Patrick Lei

All notes

Representation theory

Lectures by Ivan Mirković, UMass, Fall 2018.

Math 797RT
Representation Theory
Notes by Patrick Lei, May 2020
Lectures by Ivan Mirkovic, Fall 2018

University of Massachusetts Amherst

Disclaimer

These notes are a May 2020 transcription of handwritten notes taken during lecture in Fall 2018. Any errors are mine and not the instructor’s. In addition, my notes are picture-free (but will include commutative diagrams) and are a mix of my mathematical style (omit lengthy computations, use category theory) and that of the instructor. If you find any errors, please contact me at plei@umass.edu.

Additional Disclaimers

  1. 1.

    Ivan was in China for a week, and I missed one of the two makeup lectures. If there are any gaps in the material, this is probably a big reason why.

  2. 2.

    I did not take notes for the final lecture. I believe only the classification of semisimple Lie algebras was given.

  3. 3.

    The course has been redesigned since I took it, so these notes may not be useful. In fact, I’m told that there is no set syllabus for this course, so it may continue to change in the future.

  4. 4.

    Ivan has notes at https://people.math.umass.edu/~mirkovic/C.Graduate/TopicsCourses/7.RepresentationTheory/8.pdf, but his notes are much more extensive and in depth than the lectures. My notes are a more faithful representation of what actually occurred during lecture.

  5. 5.

    Ivan’s lectures assumed knowledge far beyond the stated prerequisites of the course, which were just graduate algebra. It may be clear that I did not understand what was going on half the time.

1 Groups

1.1 Overview

We begin with an overview of various aspects of representation theory.

  1. 1.

    First, the principle that any subject should be organized by its symmetries (harmonic analysis). We want to spot symmetries in a problem, and this is related to physics and QFT.

  2. 2.

    We want to consider symmetries related to subtle actions of space. For example, for a scheme X and sufficiently nice group scheme G, we can consider the quotient stack X/G. The projection from X is a principle G-bundle, and in fact, we can do this for X=Spec⁡k, which is a point. Also, we will need the idea of the fiber product X×ZY later.

  3. 3.

    Groups are used to describe objects that are locally trivial but not globally (for example cohomology classes).

  4. 4.

    Groups can be recovered from their categories of representations via Tannaka-Krein duality by considering automorphisms of the forgetful functor 𝖱𝖾𝗉⁡G→𝖵𝖾𝖼.

  5. 5.

    The most important class of groups is the compact Lie groups.

Recall the notion of a group action on a set X. If we view X∈𝒞 for some category 𝒞, then we want our group G to act by automorphisms in 𝒞. Therefore we can define topological group actions on topological spaces and actions of groups on vector spaces.

Lemma 1.1.

For a group G and vector space V, then an action of G on V is equivalent to a morphism π:G→G⁢L⁢(V).

Example 1.2.

If G acts on a set X, then the space 𝒪⁡(X)=Hom⁡(X,k) of functions on X is a representation of G.

We can generalize the above example to where X has additional structure, where we can consider functions in various categories. For example, if we consider G=S⁢L2⁢(ℝ) acting on the upper half plane ℋ, then we can consider the space 𝒪ℋ⁢(ℋ) of holomorphic functions on ℋ.

1.2 Category of Representations

Lemma 1.3.
  1. 1.

    Representations of G on vector spaces over a field k form a category 𝖱𝖾𝗉G⁡(k)=𝖱𝖾𝗉⁡(G). Note that morphisms are

  2. 2.

    𝖱𝖾𝗉G⁡(k) is an abelian category.

    Recall that an abelian category is a category with the following:

    • •

      A zero object 0. Here, just take the zero representation.

    • •

      A direct sum (coproduct) operation ⊕. Here, take g⁡(u⊕v)=g⁢u⊕g⁢v.

    • •

      Subobjects, which in this case are G-invariant subspaces.

    • •

      Quotients by subobjects. This is obvious.

    • •

      𝖱𝖾𝗉⁡G has all (co)kernels and (co)images, and images and coimages are isomorphic.

    Some examples of abelian categories are categories of modules over a ring, but in fact 𝖱𝖾𝗉⁡G is equivalent to k⁡[G]⁢-⁢𝖬𝗈𝖽.

  3. 3.

    𝖱𝖾𝗉⁡G is a monoidal category. For module categories, this is only true in general for commutative rings. For two representations U,V of G, we can define an action on U⊗V by g⁡(u⊗v)=g⁢u⊗g⁢v.

    Recall that a monoidal structure on a category 𝒞 is an associative and unital functor 𝒞×𝒞→𝒞. The unit in 𝖱𝖾𝗉⁡G is k with the trivial action.

  4. 4.

    𝖱𝖾𝗉⁡(G) is a closed monoidal category. This means that there is an internal Hom functor

    that is the right adjoint to the tensor product.

Examples 1.4.

We give some examples of closed monoidal categories.

  1. 1.

    In the category of sets, we see that

    by the maps

  2. 2.

    In the category of vector spaces, we again note that

    by the same currying morphism.

We can check that the internal Hom in the category of representations is simply the vector space Hom. To see that this is a representation, define the action (g⁢φ)⁢(u)=g⁢φ⁢(g−1⁢u).

Recall that a right adjoint G:B→A to a functor F:A→B satisfies

as sets in a manner that is natural in a,b. Being a closed monoidal category means that −⊗b is a left adjoint to Hom¯⁢(b,−).

1.3 Representations of Finite Groups

First we consider invariants of V∈𝖱𝖾𝗉⁡G, which correspond to morphisms from the trivial representation.

Lemma 1.5.

We have the two identities HomG⁡(k,V)=Vg and Homk⁡(U,V)G=HomG⁡(U,V).

Definition 1.6.

A representation V∈𝖱𝖾𝗉⁡G is irreducible if V≠0 and has no nontrivial subrepresentations.

Then define IrrG to be the set of isomorphism classes of irreducible representations of G. We will use the following strategy to understand 𝖱𝖾𝗉⁡G:

  1. 1.

    Find all irreducible representations;

  2. 2.

    Study each irreducible representation;

  3. 3.

    For any interesting representation, decompose it into irreducibles.

For a finite group G, consider the set of functions 𝒪⁡(G)=Hom⁡(G,k). Then we have two subrepresentations 𝒪+⁢(G)≔𝒪⁢(G)G and 𝒪−⁢(G)={f:G→k∣∑g∈Gf⁡(g)=0}.

Lemma 1.7.

If the characteristic of k does not divide |G|, then 𝒪⁡(G)≃𝒪+⁢(G)⊕𝒪−⁢(G).

Proof.

The dimensions match up, so the only constant satisfying c⁢|G|=0 is c=0 by assumption. □

Example 1.8.

If G=S2={e,a}, then 𝒪±⁢(S2) are irreducible, but they are actually the same subrepresentation of 𝒪⁡(S2) over 𝔽2.

Definition 1.9.

A representation V is semisimple if it is a direct sum of irreducibles.

Theorem 1.10 (Schur Lemma).

Let U,V be two irreducible finite-dimensional representations of G.

  1. 1.

    If 0≠α∈HomG⁡(U,V), then α is an isomorphism;

  2. 2.

    EndG⁡(U) is a finite-dimensional division algebra;

  3. 3.

    If k=k¯, then EndG⁡(U)=k.

Proof.
  1. 1.

    Note that ker⁡α is a subrepresentation and U is irreducible, so ker⁡α=0. Similarly, the image is nonzero and is a subrepresentation of V, so it must be all of V.

  2. 2.

    Any nonzero endomorphism must be an isomorphism.

  3. 3.

    If α∈EndG⁡(U) is nonzero, it has an eigenvalue λ with eigenvector v. Then ker⁡(α−λ) is a nonzero subrepresentation, so it must be everything.

□

Corollary 1.11.
  1. 1.

    V is semisimple if and only if it can be written as a direct sum

  2. 2.

    Let U be irreducible. Then HomG⁡(U,V)=HomG⁡(U,U)⊗MU.

Proof.
  1. 1.

    Note that if V is a direct sum of irreducibles, then we have

  2. 2.

    Hom is an additive functor, so just use Schur’s Lemma.

□

Corollary 1.12.

If k is algebraically closed, then HomG⁡(U,V)=MU for V semisimple and U irreducible.

Theorem 1.13.

Let G be abelian and k be algebraically closed. Then all finite-dimensional irreducible representations of G are one-dimensional.

Proof.

Let V∈Irr⁡(G) and g∈G. Then g has an eigenvalue λ with eigenvector v. Therefore, ker⁡(g−λ)=V is a subrepresentation because G is abelian, so all subspaces are invariant and thus dimV=1. □

This tells us that Irr⁡(G)=Hom⁡(G,k∗)≕Ĝ. This is known as the group of characters1 of G.

Lemma 1.14.

Subrepresentations of semisimple representations are semisimple.

Proof.

Let V′⊆V be a subrepresentation of an irreducible representation. Then V′ has an irreducible subrepresentation V1′ and thus 0≠Hom⁡(V1′,V)=⨁i∈IHom⁡(V1′,Vi), so there exists i such that Hom⁡(V1′,Vi)≠0, so V1′⊆V. Then V′/V1′⊆V/V1′=⨁j≠iVj, so we have an exact sequence

Therefore V=Vi⊕C, so V′=(Vi⊕C)∩V′. □

Theorem 1.15.
  1. 1.

    Sums of semisimple representations are semisimple.

  2. 2.

    V is semisimple iff and only if for every subrepresentation V′⊆V, there is another subrepresentation V′′ such that V=V′⊕V′′.

Proof.

We will only prove the second part (the first part is obvious). Note that if the condition holds, then choose an irreducible subrepresentation V′, apply the condition, and use induction on the dimension.

Now suppose V is semisimple. Then apply Lemma 1.14 to V′=⨁HomG⁡(U,V′)⊗U⊆⨁HomG⁡(U,V)⊗U and we see that there is a complement. □

Theorem 1.16.

Let G be a final group and k=ℂ. Then every V∈Irrfd⁡(G) has a G-invariant Hermitian inner product and is semisimple.

Proof.

To construct an invariant inner product (−,−), just take any Hermitian inner product ⟨−,−⟩ and define

If we have an inner product, let W⊆V be a subrepresentation. Then W⟂ is also a subrepresentation, and clearly W⊕W⟂=V, so V is semisimple by Theorem 1.15. □

Theorem 1.17 (Maschke).

If char⁡k∤|G|, then all finite-dimensional representatons of G are semisimple.

We will now study in detail the characters Hom⁡(G,k∗) of G.

Lemma 1.18.
  1. 1.

    Hom⁡(G,k∗) is a group with operation the multiplication of functions.

  2. 2.

    Hom⁡(G,k∗)=Irr1⁢d⁡(G) because G⁢L1⁢(k)=k∗.

  3. 3.

    Hom⁡(G,k∗)=Hom⁡(Gab,k∗).

If A is abelian, then we call the group of characters Â≔Hom⁡(A,k∗) the Pontryagin dual of A.

Examples 1.19.
  1. 1.

    Note that ℤ̂=k∗ because 𝖠𝖻=ℤ⁢-⁢𝖬𝗈𝖽.

  2. 2.

    We can see that ℤ/n⁢ℤ̂=μn⁢(k).

  3. 3.

    If k is algebraically closed, and char⁡k∤n, then ℤ/n⁢ℤ̂≃ℤ/n⁢ℤ.

Note that Hom⁡(G,k∗)×Hom⁡(H,k∗)≅Hom⁡(G×H,k∗) by the categorical properties of the product. Consider the functor 𝖠𝖻𝗈𝗉→𝖠𝖻 given by A↦Â=Hom⁡(A,k∗). Note that there is a natural2 morphism ι:A→Â̂ given by ι⁡(g)⁢(χ)=χ⁡(g) for g∈A.

Lemma 1.20 (Pontryagin Duality).

If k is an algebraically closed field of characteristic 0, then −̂:(𝖠𝖻𝖿)𝗈𝗉→𝖠𝖻𝖿 is an equivalence of categories. In fact, the natural transformation ι defined above is an isomorphism.

Proof.

This is obviously true for finite cyclic groups, and then use the structure theorem for finite abelian groups and the properties of the product. □

Remark 1.21.

If char⁡k>0, then the result extends to finite abelian group schemes. If we take locally compact abelian groups and Hom⁡(A,S1), then this result still holds.

Now we will study representations of products of groups. Consider the box product

with the operation the tensor product with G acting from the left and H acting from the right.

Lemma 1.22.

Let W∈Irr⁡(G×H). Then there exist V∈Irr⁡(G) and U∈Irr⁡(H) such that W=V⊠U.

Proof.

Let V∈𝖱𝖾𝗉⁡G. Then consider the morphism e⁢v:V⊗HomG⁡(V,W)→W. We claim this is a representation of H. To see this, define (h⁢φ)⁢(v)=(1,h)⁢(φ⁢v), which commutes with the action of G on W. Clearly, evaluation is surjective, so we can consider ker⁡(e⁢v)=V⊗K′ for some representation k′ of H. Therefore we can write

for some H′∈Irr⁡(H) (if H is reducible, then so is the entire representation). □

Lemma 1.23.

If V∈Irr⁡G and U∈Irr⁡H, then V⊠U∈Irr⁡G×H.

Proof.

Consider an irreducible subrepresentation V′⊠U′. Then

so V′=V and U′=U. □

Theorem 1.24.

The box product gives an isomorphism Irr⁡G×Irr⁡H→⊠Irr⁡G×H.

Now let V be a representation of G and consider the matrix coefficient map

This is equivariant with respect to G2, where (g,h)⁢f⁢(x)=f⁡(h−1⁢x⁢g). Also, the matrix coefficient map cV:V⊗V∗→𝒪⁡(G) is an embedding because V,V∗ are both irreducible.

Corollary 1.25.

Any finite group G has finitely many irreducible representations.

Theorem 1.26.

Let k be an algebraically closed field of characteristic 0. Then

as a representation of G2

Proof.

We compute that as representations of 1×G,

by Frobenius reciprocity. □

It is not hard to see that an explicit isomorphism can be given by cV, which is the same thing as evaluation.

1.4 Characters

From now on, we will always work over the complex numbers. It turns out that each irreducible representation of G gives us one conjugation-invariant function G→ℂ, the trace. Denote the character of a representation V by χV⁢(g)≔TrV⁡g.

Lemma 1.27.
  1. 1.

    χU⊕V=χU+χV;

  2. 2.

    χU⊗V=χU⋅χV.

These results follow from linear algebra of direct sums and tensor products.

2 Representation Theory of the Symmetric Group

Now we will construct irreducible representations of the symmetric group. Because irreducible representations correspond to conjugacy classes, there are bijections between the following:

  1. 1.

    Irreducible representations of Sn;

  2. 2.

    Conjugacy classes of Sn;

  3. 3.

    Partitions λ of n;

  4. 4.

    Young diagrams with n squares;

  5. 5.

    Conjugacy classes of nilpotent matrices of size n;

  6. 6.

    Finite (length n) subschemes of 𝔸2.

We want to construct an irreducible representation for each partition λ. For a partition λ=(λ1,…,λm), write Sλ=Sλ1×⋯×Sλm. We will write Mλ=CoindSλSn⁡τλ, where τ is the trivial representation. Similarly, define Nλ=CoindSλSn⁡σλ, where σ is the sign representation.

Theorem 2.1.

If C,R are conjugate partitions, then there exists a unique irreducible representation πC that is common to MR and NC.

Now, we will consider the dominance order on the set Πn of partitions of n. We say that λ≥μ if λ1≥μ1, λ1+λ2≥μ1+μ2, and so on.

We may also consider a more geometric notion. Let 𝒩n be the cone of nilpotent n×n matrices. Then every nilpotent matrix has a Jordan canonical form that corresponds to a shift operator eλ (shift left on a Young diagram λ). Then 𝒩n is a stratified space with strata given by orbits 𝒪λ corresponding to partitions λ. Then we say that λ≥μ if 𝒪¯λ⊇𝒪μ.

Lemma 2.2.

The dominance and closure orders coincide.

2.1 Geometry of G⁢Ln

The correspondence between irreducible representations of Sn and orbits of nilpotent matrices under conjugation by G⁢Ln, due to Springer, is the origin of geometric representation theory. This also extends to other reductive algebraic groups. For this, we will need to do some geometry.

First recall the Grassmannian G⁢rp⁢(V) which parameterizes dimension p subspaces of V. Then we can consider generalized flag varieties G⁢r1≤p1≤⋯≤pk≤n, which parameterizes flags of subspaces with dimension p1,…,pn. Finally, we have the flag variety

Lemma 2.3.

These spaces are all homogeneous spaces of G⁢Ln which are smooth projective varieties.

Now recall that the nilpotent cone 𝒩n=⨆λ∈Πn𝒪λ is a disjoint union of smooth pieces. By definition of the dominance order, we know that

However, note that 𝒩n itself is a singular space, so we should be able to extract information from the singularities.

Definition 2.4.

Let X be a singular variety. Then a resolution of X is a proper birational map π:X~→X from a smooth variety X~.

Example 2.5.

We can resolve the variety (x⁢y=0)⊆ℂ2 by taking the disjoint union of two lines.

If we consider 𝒩2, then this is just a quadric cone in 𝔸3, so we can resolve by blowing up the origin to get 𝒩~2={(e,L)∈𝒩2×ℙ1∣e∈L}.

Lemma 2.6.

ℱ=G⁢Ln/B0, where B0⊆G⁢Ln is the subgroup of upper-triangular matrices, also called the standard Borel. In addition, ℱ has a natural identification with the space of Borel subgroups of G⁢Ln.

Remark 2.7.

Note that G⁢Ln is a nonabelian reductive group. G has a maximal Borel B0, and we call the subgroup of B0 with 1s on the diagonal N0. Finally, we write T0=B0/N0, and this is called the standard Cartan, or the maximal torus, because T0≃𝔾mn.

2.2 Springer Theory for G⁢Ln

We want to resolve 𝒩n in general. To do this, we will consider 𝔤=Mn⁢(ℂ)=𝔤⁢𝔩n. Then define

so we have a natural morphism π:𝔤~→𝔤 that forgets the flag. Similarly, define

Again, there is a map π:𝒩~→𝒩. Write ℱx=π−1⁢x.

Lemma 2.8.

Let s∈𝔤r⁢s be a regular semisimple matrix. In other words, s has n distinct eigenvalues. Then |ℱs|=n! and ℱs is a torsor for Sn.

Here are some examples of torsors:

  1. 1.

    The set of bases of ℂn is a torsor for G⁢Ln.

  2. 2.

    The Möbius strip with the zero section removed is a torsor for ℝ∗.

  3. 3.

    Rank n vector bundles over a base space B are the same thing as G⁢Ln-torsors over B.

If G acts on X and Y, we can create a new G-set X×GY≔(X×Y)/Δ⁢G, where the diagonal action of G is g⁡(x,y)=(x⁢g−1,g⁢y). We can see that X×GY is a Y-bundle over X/G, so if X is a G-torsor, then X×GY≃Y.

Note that the set of regular semisimple matrices is a Zariski-open subset of 𝔤. We now consider the special fibers:

  1. 1.

    Above x=0, the fiber is simply all of ℱ.

  2. 2.

    If x corresponds to the Young diagram ​, then ℱx is a point.

  3. 3.

    For n=3, the fiber above the Young diagram ​ is two copies of ℙ1 intersecting at a point.

Lemma 2.9.

As spaces, 𝔤~=G×B𝔟. This means that 𝔤~ is a vector bundle over G/B=ℱ. Because 𝔟 is a representation of B, H0⁢(G/B,𝔤~) is a representation of G.

Remark 2.10.

These Associated Bundle constructions allow us to construct G-equivariant vector bundles from representations of B. If we consider the character (m,n) that sends (ℂ∗)2∋(a,b)↦am⁢bn as a representation ℂm,n of B, then we can define the vector bundle Vm,n=G×Bℂm,n. Then we can consider the representation Lm,n=H0⁢(G/B,Vm,n). Then each Lm,n is either zero or irreducible or zero and all irreducible representations of G are of this type. This is the Borel-Weil theorem.

Example 2.11.

For G=G⁢L2, then it is easy to see that G/B=ℙ1.

Lemma 2.12.

As spaces, 𝒩~=G×B𝔫=T∗⁢ℱ.

Lemma 2.13.

If G acts on X, then any G-equivariant map X→G/A identifies X with an associated bundle G×AXe⁢A.

Note that there is a natural symplectic structure on T∗⁢ℱ and that 𝒩~→𝒩 is a resolution. Here, over a regular nilpotent matrix, the map is an isomorphism and the map is proper because fibers ℱ is projective.

We will call the partition (n−1,1) subregular and the partition 2,1n−2 minimal. In general, these classes exist for all semisimple Lie algebras and we can perform the Springer construction for all of them. We now guess that dimπλ is the number of irreducible components of ℱλ.

Remark 2.14.

Semisimple Lie algebras are the same thing as Dynkin diagrams, which are displayed below:

Theorem 2.15.

All irreducible components of Springer fibers are equidimensional.

We now guess that πλ=ℂ⁡[Irr⁡(ℱλ)]. However, it is not clear what the action Sn is. For this, we will need a continuity argument. For this, we will need sheaf cohomology.

Theorem 2.16.

ℱλ is paved. This means that it is a disjoint union of affine spaces.

Being paved means that it is easy to compute the cohomology with coefficients in ℤ.

2.3 Sheaves

A presheaf on a space X valued in a category 𝒞 is a presheaf Xop→𝒞. A presheaf ℱ is a sheaf if for an open set U and a cover Ui with sections fi∈ℱ|Ui such that fi|Ui∩Uj=fj|Ui∩Uj, then there exists a unique section f∈ℱ|U such that f|Ui=fi.

Let f:X→Y be a morphism of spaces. Then there is a pushforward

and a pullback

There operations define functoriality for categories of sheaves.

If 𝒞 is an abelian category, then so is S⁢h⁢(X,𝒞), so we may consider complexes of sheaves. Then we may consider the derived category of complexes. In this setting, we have derived versions of f∗,f∗ as well as the compactly supported Rf! and exceptional inverse image functor Rf!.

Given a sheaf ℱ and a point x, the stalk ℱx is defined as the colimit

Alternatively, given the inclusion ι:x→X, the stalk is simply ι∗⁢ℱ. Then we can define the support of a sheaf as the closure of the set of points with nonzero stalks.

Remark 2.17.

The derived category is not an abelian category, but it is an triangulated category. This means we have exact triangles

There is a Verdier duality1 functor 𝔻X:D⁡(X)→D⁢(X)op such that 𝔻f∗𝔻=f! and 𝔻f∗𝔻=f!.

Example 2.18.

If a:X→pt, then a∗⁢ℱ=ℱx and a!k=ωX, the dualizing sheaf.

Under base change, we have the standard push-pull trick, and if f is proper, then f!=f∗. If f is a smooth morphism, then f!=f∗[dy−dx]⊗orf, where ℴ⁢𝓇 is the orientation sheaf.

Note that cohomology is simply the derived functor of global sections, so we will treat everything as a derived functor.

Definition 2.19.

A local system on X is a locally constant sheaf, and a sheaf ℱ is constructible if there exists a stratification X=⨆Si such that ℱ|Si is a local system.

Remark 2.20.

All of the derived category and functoriality constructions apply in the case of constructible sheaves.

2.4 Springer Fibers Again

Returning to our representation theory, we want to produce a∗a!ℂ from our setup. Therefore, we will push the constant sheaf ℂ forward from 𝔤~ to 𝔤. The resulting complex of sheaves is called the Grothendieck sheaf.

Remark 2.21.

Note that Sn is embedded in G⁢Ln and is embedded in the normalizer of the standard Cartan. In addition, Sn≅NG⁢(T)/T.

If π:𝔤~→𝔤 is the projection that forgets the flag, denote the Grothendieck sheaf by 𝒢≔π∗⁢k. Then the stalk 𝒢x is the same as the cohomology H∗⁢(ℱx). We know that Sn acts on 𝒢rs=𝒢|𝔤rs, so we need to extend this action.

Proposition 2.22.

𝒢 is obtained from its restriction to the regular semisimple matrices by a procedure called “intersection cohomology extension,” and this works because π is “small,” which means that the fibers increase slowly in some sense.

Definition 2.23.

A morphism π:Y→X is semi-small if dx=dimπ−1⁢x is generically zero and the closed set where dx≥k has codimension at least 2⁢k. The morphism π is small if it is semi-small and the set of points with dx>0 has codimension at least 3.

Theorem 2.24.

The morphism π:𝔤~→𝔤 is small.

This means that Sn acts on the entire Grothendieck sheaf 𝒢, so in particular, it acts on the stalks. This tells us that each stalk is a representation of Sn.

Lemma 2.25.

There is an isomprhism ℂ⁡[Sn]≃End⁡(𝒢)=Htop⁢(m⁢f⁢g~×𝔤𝔤~).

We conclude from our geometric construction that

  1. 1.

    Interesting objects have geometric realizations.

  2. 2.

    Geometric methods then give construction of irreducible representations.

  3. 3.

    This allows us to study Irr more efficiently.

3 Lie Algebras

3.1 Basic Notions

First, we will define manifolds over k=ℝ,ℂ. These are topological spaces M locally isomorphic to kn with transition functions that are (Ck, analytic, holomphic). An example of a manifold is a Riemann surface. Vector fields are simply derivations of 𝒪M and the tangent space at a is simply the derivations at that point.

Example 3.1.

If M=kn, then the space of vector fields is simply V⁡(M)=∑𝒪M⁢(M)⁢∂∂xi and the tangent spaces are Ta⁢M=⨁k⁢∂∂xi|a.

The tangent sheaf is locally free, so it is the sheaf of sections of the tangent bundle. If 𝒪M,x is the local ring at the point x and Mx is the maximal ideal, the cotangent space is isomorphic to Mx/Mx2.

Definition 3.2.

A Lie Group is a group object in the category of manifolds.

Example 3.3.

The group G⁢Ln is a Lie group, given by the equation det⁡g⋅z−1=0 in ambient space Mn⁢(k)×𝔸z1. Similarly, the group S⁢Ln is given by det⁡g=1, and the group S⁢p⁢(V) is the set of invertible matrices that preserve the symplectic form. Finally, the group O⁡(V) is the group of matrices that preserve a nondegenerate symmetric bilinear form.

Remark 3.4.

A symplectic form is a nondegenerate alternating bilinear form.

Definition 3.5.

A Lie algebra over k is a vector space 𝔤 together with an anticommutative product [−,−]:𝔤⊗𝔤→𝔤 that satisfies the Jacobi identity.

Example 3.6.

Any associative algebra can be made into a Lie algebra with the commutator as Lie bracket. Similarly, if A is an associative algebra, then Der⁡A is a Lie algebra with the commutator as Lie bracket.

Example 3.7.

Let G be a Lie group. Then the space Te⁢G is a Lie algebra and is the same as the left-invariant vector fields. We will denote this Lie algebra by Lie⁡G.

  1. 1.

    It is easy to see that Lie⁡G⁢Ln=𝔤⁢𝔩n=Mn.

  2. 2.

    The Lie algebra 𝔰⁢𝔩n of S⁢Ln is simply the set of trace zero matrices.

  3. 3.

    The Lie algebra 𝔰⁢𝔭2⁢n is the set of matrices where ω⁡(x⁢u,v)+ω⁡(u,x⁢v)=0.

  4. 4.

    The orthogonal Lie algebra 𝔬n is the set of matrices x where x+xT=0.

Theorem 3.8.

There is an exponential map exp:𝔤→G such that

  1. 1.

    exp⁡(0)=e;

  2. 2.

    d0exp:𝔤→𝔤 is the identity. In particular, exp is an isomorphism near 0 and e;

  3. 3.

    dds|s=0⁢exp⁡(s⁢x)⁢exp⁡(s⁢y)⁢exp⁡(s⁢x)−1⁢exp⁡(s⁢y)−1=[x,y]. This tells us that the Lie algebra commutator is an infinitesimal version of the group commutator.

Theorem 3.9.

We will see that 𝔤 controls Ge/Z⁡(Ge).

  1. 1.

    Any morphism π:G→G′ of Lie groups is determined on Ge by deπ:𝔤→𝔤′.

  2. 2.

    If τ:𝔤→𝔤′ is a map of Lie algebras and π1⁢(G)=0, then τ=deπ for a unique map of groups π:G→G′.

The universal cover of any connected Lie group is still a Lie group, so we can preserve most of the information by studying simply connected Lie groups.

3.2 Representation Theory and Classification

A representation of a Lie group is a morphism π:G→G⁢Ln and a representation of a Lie algebra is a morphism 𝔤→𝔤⁢𝔩n.

Lemma 3.10.

Representations of G differentiate to representations of 𝔤. In particular, if G is simply connected, differentiation is an equivalence of categories.

A Lie algebra is simple if it has no nontrivial ideals 𝔞 such that [𝔤,𝔞]⊆𝔞. A solvable Lie algebra is a Lie algebra where the series [𝔤,𝔤],[[𝔤,𝔤],[𝔤,𝔤]],… eventially arrives at 0.

We will now state the classification of finite-dimensional semisimple Lie algebras over ℂ. These are simply sums of simple Lie algebras, so we simply need to classify the simple Lie algebras, which are simply given by Dynkin diagrams:

The correspondence is given by the vertices corresponding to simple roots in 𝔤.

Definition 3.11.

A connected Lie group is semisimple if its Lie algebra is semisimple.

Our slogan will be that 𝔰⁢𝔩2∈𝔰⁢𝔩n⊂semisimple⊂Kac-Moody. A Kac-Moody Lie algebra is a generalization of the semisimple or affine Lie algebras. An important example of such an object is the loop group. For a real manifold M, the loop space of M is L⁢M=Map⁡(S1,M). For a complex manifold ℳ, then L⁢M=Hom⁡(ℂ∗,𝔐), and finally if X is an algebraic variety, L⁢X=Hom⁡(𝔸1∖0,X).

If G is a Lie group, then L⁢G is an infinite-dimensional version of a Lie group and Lie⁡(L⁢G)=L⁢Lie⁡(G). If G is an algebraic group, then L⁡(G)⁢(R)=G⁡(R⁡((t))) for any ring R, where we think of G and L⁢G as functors.

Now we will consider the representation theory of S⁢Ln. Note that G⁢L1⁢(ℂ)=𝔾m=S1×ℝ. The torus is a complex Lie group T≃𝔾mn. We will consider a maximal torus T⊆S⁢Ln. There are also called Cartans.

Theorem 3.12.

All maximal tori are conjugate to the set of diagonal matrices.

We will denote the set of roots of T in G=S⁢Ln by Δ. We will aim to describe 𝔤 in terms of 𝔱=Lie⁡T.

Lemma 3.13.

Let U be a compact Lie group. Then it has a left-invariant Haar measure that is unique up to constants and 𝖱𝖾𝗉fd⁡(U,ℂ) is semisimple.

Proof.

We need to find a left-invariant differential form. Fortunately, the tangent and cotangent bundles are trivial, so in particular, the canonical bundle is trivial.

Once we have a Haar measure, choose an inner product h0. Then we can write an invariant inner product by

Then we finish the proof as in the case of finite groups. □

Corollary 3.14.

Let G be a connected complex Lie group with a compact real form U. Then let V∈𝖱𝖾𝗉⁡G and h be the U-invariant hermitian inner product on V. Then h is invariant under G.

Proof.

Because G is connected, we can transfer the problem to the Lie algebra. The condition that h⁡(g⁢x,g⁢y)=h⁡(x,y) becomes h⁡(α⁢x,y)+h⁡(x,α⁢y)=0. The other direction is true when g∈exp⁡(𝔤) because exp⁡(𝔤) generates G. Therefore, we can reduce the problem to 𝔲 because 𝔤 is its complexification. However, we can finally reduce from 𝔲 to U, so we are done. □

Example 3.15.

Let T=(ℂ∗)n, Then Tℝ=(S1)n is compact.

Lemma 3.16.

For any finite dimensional representation V of a torus,

Here X∗⁢(T)≔Hom⁡(T,ℂ∗).

Note that any character χ:ℂ∗→ℂ∗ is given by z↦zn for some n∈ℤ, so we see that χ∗⁢(T)≃ℤn.

Because S⁢Ln⁢(ℂ) has a compact real form S⁢U⁢(n), this means that 𝖱𝖾𝗉⁡S⁢Ln is a semisimple category. Then 𝔤=𝔲+i⁢𝔲, and there is an isomorphism G=U×𝔲 given by (u,x)↦u⁢exp⁡(x). Therefore π1⁢(G)=π1⁢(U). In particular, for n=2, S⁢U⁢(2)=S3, so it is simply connected. We also have a fiber sequence as below:

Given a maximal torus T, its Lie algebra 𝔱 acts on V∈𝖱𝖾𝗉⁡T. If T acts by χ, then 𝔱 acts by deχ:𝔱→ℂ, so deχ∈𝔱∗. We will denote these functionals by λ, so define

Lemma 3.17.

The eigenspaces of χ and deχ are the same. In particular, differentiation is injective.

Lemma 3.18.

For all x∈𝔤 there exists a unique Θx:(ℝ,+)→G such that d0⁢(Θx)⁢(1)=x. We denote exp⁡x≔Θx⁢(1).

For G=ℂ∗, we see that exp⁡τ=eτ, so es⁢z=χs⁢(ez). If we set z=2⁢π⁢i, then we see that s∈ℤ. Alternatively, we can differentiate z↦zn at z=1.

We will define the set of coroots by X∗⁢(T)=Hom⁡(ℂ∗,T). Then it is easy to see that X∗⁢(T) and X∗⁢(T) are dual lattices under the pairing

Corollary 3.19.
  1. 1.

    Differentiation at 1 gives an injection X∗⁢(T)↪𝔱;

  2. 2.

    Differentiation at the identity gives an injection X∗⁢(T)↪𝔱∗;

  3. 3.

    The map X∗⁢(T)⊗ℂ→𝔱 given by η⊗z↦z⁢d1⁢η is an isomorphism.

We will now study the system of roots. Let 𝕍=X∗⁢(T)⊗ℝ≕𝔱ℝ∗. Then for V∈𝖱𝖾𝗉⁡T, write

A Cartan subalgebra of a Lie algebra 𝔤 is a maximal toral subalgebra. Then for all s∈𝔱, the Lie bracket [s,−]:𝔤→𝔤 is a semisimple operator.

Lemma 3.20.

If T⊆G is a Cartan, then 𝔤=⨁α∈𝒲⁡(g)𝔤α

Example 3.21.

In both G⁢Ln and S⁢Ln, the maximal torus is the set of diagonal matrices, and for G=G⁢Ln, 𝔱 is the set of diagonal matrices. In S⁢Ln, the Cartan subalgebra is the set of trace zero diagonal matrices.

Define the set of roots Δ𝔱⁢(𝔤)=𝒲⁡(𝔤)∖{0}. This tells us that 𝔤=⨁x∈Δ𝔤α⊕Z𝔤⁢(𝔱) and that for V∈𝖱𝖾𝗉⁡G, if λ∈𝒲⁡(V) and α∈𝒲⁡(𝔤), then gα⁢Vλ⊆Vλ+α.

3.3 Representation Theory of 𝔰⁢𝔩n

Lemma 3.22.
  1. 1.

    For 𝔤=𝔰⁢𝔩n, Δ={αi⁢j=εi−εj∣i≠j};

  2. 2.

    The weight spaces are simply 𝔤αi⁢j=ℂ⁢Ei⁢j;

  3. 3.

    𝔤=⨁α∈Δ𝔤α⊕𝔱;

  4. 4.

    For all α∈Δ, dim𝔤α=1;

  5. 5.

    If α,β,α+β∈Δ, then [𝔤α,𝔤β]⊆𝔤α+β.

Lemma 3.23.
  1. 1.

    If α,β∈Δ but 0≠α+β∉Δ, then [𝔤α,𝔤β]=0;

  2. 2.

    [𝔤α,𝔤−α]=ℂ⁢(αi⁢j)∨⊆𝔱.

In the case n=2, we can write a basis for 𝔤:

with the identities

Lemma 3.24.

Any root α∈Δ defines an 𝔰⁢𝔩2-subalgebra Sα=𝔤α⊕𝔤−α⊕[𝔤α,g−α].

  1. 1.

    There exists a Lie algebra map φ:𝔰⁢𝔩2↪𝔤 with φ⁢e∈𝔤α, φ⁢f∈𝔤−α;

  2. 2.

    The image is a Lie subalgebra;

  3. 3.

    The image is isomorphic to 𝔰⁢𝔩2;

  4. 4.

    φ⁢h is independent of φ.

Proof.

The first three parts are easy. Suppose ψ⁢e=a⁢φ⁢e and ψ⁢f=b⁢φ⁢f. Then

so a⁢b=1. □

Corollary 3.25.

We see that ⟨α,αˇ⟩=2.

Now we will consider reflections. If v∈V and v∗∈V∗, then if ⟨v,v∗⟩=2, there is an involution sv,v∗⁢(x)=x−⟨v∗,x⟩⁢v. Then we see that

The group W generated by all reflections is called the Weyl group of Δ. For G=S⁢Ln, W=Sn.

Lemma 3.26.

For w∈W and α∈Δ, w⁡(Sα)=Sw⁢α.

Lemma 3.27.

V has a W-invariant inner product.

The inner product is simply the standard inner product on V0∗=⨁ℝ⁢εi=ℝn. This restricts to V∗, so we can compute the inner product of the indices. An inner product of 0 corresponds to an angle of π/2, 1 corresponds to an angle of π/6, and 2 corresponds to an angle of 0.

Proposition 3.28.

Define the coroot αˇ=2⁢α(α,α)∈V≃V∗.

  1. 1.

    For all α∈Δ, sα⁢Δ=Δ;

  2. 2.

    For all α,β∈Δ, (α,βˇ)∈ℤ.

  3. 3.

    Δ spans V.

We will now describe the irreducible representations of 𝔰⁢𝔩2. Note that Δ={α,−α} and X∗⁢(T)=ℤ⁢α. Thus we will see that the irreducible representations of 𝔰⁢𝔩2 are simply natural numbers, where

We see that e increses the weight by 2 and f decreases the weight by 2. In addition, if (aa−1)∈T, then

where ρ⁢(aa−1)=a. The representation given above is the only irreducible representation of 𝔰⁢𝔩2 with highest weight n. In general, we have irreducible representations with highest weight λ.

Note that ℕ⁢ρ⊂ℤ⁢ρ is a direction in X∗⁢(T), which comes from a choice of Borel, so the roots come from 𝔫,𝔟.

For 𝔰⁢𝔩n, define the dominant cone X∗⁢(T)+⊂X∗⁢(T) as the set of weights λ∈X∗⁢(T) which are positive with respect to all roots in 𝔫,𝔟.

Definition 3.29.

In X∗⁢(T), we say that λ≥μ if λ−μ is in the ℕ-span of Δ+.

Definition 3.30.

In a representation V∈𝖱𝖾𝗉⁡𝔰⁢𝔩n, v is primitive of weight λ if v∈Vλ for some maximal weight λ.

Theorem 3.31.

For all λ∈X∗⁢(T)+, there exists a unique representation L of 𝔤 such that

  1. 1.

    L has highest weight λ;

  2. 2.

    There exists a primitive vector v of weight λ, where 𝔫⋅v=0.

Proof.

We know that

Because λ is maximal, then Vλ+α=0. □