel , then we can ask for all parabolic subgroups containing . Because is connected, we can reduce to the corresponding problem for Lie algebras, namely finding such that . We may consider the adjoint action of on , which decomposes as
The first two factors give , so our goal is to determine the possible . We will not discuss the details and only state the end result.
Proposition 2.84.
There is a one-to-one correspondence between and subsets of the simple roots. It is given by
If we restrict the Cartan matrix to , we obtain a root system
which decomposes into irreducible root systems. We can then decompose
where is the nilradical and is the Levi factor, then has root system determined by . The center is given by
In particular, we have
For example, in the example, if we choose to be one simple root, then we see only roots living in the span of that simple root, which gives us a copy of .
Example 2.85.
If and is the upper-triangular matrices, then the parabolic subalgebras are simply block upper-triangular matrices and the Levi factor is given by block-diagonal matrices. The root system of the Levi is determined by the sizes of the blocks, and the center is given by scalar matrices in each block.
Remark 2.86.
The varieties are known as partial flag varieties. In the example, they are given by flags
where we require that is the size of the -th block.
2.10 Representation theory.
We will give a classification of irreducible representations of .
Theorem 2.87.
Fixing , the (isomorphism classes of) irreducible representations of are in one-to-one correspondence with dominant weights , where
is the nonnegative integral span of the fundamental weights.
In the example, this corresponds to the shaded region in Figure 5.
Remark 2.88.
The case of groups is more subtle because of fundamental group issues. However, representations of are “the same” as representations of the simply connected form . This happens because we can integrate to . In particular, this means that irreducible representations of are the same as positive coweights .
The way we do this is by considering the restriction to for any -representation. These are all semisimple by preservation of Jordan decomposition. This gives a decomposition
so because we can integrate to , the -action upgrades to a -action, and we have a decomposition
We can now consider the action of the unipotent radical on , which will not preserve the weight decomposition.
has a nonzero fixed point because the -action on has a fixed point by Borel’s fixed point theorem. If we consider the action of on , then
where is the positive part of the root lattice. We will say that if . In this language, we have
Definition 2.89.
A highest weight vector of is a nonzero vector such that .
Theorem 2.90.
Every irreducible representation has a unique highest weight vector up to scaling, and this is called the highest weight. Conversely, is uniquely determined by its highest weight.
For an arbitrary finite-dimensional representation , we can compute the decomposition
by considering the action of on . To decompose, we look for the fixed subspace , which is the span of the highest weight vectors. Finite-dimensional representations of are classified by . For all , let be the corresponding irreducible representation of . Then the associated line bundle
on is -equivariant.
Theorem 2.91 (Borel-Weil).
The global sections are isomorphic to as -representations.
Remark 2.92.
There is a generalization of this result, called the Borel-Weil-Bott theorem, which calculates the higher cohomology of these line bundles.
We will not give a detailed proof of the Borel-Weil theorem. However, the idea is that a section of is a pair where (here, we always use the simply connected form and drop the subscripts), and we must be invariant under the -action
and therefore we must have
for all and . Equivalently, we can ask for
Then we are looking for . To identify this representation, we need to find the highest weight with respect to or . It is a bit easier to work with , so we can pass to . Then we consider
Using the Bruhat decomposition
we will use the longest element (for this is the permutation which exchanges and , and , etc.) to write
Then is open, so we can consider the behavior of functions on this open. This decomposes as the Cartesian product
so we are looking for functions in . This gives us the -dimensional space , so the highest weight space we are trying to calculate is at most -dimensional. It remains to show that has no poles outside of . Using the Bruhat decomposition again, the codimension orbits with respect to are given by for simple reflections .
Lemma 2.93.
For all simple coroots , let be the dual fundamental weight. Then .
Now if , we see that is globally defined and that
If , then
with some of the being negative, so has a pole along of order . Thus, acts on with weight , and the end result is that the representation must be .
Remark 2.94.
Note that because of the Bruhat decomposition
and therefore the the exact sequence
has the first object given entirely by principal divisors . Equivalently, we see that is a UFD.
However, if we move away from the simply-connected form, then there are nontrivial line bundles, and
Theorem 2.95 (Peter-Weyl).
Consider the action of on . Then we have a decomposition
We may obtain this decomposition by considering -fixed vectors and obtaining
which has weight for in the first factor and weight for in the second factor. In the other direction, we associate the tensor with the function
3 Springer theory
3.1 The Grothendieck-Springer resolution.
Proposition 3.1.
Let and be two Borel subalgebras of . Then there is a canonical isomorphism .
Proof.
Because all Borels are conjugate, we have
This isomorphism preserves the nilradical, so it induces an isomorphism on the quotients. For a different that conjugates to , we have for some , and it remains to check that
is the same as
Passing to the infinitesimal version, the action of on is zero because the adjoint action of on itself sends everything to the nilradical, so it integrates to the trivial action of . □
Definition 3.2.
The quotient is called the universal Cartan of and is denoted by .
Remark 3.3.
We must note that is not a subalgebra of . Instead, we should view it as a pair . Indeed, for all such pairs, we can identify and by
If is a different Borel, then , and therefore relating the sequences via
requires conjugation by and thus requires the choice of a Borel. However, we do end up with a pair , which gives a universal set of positive roots containing a universal set of simple roots. We will call this a “pinned” version of the usual Cartan. This allows us to discuss the Weyl group and the universal simple reflections. Here, we note that any choice of gives a canonical isomorphism
Lemma 3.4.
For any Borel subalgebras and there exists a Cartan subalgebra .
Proof.
Using the Bruhat decomposition relative to , we write
There exists some and such that
We consider and . In particular, we see that
Applying to this inclusion gives us the desired result. □
Remark 3.5.
If we consider the set of pairs , this is actually isomorphic to . Forgetting corresponds to the map , and the fiber over the set
Definition 3.6.
Define the relative position of two Borel subalgebras and as follows. Choose a Cartan . The pair determines a choice of positive roots, and we also know that . We then have canonical identifications and . Therefore, we have
for a unique .
The reason this is well defined is because for any two , we have for some . This implies that in the diagram
we actually have , so they are sent to the same .
In practice, if we fix a pair , then given and , we first write , then apply
and then compute which Bruhat cell lies in.
Remark 3.7.
- (1)
If we consider the action of on , then the orbits are canonically indexed by .
- (2)
If we have , the relative position of with respect to is not determined by and . However, if , then the relative position is indeed .
Example 3.8.
If , then . The -action decomposes this into
and the diagonal corresponds to the identity element (so ), while the other element corresponds to the large open set. If we consider
then we can get either or depending on whether .
In general, suppose and
Suppose that . Applying to the second arrow, we see that
which tells us that and hence that .
Definition 3.9.
The Grothendieck-Springer resolution is the incidence variety
This variety admits two maps
which are both -equivariant. Over , the fibers are simply given by , so we have the tautological vector bundle over . Therefore, we have a morphism
over .
The fiber
which is the zero locus of the vector field on corresponding to is called the Springer fiber. Later, we will restrict to in the nilpotent cone. The first extreme case is if , in which case . On the other hand, if , then for some Cartan, so for any , and thus the Springer fiber is simply
which is a torsor for the Weyl group. In particular, is actually a -torsor over .
Proposition 3.10.
The map is proper.
Proof.
Consider the commutative diagram
and use the fact that the horizontal arrow is a closed immersion and the fact that proper morphisms are closed under composition. □
Now consider the morphism
Note that if we consider only, then the target must be because there is no preferred order of the eigenvalues (in the case, the others are similar). We now have the correspondence
| (1) |
Theorem 3.11 (Chevalley restriction theorem).
Fix a Cartan . Then there is the restriction of functions , which induces
This is an isomorphism.
Proof.
We begin with injectivity. If satisfies , then . This is dense, so .
Now let . Our goal is to find such that . Note that determines on , but we need to ensure that extends to all of (since the boundary has codimension ).
First, using the correspondence Equation 1, we have after we choose some . Because is -invariant, this in fact does not depend on the choice of . We now have a function on , which is -invariant because
Our goal now is to descend it to . Let . Note that is a -torsor and thus . In particular, we see that
Note that is -invariant, so it descends to . Moving to the analytic topology, we want to show that for all and any small ball , then is bounded. Note that is compact because is proper. Note that is bounded, so is bounded and thus extends to a holomorphic function on .
Finally, we need to check that . For any and any choice of , we see that
Example 3.12.
For , we can write
and . Transported to , these are simply the elementary symmetric polynomials in the eigenvalues.
Note also that the Chevalley restriction theorem gives a canonical isomorphism . Later, we will see that is a polynomial ring in variables, so is an affine space of dimension . The embedding gives the adjoint quotient map
Remark 3.13.
We can in fact choose homogeneous generators of degrees , and we will have and being the Coxeter number .
Proposition 3.14.
The diagram
commutes.
Proof.
It suffices to check that
because is affine. This is true because
essentially by definition. □
Proposition 3.15.
Let be a Borel, , and . Then for all , the restriction is constant.
Proof.
For all , we have
Remark 3.16.
If we use the Jordan decomposition, then in fact for any because .
3.2 Nilpotent cone.
Definition 3.17.
The set of nilpotent elements
is closed, -stable, and -stable and is called the nilpotent cone.
Recalling the Grothendieck-Springer resolution , we define . Now we can consider the diagram
We see here that
Note in fact that the condition of being nilpotent is equivalent to for all . This identifies as a -equivariant vector bundle over with fiber , and in particular that
Under this presentation, we see that .
Theorem 3.18.
There is a natural identification .
Proof.
Note that if , we have , so the cotangent space . Using the Killing form, we can identify , and under this identification, . Therefore, we have a canonical isomorphism . □
Later, we will see that is the moment map for the -action on induced by the -action on . Let be the functions which vanish at the origin.
Proposition 3.19.
if and only if for all .
Corollary 3.20.
The nilpotent cone is an irreducible variety of dimension .
Proof.
The map is proper and surjective, and is irreducible. We also see that . On the other hand, we know , and the morphism expresses and shows that . □
Proposition 3.21.
There are finitely many -orbits in .
Idea of proof.
This is easy to see for via the Jordan normal form. In the general case, we choose a faithful representation , which embeds nilpotent cones by preservation of Jordan decomposition. The following lemma finishes the proof using the fact that is semisimple (so it has a -stable complement in ). □
Lemma 3.22.
Let such that has an -stable complement . Let and be connected algebraic groups with Lie algebras and . For all -orbits in , any irreducible component of is a single -orbit.
Proof.
Let be any irreducible component. For any , we have
This implies that , so is open and dense in . This implies that for any , , so the two orbits coincide. □
Corollary 3.23.
The nilpotent cone contains an open dense orbit, which is exactly
Proof.
For all , we want to check that . This follows from the fact that
Example 3.24.
In the case of , we see that
See Figure 9 for a visualization of this variety. We can interpret as
and therefore anything in the open part is similar to and clearly has a one-dimensional kernel, while the origin has a two-dimensional kernel and thus its fiber is all of .
Now fix a . This gives an action of on . Choose a basis for . Define the subset
Proposition 3.25.
The set is a single -orbit consisting of regular nilpotent elements.
Proof.
Note that is open and dense in (here, must be open and dense in ), so must contain a regular element . The image of in is a single -orbit, so it remains to show that is a single -orbit. On the level of tangent spaces, we see that
and that this holds on all of . This tangent space has dimension
and therefore is open and dense. However, the fact that is unipotent and is affine implies that it is also closed, so . □
Proposition 3.26.
Every regular nilpotent element is contained in a single Borel.
In particular, this means that the morphism is an isomorphism over the regular locus.
Proof.
Fix a regular nilpotent element
where each . Because this consists entirely of elements from the simple root spaces, there exists such that . This is determined uniquely by the condition for every simple root . This implies that for all positive roots , and thus .
Note that this determines a one-parameter subgroup . The adjoint action is given by and lifts to an action of on . For all , we see that
and because is finite, the action is trivial. In particular, this implies that is a fixed point of the action of on induced by . This happens if and only if , and because is regular, we see that . The set of Borels satisfying this form a -torsor, but any containing must contain all simple root spaces , and there exists a unique Borel satisfying this condition. □
In particular, we see that is a resolution of singularities, called the Springer resolution.
Remark 3.27.
If , then is a singleton. On the other hand, if , then is a -torsor. In general if , then is finite. More generally, the map sends (after conjugating to be in ). In particular, the set can be identified with the set .
3.3 Jacobson-Morozov theorem.
The idea is that for every nonzero , we can find an -triple .
Theorem 3.28.
Let . Then there exist such that is semisimple, is nilpotent, , , and .
Note that in , the matrices
satisfy these relations. For a general , these give an embedding sending the standard basis to . Also, in the general case, is a transverse slice to , and if is regular, it is called the principal slice or Kostant slice and satisfies .
Corollary 3.29.
For all , there exists a rational group homomorphism sending to .
Proof of Theorem 3.28.
Recall the Killing form
The fact that this is -invariant means that
To simplify things, we will write and for .
Lemma 3.30.
Assume that and consists entirely of nilpotent elements. Then .
Proof.
We need to show that for any , the trace . However, both and are nilpotent, and they commute, so their product is also nilpotent. □
Our goal now is to reduce to the case where so that we can induct on the rank of . Assume that if the rank is strictly less than , then the theorem holds. For all , we consider the set . If , then there exists which is not nilpotent. Writing , we see that implies that both and , so . Because is semisimple and nonzero, is a proper Levi subalgebra and is hence reductive. Because is nilpotent, we have
which is semisimple and has strictly smaller rank than .
In the case where , we use the -invariance of the Killing form to obtain
and therefore that . Because the Killing form is nondegenerate, the two are actually equal. This implies that and thus that there exists such that . Decomposing into a semisimple and nilpotent part, we see that . Therefore, we can replace by .
Our final goal is to find . Consider the weight decomposition of . This decomposes
Here, note that and , and because , we see
Here, the calculation is
We want to find such that , or equivalently, show that . Because and , we see that the subspace is a solvable subalgebra in . By the theorem of Lie and Engel, there exists a basis of such that is upper-triangular and for any , is strictly upper-triangular. This implies that
for any because is strictly upper-triangular. This implies that , so there exists such that . Because , we can choose . □
Proposition 3.31.
All -triples containing are conjugate under .
For a fixed triple , we can consider the induced representation of from any representation of . Recall that representations of are classified by their highest weight, which is a positive integer . See Figure 10 for a visualization. For non-irreducible representations, we can draw a diagram with all of the irreducible components.
Specifying to , we see that . If we decompose
note that .
Remark 3.32.
Note that and equality holds if and only if all weights are even.
If is regular, then , and therefore is regular and , and therefore are all regular. In this case, we have a principal -triple. Note here that all weights are even. Also, note that
but is not necessarily contained in the strictly positive part. By definition, set
By definition, we have , and if there is no trivial part under , then is contained in the strictly positive part. Now lift to a unipotent subgroup . Note that may be disconnected, but is connected.
Lemma 3.33.
The affine space is a single -orbit.
Proof.
First, note that if , then because consists of weight spaces. In fact, they are actually equal because . By a standard argument, the orbit is open, but is unipotent, so it is also closed. □
Proof of Proposition 3.31.
Fix and . Note that , so . In particular, we see that because we also have . Therefore, , so there exists such that .
Now we have . Note that
and therefore . This implies that , but while is concentrated in nonnegative weights, so . Therefore, . □
Remark 3.34.
We have in fact also shown that if we fix and , then is unique.
We will now consider the structure of .
Proposition 3.35.
The unipotent subgroup is the unipotent radical of and is the maximal reductive subgroup of .
Example 3.36.
If is regular, then so are and . When we described , recall that consisted of all elements which are nonzero on all simple root spaces. If we let
thhen is actually unique is given by
while is lower-triangular with nonzero entries on the subdiagonal. In this case, we have . For all , we see that , and so under some choice of Borel containing , we see that
and therefore if , this is equivalent to for all simple roots , which is equivalent to .
Because is regular, we see that . Using the proposition, we see that , and therefore that . Here, we see that
Looking at the decomposition of with respect to this , then in fact
or namely all weights are even. There is no trivial representation appearing as a summand of because is discrete.
Proof of Proposition 3.35.
Our goal is to show that and that is trivial. We already know that because is an ideal in .
We will use the fact that the centralizer of a reductive subgroup in is always reductive. Because the image of in given by is reductive, so is . We also know that , and because is reductive, the intersection is trivial.
Finally, we need to prove that . For any , recall that
for some . Therefore, we see that and therefore (here, we use normality of ). □
3.4 Regular -triples.
Let and consider the Kostant slice . Recall that the map restricts to an isomorphism
Note that is an isomorphism of graded rings.
Theorem 3.37 (Chevalley-Shephard-Todd).
We have , where are homogeneous elements of degrees which are independent of the choice of generators. These are called the exponents of .
Theorem 3.38 (Kostant).
Let be a regular -triple. Recall that the decomposition
into irreducible -representations satisfies , where . Order such that . Then and .
Proof.
Consider the decomposition
For each , the lowest weight is . Because is the lowest weight space of , we can let the set of be a basis of . Define a map
We will show that is surjective everywhere, or namely that is a submersion. Because this morphism is -equivariant, it suffices to check when . Let . Then
Our goal is to prove that this is surjective. We compute
Now consider the -action on , which is given by
Note that the map is equivariant with respect to this action, where acts on by . Note that the limit contracts everything to the origin, so it was enough to check that we had a surjection there. Because the differential is surjective, is open and dense in , and therefore the restriction
is injective.
For a generator , let be given by
Because are algebraically independent, so are . We now use the Euler identity, which says that for all homogeneous , we have
Denote this vector field by . Applying this to the on , let denote the tangent vector at . Clearly, we have
Now let
Then we see that
The polynomial is a sum of monomials of the form , and we have the identity
for every monomial in .
We now want to show that each . Suppose there exists such that . Then for every , we must have . On the other hand, for all , we have while for . This implies that depend only on , so they cannot be algebraically independent, which is a contradiction.
Finally, we need to show that . We will follow Kostant’s proof.
Proposition 3.39.
Let be semisimple of rank and be the exponents of . Then
Assuming the proposition, we see that
so we must actually have an equality everywhere. □
Proposition 3.40.
The morphism
is an isomorphism.
2 Universal centralizers
4 Defining the universal centralizer
Definition 4.1.
Define the universal centralizer to be
Note that this has a morphism to given by forgetting and applying . This map is flat, and the general fiber is a maximal torus.
This universal centralizer admits a symplectic structure. To see this, we will realize it as a Hamiltonian reduction.
Definition 4.2.
A symplectic manifold is a smooth manifold of dimension equipped with a closed non-degenerate -form .
Note that induces an isomorphism .
Example 4.3.
The most basic example is the cotangent bundle
where are local coordinates on and are the corresponding covectors in the fiber.
Example 4.4.
A Kähler manifold has a symplectic form given by the imaginary part of the Kähler metric. For example, we can consider with the Fubini-Study metric.
We will consider the action of on . If there is an action of on , then there is an induced action of on given by
Exercise 4.5.
Check that this action preserves the symplectic form.
If we consider the infinitesimal action, then for every there is a vector field on . We then consider the -form
for some Hamiltonian function . The moment map is defined by the relation
Note that this is only defined up to a constant, but can be fixed by requiring that is a Poisson map. In particular, the formula is given by setting
Example 4.6.
Consider the left and right actions of on . The right action gives an isomorphism given by
We can now try to compute the moment maps. The moment map for the left action is given by
and the moment map for the right action is given by
To see this, note that
for the left action.
Proposition 4.7.
The universal centralizer is isomorphic to a Hamiltonian reduction of .
Here, we can consider the left and right actions of the unipotent . We then consider
We can now compose this to by using the Killing form to identify . Now consider the inclusion
and recall that is the sum of the simple root spaces. Dualizing, this is an injection
Now choose . This is invariant under the adjoint action of , and therefore
In general, we can consider a coadjoint orbit and the Hamiltonian reduction . In this case, we are considering
Proposition 4.8.
- (1)
;
- (2)
If is the Kostant slice, then .
Proof.
- (1)
We will prove that . Write the decomposition
with respect to the -triple . Now consider the ascending filtration
Let . We see that preserves the filtration, so has an induced filtration
We will show that the image of lies in . Choose any element . We can write where and . Then we see that
This implies that . Now we compute
and therefore is regular.
- (2)
We will not give a full proof of this. The idea is that the tangent map of the opposite version lands inside and is given by . The rest of the proof is to define a -action on both sides which contracts everything to a point and such that is equivariant. On the right, we define
where . Note that , so acts on it with non-negative weights and thus the action has positive weights. On the left, note that is generated by the negative root groups. Then has positive weight on it, so clearly both factors are contracted to the identity. The proof concludes with the following lemma. □
Lemma 4.9.
Let and be smooth irreducible affine varieties with contracting -actions. Let be -equivariant such that the tangent map at the fixed point is an isomorphism. Then is an isomorphism.
Theorem 4.10.
The composition is a bijection.
Proof.
Choose a set of homogeneous generators of . Then we have a diagram
We know that is an isomorphism. We need to prove that is injective. To do this, we will show that separate -orbits on . Suppose that have the same semisimple part. We claim that they lie in the same -orbit. This is because their nilpotent parts and must lie in , so they are in the same nilpotent orbit in . Because of this, we can see that . This happens because for any nonzero constant, is conjugate to . Therefore, we see that , and taking the limit gives us .
It now suffices to show that separates -orbits. Let and . We claim that there exists such that and . To see this, we start with such that and for all and the apply the averaging trick to obtain
We will now return to the problem of considering the moment map
Recall here that descends to and that we want to consider , which is called the bi-Whittaker reduction . Here, we have
Because , we can kill the factors and make the identification
4.1 Bruhat Decomposition.
Recall that and thus has a map to , which has the Bruhat decomposition
It is natural to ask what the preimages are.
Proposition 4.11.
- (1)
appears in the image of if and only if for some , where is the longest element and .
Denote the Levi subgroup associated to by . This can be obtained by integrating
Alternatively, choose a generic element , and then .
- (2)
Let . Then
Note that the second factor is an affine space and is a torus which may not be connected.
Example 4.12.
Let and make the standard choice of . Then let
Then the general fiber over is but the central fiber has two components given by
Away from the central fiber (or in general over the regular locus), we have
There is a unique simple root , and so we see that
which is the union of Kostant sections. Note that this holds generally for any semisimple .
In general, the complement of the Kostant sections is .
Remark 4.13.
If , then is open. Note that this contains as another open subset.
We will now begin the proof of the proposition. Let . To simplify the proof, consider a torsor of over given by
Now our Bruhat decomposition is with respect to , and we can get rid of the everywhere.
Up to the action of , we can always reduce for some representatives and . We need to determine for which there exists such that . This requires that , so let . Then must send to . Also, we see that sends to and thus we have
for all . This implies that and therefore . Because , we see that . In summary, if , then , so we have proved that if , then .
Now fix some and let for some . Write , where , , , and . We need to show that
Note here that , which we need to equal , is sent to an element of , , and is sent to , so we need . This implies that
In summary, we need to satisfy
Note that there are some “free variables” like and , so we can choose such that , and thus the choices of form a torsor for . In particular, we see that . Taking the quotient by conjugation of , we see that the choices can be identified with . Therefore, we have proved that
If we denote the previously defined -torsor by , we see that this preimage is actually an isotropic submanifold in .
Note that there is a canonical open symplectic embedding
which is given by the formula
If we consider the composition (another integrable system), then the map
is proper.
4.2 Wonderful compactification.
Note that the horizontal integrable system has target
By the Peter-Weyl theorem, we see that this is actually
The monoid in question is a free monoid generated by the fundamental weights if is simply connected. Our goal is to produce a partial log compactification
where is a complete toric variety and the map
is finite. Roughly, it will be something like
For a reference, read Brion-Kumar’s book on Frobenius splitting methods.
More generally, we can consider equivariant embeddings of , where we are looking for a normal variety with an action of and an open equivariant embedding .
- (1)
When is a torus, then these are just toric varieties;
- (2)
If , then there is the wonderful compactification .
The wonderful compactification satisfies the following properties:
- (1)
is smooth and complete;
- (2)
The boundary is a union of smooth divisors with simple normal crossings;
- (3)
The closures of the smaller -orbits are given by where is nonempty.
In general, equivariant embeddings will combine the features of toric varieties and the wonderful compactification. The idea is that , while .
Let and . Choose a regular dominant weight . Note that there is a map
which induces a map
Lemma 4.14.
For all regular dominant weights , satisfies the following:
- (1)
The -eigenspace of weight is a -invariant line. All the other weights are strictly smaller than in the dominance order;
- (2)
For all positive roots , we have and the morphism sending to is a closed embedding;
- (3)
Dually, and the morphism to sending to is a closed embedding;
- (4)
The action of (resp ) on factors through a faithful action of (resp ).
Note that acts on by a character . If we consider the other weights
all , so the action factors through . The same applies to the Lie algebra action. On the other hand, if fixes every line in , then is in every Borel, so . This implies that the action of is faithful, so there is an embedding .
Now fix as above and let . Then consider the -module and let be the identity matrix.
Lemma 4.15.
The orbit is isomorphic to
Proof.
Suppose that . This is equivalent to for a fixed scalar for all . This happens if and only if . Therefore, the stabilizer of is given by . □
Now let and . We need to prove that is smooth and that the boundary is snc. To study , one key ingredient is to analyze the closure of .
Lemma 4.16.
Let be a reductive group and be a -variety. Assume that is such that is open in and that the isotropy group contains . Set . Then:
- (1)
All -orbits in intersect ;
- (2)
contains finitely many -orbits;
- (3)
is complete if and only if is complete.
Remark 4.17.
If is not adjoint, then we need to consider toroidal compactifications, which depend on a choice of -equivariant toric compactification of .
Let be a basis of -eigenvectors of and be the dual basis. Then . There are some summands like and where is the highest weight. The vector is a -eigenvector. Let . Now let
and . Then is -stable because applying the action to
does not kill the leading term.
Lemma 4.18.
- (1)
The inclusion extends to an isomorphism .
- (2)
The action of on induces an action of , and
is a smooth complete toric variety whose fan consists of the Weyl chambers and their faces.
Proof.
- (1)
Because fixes , we can compute that
where . Then the embedding
is given by
It is easy to see that .
- (2)
Note that
By definition, the fan of is the negative Weyl chamber because the functions are for , but when we apply , this corresponds to moving to a different chamber. Gluing these together, will contain the toric variety with the complete fan given by all Weyl chambers and their faces. Clearly, these two things must be equal. □
Proposition 4.19.
The map
is an -equivariant isomorphism.
Proof.
The restriction of to identifies the source with the open Bruhat cell in . To show that is an isomorphism, we need to construct an -equivariant morphism and then show that . Note that , but it is easy to see that
so clearly we see that .
We will now construct . Regard as a space of rational maps . By definition, consists exactly of those rational maps defined at and send this to . For all , it gives a regular self-map of preserving . By the Bruhat decomposition, we see that
so we can consider the action of each Bruhat cell on and see that
There is a map and we want to factor it through a rational map . Because the image of is already closed, the map extends to some partial compactification. This gives a map which is clearly equivariant. The map is defined dually. □
Theorem 4.20.
- (1)
is covered by the -translates of (and in particular is smooth);
- (2)
The boundary is the union of smooth prime divisors with normal crossings;
- (3)
For all subset , the stratum
is the closure of a unique -orbit . There is a unique closed orbit .
4.3 Log compactification of the universal centralizer.
Recall that . Here, we will need to be adjoint, but the first step will be to replace the cotangent bundle with the log cotangent bundle of the wonderful compactification. To understand , we will consider . This is a toric variety whose fan is the single cone given by the negative Weyl chamber, so the functions are simply for the simple roots . The -orbits are indexed by with standard points
Everything in can be translated to by the Weyl action. Every -orbit in intersects , so they are also indexed by -orbits and will be called . These are stabilized by
where is the parabolic subgroup associated to . In particular, there is a fiber sequence
The orbit closures are given by
and the fibers of are the wonderful compactification of . In fact, there is an isomorphism
and from we see that the boundary is normal crossings.
Now consider the log tangent bundle where local sections are vector fields tangent to the stratification by . Explicitly, if we have local coordinates such that is defined by , then the local sections are generated by . The log cotangent bundle is the dual of this, and its local sections are generated by
Now inherits an action of and in particular is log homogeneous for the action of . In particular, there is
We will denote the kernel by , which is called the bundle of isotropy subalgebras. We can show that is the restriction of the tautological bundle of
where . It is in fact a Lagrangian subbundle of with respect to on , where is the Killing form. Now note that
Lemma 4.21.
There is an identification
Now consider the compactified moment map
where we view .
Lemma 4.22.
- (1)
The image of is ;
- (2)
The variety is normal;
- (3)
has connected fibers.
Sketch of proof.
For the first part, is clearly proper, so the image is closed. Note that the image contains because the uncompactified moment map is contains this locus. The image of must be its closure, so it must be .
For the second part, is a complete intersection defined by
where are generators of . Now is open and smooth, and the complement has codimension at least because the complement of has codimension at least , so is smooth in codimension and is therefore normal.
For the third part, we know that is proper and its image is normal. We only need to show that a general fiber is connected, in which case all fibers are connected by Stein factorization. For this, we will use the general fact that is a connected toric variety. □
Now define a fiberwise partial compactification of by
Theorem 4.23 (Balibanu).
is smooth and the boundary is a divisor with simple normal crossings. Moreover, has a natural log symplectic structure extending the symplectic structure on , which is a closed nondegenerate -form with logarithmic poles along the boundary.
Sketch of proof.
Note that
Now consider . Here, acts freely on , so we can use a logarithmic version of Hamiltonian reduction. Here, we need to be a divisor, and it certainly avoids , and is connected and smooth, and hence irreducible. This also proves that is open and dense. □
Theorem 4.24 (Jin, to appear).
Let . This compactifies to
where is the complete toric variety associated to the fan whose rays are for and for and whose dual polytope is homeomorphic to a cube. For each maximal cone in the fan, there is the identification
where is the closure of in . Finally, we have