Representation Theory IV

5 Problems for Epiphany

5.1 Problems for section 1

Problem 1.
  1. 1.

    Compute exp(X) for X equal to (t00s), (0t-t0), and (0tt0) (where s,t).

  2. 2.

    Let Ea,b be the elementary n×n matrix with 1 in the (a,b)-entry and 0 elsewhere. Compute exp(tEa,b) for ab and a=b.

Problem 2.

Show that

exp(tX)exp(tY)=exp(t(X+Y)+t22[X,Y]+O(t3))

as t0, where

[X,Y]=XY-YX.
Problem 3.

Let 𝔫 be the -vector space of strictly upper triangular matrices (0’s on the diagonal) and let N={gGLn():g=I+X,X𝔫}.

In this problem we will see that the restriction of the exponential to 𝔫 is a homeomorphism onto N (i.e. a continuous bijection with continuous inverse).

  1. 1.

    Let X𝔫. Show that Xn=0.

  2. 2.

    Show that exp(X)N for X𝔫.

  3. 3.

    Show that, for gN, the logarithm log(g)=k=1(-1)k+1(g-I)kk is in fact a finite sum (and hence converges).

  4. 4.

    Show that exp|𝔫 and log|N are inverses of each other. Hint: this boils down to an identity of formal power series, which you can actually deduce from the corresponding fact over .

Problem 4.
  1. 1.

    Using the previous question, fill in the gaps of the proof from the notes that

    exp:𝔤𝔩n,GLn()

    is surjective.

  2. 2.

    (+) Is the exponential map exp:𝔰𝔩2,SL2() surjective? What about exp:𝔤𝔩2,GL2+()?

Problem 5.

Let v3 be a unit vector and let f:SO(3) be the map with f(θ) being rotation by θ about the axis v (the angle is measured anticlockwise as you look along the vector from the origin).

Show that f is a one-parameter subgroup and find its infinitesimal generator in terms of v.

Problem 6.

Prove that the Lie algebra of U(n) is

𝔲n={X𝔤𝔩n,:X+X=0}

and find its (real) dimension. Is it a complex vector space?

Problem 7.

Let Ip,q=(Ip-Iq), where Ik denotes the identity matrix of size k. Let n=p+q. Let

O(p,q)={gGLn():gIp,qgT=Ip,q}

be the orthogonal group of signature (p,q). Let SO(p,q)=O(p,q)SLn(). We let 𝔬p,q and 𝔰𝔬p,q be their Lie algebras.

Show that the Lie algebra 𝔬p,q is given by

𝔬(p,q)={XMn():XIp,q+Ip,qXT=0}

and that 𝔰𝔬p,q=𝔬p,q.

Problem 8.
  1. 1.

    Show that the Lie algebras 𝔰𝔬3 and 𝔰𝔲2 are isomorphic. (Later on, we will see a conceptual reason for this).

    Hint: it is enough to find a basis for 𝔰𝔬3 and a basis for 𝔰𝔲2 which satisfy the ‘same’ Lie bracket relations. Try using the basis of 𝔰𝔬3 consisting of infinitesimal generators for rotations around the axes, and a basis for 𝔰𝔲2 related to the quaternions.

  2. 2.

    Show that the Lie algebras 𝔰𝔬2,1 and 𝔰𝔩2, are isomorphic. See Problem 7 for the definition of 𝔰𝔬2,1.

  3. 3.

    (+) Show that the Lie algebras 𝔰𝔬3,1 and 𝔰𝔩2, are isomorphic (as real Lie algebras). See Problem 7 for the definition of 𝔰𝔬3,1.

Problem 9.

Show that:

  1. 1.

    If X𝔰𝔭2n, then tr(X)=0.

  2. 2.

    (+) Show that, if gSp(2n), then det(g)=1.

Problem 10.

Prove that if G is a Lie group and G0 is the connected component of the identity, then the subgroup G0 is normal.

Problem 11.
  1. 1.

    Give a direct proof that SO(3) is connected, by constructing a path from an arbitrary element of SO(3) to the identity. Hint: every element of SO(3) is rotation by some angle about some axis.

  2. 2.

    Prove by induction on n that SO(n) is connected for all n1.

Problem 12.

Show that a general element of SU(2) may be written

(a-b¯ba¯)

for a,b with |a|2+|b|2=1.

Deduce that SU(2) is homeomorphic to the three-sphere S3={v4:|v|=1}.

In other words, write down a bijection SU(2)S3 with continuous inverse. Don’t worry about checking that the maps are continuous, just write them down. The result of this problem implies that SU(2) is simply-connected, because S3 is.

Problem 13.

Show that, if G is a connected (linear) Lie group with Lie algebra 𝔤, then G is abelian if and only if 𝔤 is (see Definition 1.30). Hint: for the converse, consider the adjoint map GGL(𝔤).

What goes wrong if G is not connected?

Solution: see Proposition 2.14.

Problem 14.

If 𝔤 is a Lie algebra, let 𝔷 be its centre:

𝔷={X𝔤:[X,Y]=0:for all Y𝔤.}.

Suppose that G is a connected Lie group with centre Z and Lie algebra 𝔤 with centre 𝔷.

Prove that 𝔷 is the Lie algebra of Z.

You will always have Lie(Z)𝔷 but the reverse inclusion requires that G is connected.

Problem 15.

Solve the exercises in section 1.8

5.2 Problems for section 2

Problem 16.
  1. 1.

    If (ρ,V) is an irreducible finite-dimensional complex representation of 𝔤 and 𝔷 is the centre of 𝔤 (see problem 14), show that there is a linear map α:𝔷 such that ρ(Z)v=α(Z)v for all Z𝔷.

  2. 2.

    For 𝔤=GLn,, find 𝔷. Find α when V=Λkn, where n is the standard representation and 1kn. These representations are in fact irreducible, though we haven’t proved that yet; you can just directly show that α exists.

Problem 17.

Prove that, for X,Y𝔤𝔩n,

(adX)m(Y)=[X,[X,,[X,Y]]]=k=0m(mk)XkY(-X)m-k.

Hence give a direct proof that exp(adX)=Adexp(X).

Problem 18.

Consider G=U(1).

  1. 1.

    For φ a continuous function on G, we define its integral

    Gφ(g)𝑑g=12π02πφ(eit)𝑑t.

    Note that g1𝑑g=1. Show that

    Gφ(hg)𝑑g=Gφ(gh)𝑑g=Gφ(g)𝑑g

    for any hG.

  2. 2.

    Let (V,ρ) be a finite dimensional representation of G and let (,) be any Hermitian form on V. Define a new Hermitian form by

    (v,w)ρ=G(ρ(g)v,ρ(g)w)𝑑g.

    Show that (,)ρ is a G-invariant Hermitian form on V.

  3. 3.

    Conclude that every finite-dimensional representation of U(1) is completely reducible. (This is analogous to Maschke’s theorem for finite groups.)

Problem 19.

Consider the orthogonal group O(2).

  1. 1.

    Show that SO(2) has index 2 in O(2). Deduce that every element in O(2) can be uniquely written as rθ or rθs with s=(0110) and rθ the matrix for rotation by θ. Show that

    srθ=r-θs.
  2. 2.

    Mimic the method we used for dihedral groups to classify all irreducible finite-dimensional representations of O(2).

Solution

Problem 20.

Let V be the space of functions on 2 that are polynomials in the coordinates x and y. Consider the (left) action of GL2() on V given by

(gφ)(v)=φ(g-1v)

(here, think of v=(xy)2 as a column vector).

Compute the derived action for the “standard” basis of 𝔰𝔩2() given by X=(0100), Y=(0010), and H=(100-1). You should get something involving the partial derivatives x and y.

Problem 21.

Let V=2 be the standard representation of GL2().

  1. 1.

    Show that Λ2(V)det as Lie group representations.

  2. 2.

    Show that Λ2(V)tr as representations of 𝔤𝔩2(). (You could just ‘take the derivative’ of part (a), but please do it directly instead.)

  3. 3.

    Find an explicit homomorphism ρ:GL2()GL3() corresponding to Sym2(V).

5.3 Problems for section 3

Problem 22.

Let V, W be representations of 𝔰𝔩2,. Let v and w be two weight vectors of V and W respectively with respective weights α and β. Show that

vwVW

is a weight vector with weight α+β, and that if v and w are highest weight vectors then so is vw.

Problem 23.

Let V be a finite-dimensional representation of 𝔰𝔩2,.

  1. 1.

    What are the weights of the dual representation V*?

  2. 2.

    Deduce that VV*.

Problem 24.

Let (π,V) be a finite-dimensional representation of 𝔰𝔩2,. Consider the Casimir element55 5 Conventions differ; it might be more usual to call 1+2𝒞 the Casimir.

𝒞=π(X)π(Y)+π(Y)π(X)+12π(H)2.
  1. 1.

    Show 𝒞 commutes with the action of 𝔰𝔩2,. Conclude that if V is irreducible then 𝒞 acts as a scalar.

  2. 2.

    What is the scalar for V=Symn(2), the irreducible representation of highest weight n?

  3. 3.

    Compute the action of 𝒞 on the space V of polynomial functions ϕ on 2, with action the derivative of (gϕ)(v)=ϕ(g-1v) (see problem 20).

Problem 25.

If λ, show that there is a (possibly infinite dimensional!) representation of 𝔰𝔩2, with highest weight λ.

Problem 26.

Consider V=Symn(2), the irreducible representation of highest weight n of 𝔰𝔩2,. Decompose the following representations into irreducibles, and find highest weight vectors for the irreducible constituents:

  1. 1.

    Sym2(Sym2(2));

  2. 2.

    Λ2(Sym2(2));

  3. 3.

    Sym3(2)Sym2(2);

  4. 4.

    Sym3(Sym2(2)).

For the third example, find bases for the irreducible subrepresentations.

Problem 27.
  1. 1.

    For ab integers, decompose the representation Syma2Symb2 of 𝔰𝔩2, into irreducibles. (This is known as the Clebsch–Gordan formula).

  2. 2.

    (+) Can you find a general expression for the highest weight vectors for the irreducible subrepresentations? What about for the weight bases?

Problem 28.

Show that the real Lie algebras 𝔰𝔩2, and 𝔰𝔲2 are not isomorphic. Hint: consider the adjoint action of an arbitrary element of 𝔰𝔲2.

Problem 29.

We have that Symn(2) is the irreducible representation of SU(2) of dimension n+1. Let χn be its character. Every conjugacy class of SU(2) contains an element of the form

exp(itH)=(eit00e-it).

Show that

χn(exp(itH))=sin((n+1)t)sin(t).
Problem 30.

Let (ρ,V) be an irreducible representation of 𝔤𝔩2,, and let Z=(1001).

  1. 1.

    Show that ρ(Z) is a scalar.

  2. 2.

    Show that the restriction of V to 𝔰𝔩2, is irreducible.

  3. 3.

    Show that for every λ and integer n0, there is a unique irreducible representation of 𝔤𝔩2, of dimension n+1 with ρ(Z)=λI.

  4. 4.

    Which of these are derivatives of representations of GL2()? Hence classify the finite dimensional holomorphic irreducible representations of GL2().

Problem 31.
  1. 1.

    Verify the formula

    r2Δ=Jx2+Jy2+Jz2+2+

    as operators on 𝒫.

  2. 2.

    Find the image of the Casimir element from problem 24 under our isomorphism 𝔰𝔩2,𝔰𝔬3,, and compare to part 1.

Problem 32.

Let 1.

  1. 1.

    Verify that (x-iy) is a highest weight vector in .

  2. 2.

    By applying the lowering operator, find weight vectors of weights i(-1) and i(-2).

  3. 3.

    Find a basis of weight vectors in when =1 and =2 (see example 3.40).

Problem 33.
  1. 1.

    Prove that, for f𝒫,

    Δ(r2f)=r2Δ(f)+2(2+3)f.
  2. 2.

    Find a similar formula for

    Δ(r2kf)-r2kΔ(f).
  3. 3.

    Use this to give another proof that

    r2𝒫-2={0}.

    (Hint: if f is in the intersection, let f=r2kg, g not divisible by r2).

Problem 34.
  1. 1.

    Let V be the standard — three-dimensional — representation of 𝔰𝔬3. Find a basis of weight vectors for Sym2(V), and decompose it into irreducible subrepresentations.

  2. 2.

    Let 2 be the five-dimensional representation of SO(3). Decompose 22 into irreducible representations.

5.4 Problems for section 4

Problem 35.

Verify that

[(a1000a2000a3),Eij]=(ai-aj)Eij

and

[E12,E23]=E13.
Problem 36.

Consider, instead of 𝔰𝔩3,, 𝔰𝔩2,. What are the roots and root spaces? What is the relation between the weights (as linear functionals on 𝔥) and between the weights defined in section 3?

Problem 37.

Let V=(3)* be the dual of the standard representation, with basis e1*,e2*,e3* dual to the standard basis.

  1. 1.

    Show that the ei* are weight vectors with weights -Li.

  2. 2.

    Find the action of each Eij on e3* and deduce that e3* is a highest weight vector with weight -L3.

Problem 38.

Show that a1L1+a2L2+a3L3ΛW if and only if a1-a2,a2-a3. Must the ai be integers?

Problem 39.

The root lattice ΛRΛW is the subgroup of the weight lattice generated by the roots.

  1. 1.

    Draw a picture showing the root lattice inside the weight lattice.

  2. 2.

    Show that ΛR has index three in ΛW (i.e. the quotient ΛW/ΛR has order three).

  3. 3.

    What would the root lattice and weight lattice be for 𝔰𝔩2,? What is the index in this case?

  4. 4.

    Let V be a finite-dimensional irreducible representation of 𝔰𝔩3,. Show that any two weights of V differ by an element of the root lattice.

Problem 40.

Find the weights of Sym3(3) and draw the weight diagram.

Problem 41.

Using weights, or otherwise, show that

3(3)*𝔰𝔩3,

where is the trivial representation and 𝔰𝔩3, is the adjoint representation.

Problem 42.

(non-examinable) Let (ρ,V) be a representation of 𝔰𝔩3,. As SL3() is simply-connected ρ exponentiates to a representation, ρ~, of SL3(). Let

σ3=(010-100001)SL3().

Show that, for every weight α, ρ~(σ3) is an isomorphism

VαVs3α.

Here s3(a1L1+a2L2+a3L3)=a1L2+a2L1+a3L3.

Give another proof of Theorem 4.26.

Problem 43.

Let a,b0 be integers. Check that

e1a(e3*)bSyma(3)Symb((3)*)

is a highest weight vector with weight aL1-bL3.

Problem 44.

Show that, if V is a finite-dimensional representation of 𝔰𝔩3, with a unique highest weight vector (up to scalar multiplication), then V is necessarily irreducible.

Deduce that the standard representation, its dual, and the adjoint representation are irreducible.

Problem 45.
  1. 1.

    Find the weights of Sym2(3)(3)* and draw the weight diagram.

  2. 2.

    Show that

    e12e1*+e1e2e2*+e1e3e3*Sym2(3)(3)*

    is a highest weight vector with weight L1.

  3. 3.

    Let v=e12e3*. Calculate E32E21v and E21E32v and show that they are linearly independent.

  4. 4.

    Show that

    Sym2(3)(3)*V(2,1)3

    and find the weight diagram for V(2,1).

Problem 46.

(harder!) The aim of this problem is to show that, for n0,

V(n,0)=Symn(3).

It suffices to show that Symn(3) is irreducible with highest weight nL1.

  1. 1.

    Show that Symn(3) has a basis of weight vectors

    {e1ae2be3c:a,b,c0,a+b+c=n}

    and that these have distinct weights (so, every weight has multiplicity one).

  2. 2.

    Show that e1n is the unique highest weight vector in Symn(3), up to scalar multiplication.

  3. 3.

    Deduce that Symn(3) is an irreducible representation with highest weight nL1. See problem 44.

Problem 47.

(monster!) Let V=3, let W=V*, and let a,b>0. For vV,wW, define (v,w)=w(v).

Let

ϕ:Syma(V)Symb(W)Syma-1(V)Symb-1(W)

be defined by

ϕ((v1va)(w1wb))=i=1aj=1b(vi,wj)(v1v^iva)w1w^jwb

where v^i means vi is omitted (and similarly for w^j).

  1. 1.

    Show that ϕ is an 𝔰𝔩3,-homomorphism.

  2. 2.

    Show that Syma(V)Symb(W) has a unique highest weight vector of weight (a-i)L1-(b-i)L3 for each 0imin(a,b), and no other highest weight vectors.

  3. 3.

    Show that the highest weight vector from the previous part is in ker(ϕ) if and only if i=0.

  4. 4.

    Deduce that ker(ϕ)V(a,b) is the irreducible representation of highest weight aL1-bL3.

  5. 5.

    Show that ϕ is surjective, and hence decompose Syma(V)Symb(V*) into irreducibles.

  6. 6.

    Find the dimension of V(a,b). Find its weights.

This problem is hard! For a solution, see Fulton and Harris, section 13.2, but watch out for the unjustified ’clearly’ just before Claim 13.4.