Algebras and Representations MATH

3193 2016 Alison Parker

Visit to download the full and correct content document:
https://textbookfull.com/product/algebras-and-representations-math-3193-2016-alison
-parker/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Math 223a Algebraic Number Theory notes Alison Miller

https://textbookfull.com/product/math-223a-algebraic-number-
theory-notes-alison-miller/

Math 7350 Differential Graded Algebras and Differential

Graded Categories Yuri Berest

https://textbookfull.com/product/math-7350-differential-graded-
algebras-and-differential-graded-categories-yuri-berest/

Representations of Finite Groups I Math 240A Robert

Boltje

https://textbookfull.com/product/representations-of-finite-
groups-i-math-240a-robert-boltje/

An Invitation to Unbounded Representations of Algebras

on Hilbert Space 1st Edition Konrad Schmüdgen

https://textbookfull.com/product/an-invitation-to-unbounded-
representations-of-algebras-on-hilbert-space-1st-edition-konrad-
schmudgen/
Theory of Groups and Symmetries: Representations of
Groups and Lie Algebras, Applications 1st Edition
Alexey P. Isaev

https://textbookfull.com/product/theory-of-groups-and-symmetries-
representations-of-groups-and-lie-algebras-applications-1st-
edition-alexey-p-isaev/

Quaternion Algebras in Number Theory Spring 2016

Kimball Martin

https://textbookfull.com/product/quaternion-algebras-in-number-
theory-spring-2016-kimball-martin/

Gelfand Triples and Their Hecke Algebras Harmonic

Analysis for Multiplicity Free Induced Representations
of Finite Groups Tullio Ceccherini-Silberstein

https://textbookfull.com/product/gelfand-triples-and-their-hecke-
algebras-harmonic-analysis-for-multiplicity-free-induced-
representations-of-finite-groups-tullio-ceccherini-silberstein/

Ryder The Lost Breed MC 1 1st Edition Ali Parker Weston

Parker Parker Ali Parker Weston

https://textbookfull.com/product/ryder-the-lost-breed-mc-1-1st-
edition-ali-parker-weston-parker-parker-ali-parker-weston/

Axel The Lost Breed MC 2 1st Edition Ali Parker Weston

Parker Parker Ali Parker Weston

https://textbookfull.com/product/axel-the-lost-breed-mc-2-1st-
edition-ali-parker-weston-parker-parker-ali-parker-weston/
 Algebras and Representations
MATH 3193
2016

Alison Parker, Oliver King

a.e.parker@leeds.ac.uk, O.H.King@leeds.ac.uk
after notes of
Andrew Hubery

An algebra is a set A which at the same time has the structure of a ring and a vector space in
a compatible way. Thus you can add and multiply elements of A, as well as multiply by a scalar.
One example is the set of matrices of size n over a field.
A representation is an action of an algebra on a vector space, similar to how matrices of size
n act on an n-dimensional vector space. It is often the case that information about the algebra
can be deduced from knowing enough about its representations. An analogy might be that one
can begin to understand a complicated function by computing its derivatives, or more generally a
surface by computing its tangent planes.
There are lots of interesting examples of algebras, with applications to mathematics and
physics. In this course we will introduce some of these algebras, as well as some of the general
theory of algebras and their representations.

1
 Contents

Chapter 1. Quaternions 5
1.1. Complex Numbers 5
1.2. Quaternions 5
1.3. Some Remarks (non-examinable) 9

Chapter 2. Algebras 10
2.1. Basic Definition 10
2.2. Division algebras 10
2.3. Characteristic of a Field 11
2.4. Algebras given by a basis and structure coefficients 11
2.5. Polynomials 12
2.6. Group algebras 13
2.7. Matrix algebras 13
2.8. Endomorphism algebras 14
2.9. Temperley-Lieb algebras 14
2.10. Direct Product of Algebras 15

Chapter 3. Homomorphisms and Subalgebras 16

3.1. Homomorphisms 16
3.2. Subalgebras 17
3.3. Intersections and generating sets 19

Chapter 4. Ideals 21
4.1. Ideals — Definition and Example 21
4.2. Sums and intersections of ideals 22
4.3. Products of ideals 23

Chapter 5. Quotient Algebras 24

5.1. Definition of Quotient Algebras 24
5.2. The Factor Lemma 25

Chapter 6. Presentations of Algebras 28

6.1. Free Algebras 28
6.2. Relations 30

Chapter 7. Representations 34
7.1. Basic Definition 34
2
 CONTENTS 3

7.2. Examples 35
7.3. Representations of quotient algebras 36
7.4. Representations of group algebras 37
7.5. Equivalence 37
7.6. Direct product of representations 38

Chapter 8. Modules 39
8.1. Basic Definition 39
8.2. Direct product of modules 40

Chapter 9. Homomorphisms and Submodules 42

9.1. Homomorphisms 42
9.2. Endomorphism algebras of modules 42
9.3. Submodules 43
9.4. Submodules given via ideals 44
9.5. Restriction of scalars 45

Chapter 10. Quotient Modules and the First Isomorphism Theorem 46

10.1. Quotient Modules 46
10.2. First Isomorphism Theorem 46
10.3. Direct products again 47
10.4. Generating sets 47

Chapter 11. More on homomorphism spaces 48

11.1. The Opposite Algebra 48
11.2. Homomorphisms between direct products 49

Chapter 12. Semisimple Algebras 51

12.1. Semisimple Modules 51
12.2. Schur’s Lemma 52
12.3. Complements 52

Chapter 13. Wedderburn’s Structure Theorem 54

13.1. Semisimple algebras 54
13.2. Wedderburn’s Structure Theorem 55

Chapter 14. The Jacobson Radical 57

14.1. Definition of the Jacobson Radical 57
14.2. Nilpotent Ideals 58
14.3. Two key theorems 59
14.4. Examples 60

Chapter 15. Modules over the polynomial algebra 61

15.1. Submodules 61
15.2. Direct sums 62
15.3. Generalised eigenspaces 62
 CONTENTS 4

15.4. The minimal polynomial 62

Appendix A. Rotations 64
A.1. Orthogonal Matrices 64
A.2. Rotations in 2-Space 65
A.3. Rotations in 3-space 66
A.4. Rotations in n-space 67

Appendix B. Review of Some Linear Algebra 68

B.1. Vector spaces 68
B.2. Matrices 69
B.3. Linear Combinations 69
X
B.4. The example of the vector space of functions K 69
B.5. Subspaces 70
B.6. Sums and Intersections 70
B.7. Quotient Spaces 71
B.8. Linear Maps 71
B.9. Bases 73
B.10. Dimension 74
 CHAPTER 1

Quaternions

1.1. Complex Numbers

Since the work of Wessel (1799) and Argand (1806) we think of complex numbers as formal
expressions
z = x + yi with x, y ∈ R,
which we can add and multiply by expanding out, substituting i2 = −1, and collecting terms. In
other words, we have a two-dimensional real vector space C with basis {1, i}, on which we have
defined a multiplication
C×C→C
satisfying the following properties for all a, b, c ∈ C and λ ∈ R:
Associative a(bc) = (ab)c.
Unital there exists 1 ∈ C with 1a = a = a1.
Bilinear a(b + λc) = ab + λac and (a + λb)c = ac + λbc.
Commutative ab = ba.
Complex numbers have wonderful properties, for example:
p
• The conjugate of z = x + yi is z̄ = x − yi, and its absolute value is |z| = x2 + y 2 . Thus
|z̄| = |z| and z z̄ = |z|2 . Also, |zw| = |z||w| for all complex numbers z, w.
• Every non-zero complex number z = x + yi has an inverse

z −1 = (x − yi)/(x2 + y 2 ) = z̄/|z|2 ,

so they form a field.

• Rotations of the plane correspond to multiplication by complex numbers of absolute value
1.

1.2. Quaternions

Trying to find a way to represent rotations in three dimensions,

Sir William Rowan Hamilton invented the quaternions in 1843. He needed four real numbers,
not three, and also had to drop commutativity.
Quaternions are expressions a + bi + cj + dk with a, b, c, d ∈ R. They add and subtract in the
obvious way, and multiply by first expanding out (being careful with the ordering), making the
following substitutions

i2 = −1 ij = k ik = −j
ji = −k j 2 = −1 jk = i
ki = j kj = −i k 2 = −1

5
 1.2. QUATERNIONS 6

and then collecting terms. For example

(2 + 3i)(i − 4j) = 2i − 8j + 3i2 − 12ij = −3 + 2i − 8j − 12k.

The set of all quaternions is denoted H.

Remark 1.2.1. This looks a bit like the multiplication rule for cross product except i × i = 0
and not −1. So we can’t use the determinant trick to work out the product.

In other words we have a four-dimensional real vector space H with basis {1, i, j, k}, on which
we have defined a multiplication
H×H→H

satisfying the following properties for all a, b, c ∈ H and λ ∈ R:

Associative a(bc) = (ab)c.
Unital there exists 1 ∈ H with 1a = a = a1.
Bilinear a(b + λc) = ab + λac and (a + λb)c = ac + λbc.

Remark 1.2.2. Multiplication of quaternions is not commutative:

ij = k but ji = −k, so ij 6= ji.

Some basic definitions: these are all analogous to the definitions for the complex numbers.

Definition 1.2.3. Let q = a + bi + cj + dk be a quaternion.

We say that the real part of q is a, and the imaginary part is bi + cj + dk. A pure quaternion
is one whose real part is zero.
The conjugate of q is q̄ = a − bi − cj − dk. Thus if a is a real number and p is a pure quaternion,
then the conjugate of a + p is a − p.
√
The absolute value of q is |q| = a2 + b2 + c2 + d2 . Note that |q̄| = |q|.

Some elementary properties:

(1) If p is a pure quaternion then p2 = −|p|2 . For,

(bi + cj + dk)(bi + cj + dk)

= b2 i2 + c2 j 2 + d2 k 2 + bc(ij + ji) + bd(ik + ki) + cd(jk + kj)
= −(b2 + c2 + d2 ).

(2) Conjugation is a linear map, so

p + λq = p̄ + λq̄ for all p, q ∈ H and λ ∈ R.

(3) Conjugation satisfies

q̄¯ = q and pq = q̄ p̄ for all p, q ∈ H.

The first property is clear, so we just need to prove the second. Since multiplication is
bilinear and conjugation is linear, we can reduce to the case when p, q ∈ {1, i, j, k}. If
p = 1 or q = 1, then the result is clear, so we may assume that p, q ∈ {i, j, k}.
 1.2. QUATERNIONS 7

If p = q, then p2 = −1 = p̄2 . Otherwise p and q are distinct elements of {i, j, k}. Let
r be the third element. Then p̄ = −p, and similarly for q and r. Using the multiplication
rules we see that pq = ±r and qp = ∓r, so

q̄ p̄ = (−q)(−p) = qp = ∓r = ±r̄ = pq

as required.
(4) We have
q q̄ = q̄q = |q|2 .
For, write q = a + p with a ∈ R and p a pure quaternion, so that q̄ = a − p. Then
ap = pa,(as a is real), so

q q̄ = (a + p)(a − p) = a2 + pa − ap − p2 = a2 − p2 = (a − p)(a + p) = q̄ q.

Using p2 = −|p|2 we get

q q̄ = a2 − p2 = a2 + |p|2 = |q|2 .

(5) For any two quaternions p and q we have

|pq| = |p||q|.

For
|pq|2 = pq pq = p q q̄ p̄ = p |q|2 p̄ = |q|2 p p̄ = |q|2 |p|2 .
The first equality follows from property 4, the second from property 3, the third from
property 4, the fourth as |q|2 is real, and the fifth by property 4 again. Finally the answer
follows by taking square roots.

Lemma 1.2.4. Any non-zero quaternion has a multiplicative inverse; that is, H is a division
algebra.

Proof. If q 6= 0 then the inverse of q is q −1 = q̄/|q|2 . 

Lemma 1.2.5. Every quaternion can be written in the form

q = r cos( 12 θ) + sin( 12 θ)n

where r, θ ∈ R with r = |q| ≥ 0 and θ ∈ [0, 2π], and n is a pure quaternion of absolute value 1.

The use of 12 θ is traditional; the reason will become clear later.

Proof. If q = 0, then take r = 0 and θ and n arbitrary. Otherwise q/|q| is a quaternion of

absolute value 1. So it’s enough to prove it for quaternions of absolute value 1.
Now let q have absolute value 1 and write q = a + p with a ∈ R and p a pure quaternion. Then

1 = |q|2 = a2 + |p|2 ,

so we can write a = cos( 21 θ) for some unique θ ∈ [0, 2π]. (Note that as a2 ≤ 1 we have −1 < a < 1.
Hence |p| = sin( 21 θ). (Note as θ ∈ [0, 2π] that this RHS is indeed non-negative.) Finally, if
θ = 0, 2π, then p = 0 so we can take n to be arbitrary; otherwise n = p/|p| is a pure quaternion of
absolute value 1. We then have
p
q = a + p = cos( 12 θ) + sin( 12 θ) = cos( 12 θ) + n sin( 12 θ)
|p|
 1.2. QUATERNIONS 8

in the required form. 

Lemma 1.2.6. We may identify the set of pure quaternions P = {bi + cj + dk | b, c, d ∈ R}

with R3 such that i, j, k correspond respectively to the standard basis vectors e1 , e2 , e3 . We equip
R3 with the usual dot product and cross product. Then

pq = −p · q + p × q ∈ H for all p, q ∈ P.

Note that the dot product of two elements of P is in R, and the cross product is in P , so the
sum makes sense in H.

Proof. Each operation is bilinear, so it suffices to check this for p, q ∈ {i, j, k}. This gives 9
possible cases and symmetry means we only need to check 3, namely i2 , ij, and ji.

−i · i + i × i = −1 + 0 = i2 , −i · j + i × j = 0 + k = ij, −j · i + j × i = 0 − k = ji.

The following theorem explains the reason for the 21 θ in Lemma 1.2.5.

Theorem 1.2.7. If q is a quaternion of absolute value 1, then the linear transformation

Rq : P → P, Rq (p) := qpq −1

is a rotation. Explicitly, if q = cos( 21 θ) + sin( 12 θ)n, then Rq = Rn,θ is the rotation about axis n
through angle θ.
Two quaternions q, q 0 of absolute value 1 give the same rotation if and only if q 0 = ±q.

Proof. Recall that an ordered basis (f1 , f2 , f3 ) of R3 is called a right-handed orthonormal

basis provided that
fi · fj = δij and f3 = f1 × f2 .
(NB: the “ortho” means right angled, the “normal” means length one and the “right-handed”
comes from the right-handed rule for cross product.) Also, if n ∈ R3 has length 1, then the
rotation about axis n through angle θ is the linear map

n 7→ n, u 7→ cos(θ)u + sin(θ)v, v 7→ − sin(θ)u + cos(θ)v

where (n, u, v) is any right-handed orthonormal basis. With respect to this basis the matrix of the
rotation is:  
1 0 0
 
0 cos θ sin θ 
 
0 − sin θ cos θ
Now let q be a quaternion of absolute value 1. By Lemma 1.2.5 we can write q = cos( 21 θ) +
sin( 12 θ)n with θ ∈ [0, 2π] and n a pure quaternion of absolute value 1. Let (n, u, v) be a right-handed
orthonormal basis for P . The previous lemma tells us that

nu = −n · u + n × u = v and un = −u · n + u × n = −n × u = −v.

Similarly
uv = n = −vu and vn = u = −nv.
 1.3. SOME REMARKS (NON-EXAMINABLE) 9

For simplicity set c := cos( 12 θ) and s := sin( 21 θ). Then q = c + s and q −1 = c − s. Now, since
n2 = −|n|2 = −1 we have

qnq −1 = (c + sn)n(c − sn) = (c + sn)(cn − sn2 ) = (c + sn)(cn + s)

= c2 n + cs + csn2 + s2 n = (cs − cs) + (c2 + s2 )n = n.

Similarly

quq −1 = (c + sn)u(c − sn) = (c + sn)(cu − sun) = (c + sn)(cu + sv)

= c2 u + csv + csnu + s2 nv = (c2 − s2 )u + 2csv = cos(θ)u + sin(θ)v

using the double angle formula and hence

qvq −1 = qnuq −1 = qnq −1 quq −1 = cos(θ)nu + sin(θ)nv = − sin(θ)u + cos(θ)v.

This is the rotation claimed.

Any particular rotation occurs in exactly two ways, as the rotation about axis n through angle
θ, and as the rotation about −n through angle 2π − θ. The latter corresponds to the quaternion

cos(π − 12 θ) + sin(π − 12 θ)(−n) = − cos( 21 θ) − sin( 12 θ)n = −q. 

1.3. Some Remarks (non-examinable)

The set of quaternions of absolute value 1 form a group under multiplication, denoted Sp(1),
and it is not hard to see that the map q 7→ Rq from the previous theorem defines a surjective group
homomorphism R : Sp(1) → SO(3, R) to the group of rotations of R3 . This group homomorphism
is a double cover, meaning that there are precisely two elements of Sp(1) mapping to each rotation
in SO(3, R).
In fact, we can say more. A quaternion q = a + bi + cj + dk has absolute value 1 precisely
when a2 + b2 + c2 + d2 = 1, so Sp(1) can be thought of as a 3-sphere

S 3 = {(a, b, c, d) ∈ R4 : a2 + b2 + c2 + d2 = 1}.

Similarly SO(3, R) ⊂ M3 (R) ∼

= R9 . Therefore both of these sets have an induced topology on them,
and both the multiplication and inversion maps are continuous, so they are topological groups. In
this set-up the group homomorphism Sp(1) → SO(3, R) is also continuous.
Let us fix a pure quaternion n of absolute value 1. Then, as θ increases from 0 to 2π, we get
a sequence of rotations starting and ending at the identity. The sequence of quaternions, however,
starts at 1 but ends at −1. We therefore only get 4π-periodicity for the quaternions. This is
relevant in quantum mechanics for ‘spin 1/2’ particles like electrons.
We can visualise this by rotating a book held in a hand: a 2π rotation returns the book to
its original position, but a 4π rotation is needed to return both the book and the hand to their
original positions.
You can read about the ‘quaternion machine’ in J. Conway and R. Guy, The book of numbers.
 CHAPTER 2

Algebras

2.1. Basic Definition

Definition 2.1.1. Fix a base field K, for example R or C.

An algebra over K, or K-algebra, consists of a K-vector space A together with a multiplication

A × A → A, (a, b) 7→ ab,

satisfying the following properties for all a, b, c ∈ A and λ ∈ K:

Associative a(bc) = (ab)c.
Unital there exists 1 ∈ A such that 1a = a = a1.
Bilinear a(b + λc) = ab + λ(ac) and (a + λb)c = ac + λ(bc).
Remark 2.1.2. If you have seen the definition of a ring, then you will see that the distributivity
axiom has been replaced by the stronger bilinearity axiom. We can do this since our algebra is a
priori a vector space.
An alternative description would therefore be that A is both a vector space and a ring, and
that these structures are compatible in the sense that scalars can always be brought to the front,
so a(λb) = λ(ab).

Remark 2.1.3. (1) In the literature, the algebras we consider might be called unital,
associative algebras. There are other types: Banach algebras are usually non-unital; Lie
algebras and Jordan algebras are non-associative.
(2) Recall that a vector space V is finite dimensional if it has a finite basis. Not all of our
algebras will be finite dimensional.
(3) There is a very rich theory of commutative algebras, where one assumes that the mul-
tiplication is commutative, so ab = ba for all a, b ∈ A. This is related to, amongst
other things, algebraic geometry and algebraic number theory. In this course we will be
concerned with general, non-commutative, algebras.

Example 2.1.4. (1) K is a 1-dimensional algebra over itself.

(2) C is a 2-dimensional R-algebra with basis {1, i} as a vector space over R. (It is also a
1-dimensional C-algebra, as in Example 1.)
(3) H is a 4-dimensional R-algebra with basis {1, i, j, k}. Even though it contains a copy of
C, for example with basis {1, i}, it cannot be considered as a C-algebra since i does not
commute with j and k.

2.2. Division algebras

Definition 2.2.1. A division algebra is a non-zero algebra A in which every non-zero element
has a multiplicative inverse; that is, for all a 6= 0 there exists a−1 such that aa−1 = 1 = a−1 a.
10
 2.4. ALGEBRAS GIVEN BY A BASIS AND STRUCTURE COEFFICIENTS 11

Lemma 2.2.2. A division algebra has no zero divisors: ab = 0 implies a = 0 or b = 0.

Proof. If ab = 0 and a 6= 0, then 0 = a−1 (ab) = (a−1 a)b = 1b = b, so b = 0. 

Example 2.2.3. A field is the same as a commutative division algebra. The quaternions form
a division algebra, but are not commutative so do not form a field.

2.3. Characteristic of a Field

Definition 2.3.1. The characteristic of a field K is the smallest positive integer n such that
n · 1 = 0 ∈ K. If no such n exists, we define the characteristic to be zero.

Example 2.3.2. (1) The fields Q, R and C all have characteristic zero.
(2) If p ∈ Z is a prime, then Fp = Zp (the field with p elements) is a field of characteristic p.
(3) The characteristic of a field is either zero or a prime number. For, if char(K) = n and
n = ab with 1 < a, b < n, then ab = n = 0 ∈ K, whence either a = 0 ∈ K or b = 0 ∈ K,
contradicting the minimality of n.

2.4. Algebras given by a basis and structure coefficients

One immediate question which arises is how to describe an algebra. We shall give two answers
to this question, both having their advantages and disadvantages.
The first answer starts from the fact that A is a vector space, so has a basis {ei | i ∈ I}, for
some indexing set I and we need to define a bilinear multiplication on A.
Recall that to define a linear map f , we just need to specify the images of a basis, since then
P P
f ( i λi ei ) = i λi f (ei ).
Similarly, to define a bilinear map A × A → A, we just need to specify the products ei ej for
all i, j. We can then multiply arbitrary elements of A by expanding out:
X  X  X
λ i ei µj ej = λi µj (ei ej ).
i j i,j

We may display this information in the multiplication table, having rows and columns indexed by
the basis elements, writing the product ei ej in the i-th row and j-th column. This is essentially
what we did when describing the quaternions.
More precisely, each product ei ej again lies in A, so can be expressed uniquely as a linear
combination of the basis elements. Thus one only needs to define the scalars ckij ∈ K, called the
structure coefficients, such that
X
ei ej = ckij ek .
k

We next need to ensure that the multiplication is associative and has a unit.
Using that the multiplication is bilinear, it is enough to check associativity for all triples of
basis elements. So, our multiplication is associative if and only if (ei ej )ek = ei (ej ek ) for all i, j, k.
Note that we can express this entirely in terms of the structure constants as
X p X
cij clpk = clip cpjk for all i, j, k, l.
p p
 2.5. POLYNOMIALS 12

P
Similarly, to check that an element x = i λi ei is a unit, we just need to show that xej =
ej = ej x for all j. In practice, however, it is common to specify in advance that the unit is one of
the basis elements, as we did when describing the complex numbers and quaternions.
This method of describing an algebra is analogous to the method of describing a group by
giving the set of elements and the multiplication table, and for large-dimensional algebras it is just
as unwieldy as it is for large groups.

Example 2.4.1. (1) The vector space with basis {e, f } has a bilinear multiplication given
by
ee = e, ef = 0, f e = 0, f f = f.

Is it an algebra? It is associative: we have e(ee) = e = (ee)e and f (f f ) = f = (f f )f ; all

other possibilities are zero. It has a unit, namely e + f . So yes.
(2) The vector space with basis {1, a, b} has a bilinear multiplication given by

a2 = a, b2 = b, ab = 0, ba = b + 1

and with 1 a unit. Is it an algebra? We have

b(ba) = b(b + 1) = b2 + b = 2b but (bb)a = ba = b + 1

so no: it is not associative.

2.5. Polynomials

The set K[X] of polynomials in an indeterminate X with coefficients in K forms an algebra

by the usual addition and multiplication of polynomials.
This has basis {1, X, X 2 , X 3 , . . .} (so is infinite dimensional) and multiplication X m X n =
X m+n .
More generally, let A be an algebra. Then there is an algebra A[X] whose elements are
polynomials in the indeterminate X with coefficients in A, so of the form a0 + a1 X + a2 X 2 + · · · +
am X m with ai ∈ A, and multiplication
m+n
X r
X
(a0 + a1 X + · · · + am X m )(b0 + b1 X + · · · + bn X n ) = as br−s X r .


r=0 s=0

Observe that if A has basis ei and structure coefficients ckij , then A[X] has basis ei X m and multi-
plication
X
ei X m · ej X n = ckij ek X m+n .
k


In particular, we can inductively define the polynomial algebra K[X1 , . . . , Xr ] to be K[X1 , . . . , Xr−1 ] [Xr ].
This has basis the set of monomials

{X1m1 X2m2 · · · Xrmr | m1 , . . . , mr ≥ 0},

and multiplication

X1m1 · · · Xrmr · X1n1 · · · Xrnr = X1m1 +n1 · · · Xrmr +nr .

 2.7. MATRIX ALGEBRAS 13

2.6. Group algebras

One interesting way of defining an algebra is to start with a group G and take KG to be the
vector space with basis indexed by the elements of G, so the set {eg | g ∈ G}. We now use the
multiplication on G to define a bilinear multiplication on KG:

eg · eh := egh .

Since the multiplication for G is associative, it is easy to see that the multiplication for KG is also
associative. Moreover, the unit for KG is the basis element indexed by the identity element of G,
so 1 = e1 .
This is easiest to see in an example.

Example 2.6.1. Let G be the cyclic group of order 3 with generator g, so that G = {1, g, g 2 }
with g 3 = 1. Then the elements of KG are linear combinations λ + µeg + νeg2 with λ, µ, ν ∈ K,
and as an example of the multiplication we have

(1 + 2eg + 5eg2 )(2 + eg2 ) = 2 + 4eg + 10eg2 + eg2 + 2eg eg2 + 5eg2 eg2
= 2 + 4eg + 10eg2 + 2 + 5eg
= 4 + 9eg + 11eg2 .

In general we can compute that

(λ1 + µeg + νeg2 ) · (λ0 1 + µ0 eg + ν 0 eg2 )

= (λλ0 + µν 0 + νµ0 )1 + (λµ0 + µλ0 + νν 0 )eg + (λν 0 + µµ0 + νλ0 )eg2 .

2.7. Matrix algebras

The set of all n × n matrices with entries in K forms an algebra Mn (K). For a ∈ Mn (K)
we write a = (apq ), where apq ∈ K and 1 ≤ p, q ≤ n. Addition, scalar multiplication and matrix
multiplication are as usual:
X
(a + λb)pq = apq + λbpq , (ab)pq = apr brq .
r

The unit is the identity matrix, denoted 1 or I, or 1n or In if we want to emphasise the size of the
matrices.
The elementary matrices Epq have a 1 in the (p, q)-th place and 0s elsewhere. They form a
basis for Mn (K), so this algebra has dimension n2 . The multiplication is then given by

Epq Ers = δqr Eps .

(Here δij is the Kronecker delta, which equals 1 if i = j and 0 otherwise.)

More generally, if A is an algebra, then the set of n × n matrices with elements in A forms an
algebra Mn (A). The addition and multiplication are given by the same formulae, but where we
now take apq ∈ A. If A has basis {ei | i ∈ I} and structure coefficients ckij , then Mn (A) has basis
the matrices ei Epq , having the single non-zero entry ei in position (p, q), and multiplication
X
ei Epq · ej Ers = ckij δqr ek Eps .
k

Note that if A is finite dimensional, then dim Mn (A) = n2 dim A.

 2.9. TEMPERLEY-LIEB ALGEBRAS 14

2.8. Endomorphism algebras

More abstractly let V be a vector space. Recall that, after fixing a basis for V , we can represent
every linear map f : V → V as a matrix. It is sometimes convenient not to have to choose a basis, in
which case we consider the set of all such linear maps, called endomorphisms and denoted End(V ).
This is again a vector space, with addition and scalar multiplication given by

(f + λg)(v) := f (v) + λg(v) for all v ∈ V,

where f, g ∈ End(V ) and λ ∈ K.

Composition of linear maps now defines a bilinear multiplication on End(V ), so

(f g)(v) = f (g(v)) for f, g ∈ V and v ∈ V.

This is an algebra with unit the identity map idV (v) = v. (It is associative since composition of
functions is always associative.)

2.9. Temperley-Lieb algebras

Another interesting way of defining algebras is to index the basis elements by certain types of
diagrams and then to describe the multiplication in terms of these diagrams. Such algebras are
referred to as diagram algebras (although this is not a well defined term).
The Temperley-Lieb algebra T Ln (δ) for n ≥ 1 and δ ∈ K has basis indexed by planar diagrams
having two rows of n dots, one above the other, connected by n non-intersecting curves. Two such
diagrams are considered equal if the same vertices are connected.
To define the multiplication ab of two basis elements, we first stack the diagram for a on top of
that for b, and then concatenate the curves. The resulting diagram may contain some loops which
we must remove, and we multiply by the scalar δ for each such loop that we remove.
This is again easiest to understand once we have seen an explicit example. For the algebra
T L3 (δ) we have the following diagrams

1 u1 u2 p q

We have written the corresponding basis elements under the diagrams, so T L3 (δ) has basis
{1, u1 , u2 , p, q}.
To compute the product u21 we take two copies of the appropriate diagram, stacked one above
the other, then join curves and remove loops:

Since we had to remove one loop, we deduce that u21 = δu1 .

 2.10. DIRECT PRODUCT OF ALGEBRAS 15

Similarly u1 u2 = p, since

It is hopefully clear that the basis element 1 is indeed a unit for the algebra. In general the
unit is the basis element indexed by the diagram given by joining the dots vertically, so the i-th
dot on the top row is joined to the i-th dot on the bottom row.
We can also use the diagrams to see that the multiplication is associative; for, if we take three
diagrams stacked one on top of the other, then joining the curves of the top two diagrams, and then
joining them with the third is the same as joining the bottom two together, and then joining with
the top diagram. In fact, both operations agree with simply joining all three diagrams together in
one go. This proves that the multiplication is associative when restricted to three basis elements,
and hence is associative.
In general we define ui ∈ T Ln (δ) for 1 ≤ i < n to be the diagram so we have joined togther
the i-th and (i + 1)-st dots on the top row, and the same on the bottom row; all other dots on the
top row are joined to their counterparts on the bottom row.
1 i − 1i i + 1i + 2 n

1 i − 1i i + 1i + 2 n

Then the following relations always hold

u2i = δui , ui ui±1 ui = ui and ui uj = uj ui if |i − j| > 1.

In fact, in a certain sense (which we will make precise later), these relations are sufficient to
completely determine the Temperley-Lieb algebra. (See Exercise Sheet 1, Tutorial Problem 4.)

Remark 2.9.1. The Temperley-Lieb algebra was invented to study Statistical Mechanics (see
for example Paul Martin’s homepage). It is now also important in Knot Theory, and Vaughan
Jones won a Fields Medal in 1990 for his work in this area.

2.10. Direct Product of Algebras

Let A and B be algebras. Since they are a priori vector spaces, we may form their Cartesian
product, or direct product, A × B and this is again a vector space, with addition and scalar
multiplication given by
µ(a, b) + λ(a0 , b0 ) = (µa + λa0 , µb + λb0 ).
We give A × B the structure of an algebra via the following multiplication

(a, b)(a0 , b0 ) = (aa0 , bb0 ).

It is easy to check that this is associative and bilinear. Moreover the multiplication is unital, with
unit 1 = (1A , 1B ).
 CHAPTER 3

Homomorphisms and Subalgebras

3.1. Homomorphisms

Definition 3.1.1. A map f : A → B between algebras is an (algebra) homomorphism if

(a) f is a linear map.
(b) f (aa0 ) = f (a)f (a0 ) for all a, a0 ∈ A.
(c) f (1A ) = 1B .
In other words, f is a linear map which respects the multiplication and preserves the unit.

Note that if (b) holds, then condition (c) is equivalent to:

(c’) There is some a ∈ A with f (a) = 1B .
For, if f (a) = 1B , then

1B = f (a) = f (a1A ) = f (a)f (1A ) = 1B f (1A ) = f (1A ).

If f : A → B and g : B → C are algebra homomorphisms, then so too is their composition

gf : A → C.
An isomorphism is a homomorphism f : A → B which has an inverse, so there is a homomor-
phism g : B → A with both gf = idA and f g = idB . It is a useful fact that f is an isomorphism if
and only if it is a bijection.
If there is an isomorphism A → B, we say that A and B are isomorphic and write A ∼
= B.

Example 3.1.2. Recall that we can view C as an algebra over Q. Given any element z ∈ C
there is a Q-algebra homomorphism Q[X] → C sending X n 7→ z n . This is called the evaluation
map and is denoted evz .

Example 3.1.3. Recall that both C and H are algebras over R. Then the R-linear map C → H
sending x + yi ∈ C to x + yi ∈ H is an R-algebra homomorphism. In fact, if p ∈ H is any pure
quaternion of absolute value 1, then there is an algebra homomorphism C → H such that i 7→ p.

Example 3.1.4. Let A and B be algebras, and consider their direct product A × B. The
projection map
πA : A × B → A, (a, b) 7→ a

is then a surjective algebra homomorphism, and similarly for the projection map πB : A × B → B.
Note, however, that the inclusion map

ιA : A → A × B, a 7→ (a, 0)

is not an algebra homomorphism.

16
 3.2. SUBALGEBRAS 17

3.2. Subalgebras

Definition 3.2.1. A subalgebra B of an algebra A, written B ≤ A, is a vector subspace closed

under multiplication and containing the unit:
(a) B is a subspace.
(b) bb0 ∈ B for all b, b0 ∈ B.
(c) 1A ∈ B.

It follows that, using the induced multiplication, B is an algebra in its own right.

Lemma 3.2.2. If f : B → A is an algebra homomorphism, then its image Im(f ) is a subal-

gebra of A. Conversely, if B ≤ A is a subalgebra, then the inclusion map B ,→ A is an algebra
homomorphism.

In other words, subalgebras are the same as images of algebra homomorphisms.

Proof. Let f : B → A be an algebra homomorphism. Since f is a linear map, we know

that its image is a subspace. It contains the identity since f (1B ) = 1A , and it is closed under
multiplication since f (a)f (a0 ) = f (aa0 ).
Conversely, let B ≤ A be a subalgebra. Then the inclusion map B ,→ A is the restriction to
B of the identity map idA : A → A, and hence is an algebra homomorphism. 

Example 3.2.3. If K is a field and A a K-algebra, then there is a unique algebra homomor-
phism K → A, sometimes called the structure map. This sends the scalar λ ∈ K to the element
λ1A ∈ A.
For example, if A = Mn (K), then the structure map sends λ to the diagonal matrix λIn .

Example 3.2.4. The set of upper-triangular matrices

( ! )
x y
U= ∈ M2 (K) x, y, z ∈ K
0 z

is a subalgebra of M2 (K). For

! ! !
x y x0 y0 x + λx0 y + λy 0
+λ = ∈U
0 z 0 z0 0 z + λz 0
! ! !
x y x0 y0 xx0 xy 0 + yz 0
= ∈U
0 z 0 z0 0 zz 0
!
1 0
∈U
0 1

Two more examples of subalgebras of M2 (K) are

( ! ) ( ! )
x y x 0
B= x, y ∈ K and C = x, z ∈ K .
0 x 0 z

In fact, B and C are both subalgebras of U .

 3.2. SUBALGEBRAS 18

Example 3.2.5. Let A ≤ Mm (K) and B ≤ Mn (K) be subalgebras. Then the direct product
A × B is isomorphic to the subalgebra of Mm+n (K) consisting of block-diagonal matrices of the
form !
a 0
∈ Mm+n (K), a ∈ Mm (K) and b ∈ Mn (K).
0 b

Lemma 3.2.6. Let A be an algebra with basis {ei | i ∈ I} and structure coefficients
X
ei ej = ckij ek .
k

Then an algebra B is isomorphic to A if and only if B has a basis {fi | i ∈ I} with the same
structure coefficients:
X
fi fj = ckij fk .
k

Proof. If θ : A → B is an algebra homomorphism, then defining fi = θ(ei ) gives

X
X k X
fi fj = θ(ei )θ(ej ) = θ(ei ej ) = θ ckij ek = cij θ(ek ) = ckij fk .
k k k

So in particular, if θ is an isomorphism, then we have there is a basis for B with the same structure
coefficients as A.
k
P
Conversely, given basis elements fi ∈ B satisfying fi fj = k cij fk , let θ : A → B be the linear
map sending ei to fi . Then
X X X
ckij fk = ckij θ(ek ) = θ ckij ek = θ(ei ej ),


θ(ei )θ(ej ) = fi fj =
k k k
0 0 0
so by bilinearity θ(a)θ(a ) = θ(aa ) for all a, a ∈ A.
Now θ satisfies (a) and (b) and is bijective, so also satisfies (c’). Hence it is a bijective algebra
homomorphism, so an isomorphism. 

Example 3.2.7. The algebra A with basis {e, f } and multiplication

e2 = e, f 2 = f, ef = f e = 0

is isomorphic to the algebra B with basis {1, u} and multiplication u2 = u.

For, A has unit 1 = e + f , so has basis {1, e} with e2 = e.
Alternatively, B has basis
E := u, F = 1 − u,

and these satisfy

E 2 = u2 = u = E, EF = u(1 − u) = u − u2 = 0,
F 2 = (1 − u)2 = 1 − 2u + u2 = 1 − u = F, F E = (1 − u)u = 0.
∼
Lemma 3.2.8. The vector space isomorphism Θ : EndK (K n ) −
→ Mn (K) sending a linear map
to its matrix, is an algebra isomorphism.

Proof. Recall that the matrix (aij ) of f ∈ EndK (K n ) is determined by f (ej ) =

P
i aij ei .
Clearly Θ(id) = I, so we just need to check that Θ respects the multiplication.
 3.3. INTERSECTIONS AND GENERATING SETS 19

Suppose Θ(f ) = a = (aij ) and Θ(g) = b = (bij ). Then

X
X
(f g)(ej ) = f (g(ej )) = f bpj ep = bpj f (ep )
p p
X X X

= bpj aip ei = aip bpj ei .
i,p i p
P

Hence Θ(f g) is the matrix p aip bpj = ab, so Θ(f g) = Θ(f )Θ(g). 
∼
Lemma 3.2.9. A vector space isomorphism θ : V −
→ W induces an algebra isomorphism
∼
Θ : EndK (V ) −
→ EndK (W ), Θ(f ) := θf θ−1 .

Proof. We note that Θ is linear, since

Θ(f + λg) = θ(f + λg)θ−1 = θf θ−1 + λθgθ−1 = Θ(f ) + λΘ(g).

Moreover
Θ(f )Θ(g) = (θf θ−1 )(θgθ−1 ) = θf gθ−1 = Θ(f g)
and
Θ(idV ) = θidV θ−1 = θθ−1 = idW ,
so that Θ respects the multiplication and preserves the unit. Hence Θ is an algebra homomorphism,
and it is an isomorphism since it clearly has inverse

Θ−1 : EndK (W ) → EndK (V ), h 7→ θ−1 hθ. 

As a special case, let V be a finite-dimensional vector space. Choosing a basis {e1 , . . . , en } for
V is equivalent to choosing a vector space isomorphism V → K n , which then induces an algebra
∼
isomorphism EndK (V ) − → EndK (K n ) ∼= Mn (K). This is just the map sending a linear map to its
matrix with respect to the basis {e1 , . . . , en }.

3.3. Intersections and generating sets

Lemma 3.3.1. Let S and T be subalgebras of an algebra A. Then the vector space intersection
S ∩ T is again a subalgebra of A.

Proof. We know that S ∩ T is a subspace of A, and clearly 1 ∈ S ∩ T . If x, y ∈ S ∩ T , then

x, y ∈ S so xy ∈ S, and similarly x, y ∈ T so xy ∈ T . Thus xy ∈ S ∩ T proving that S ∩ T is a
subalgebra. 
T
More generally, if Sj is a collection of subalgebras, then j Sj is again a subalgebra. In partic-
ular, it is now possible to define the smallest subalgebra containing any subset X ⊂ A — we just
consider the intersection of all such subalgebras containing X. We call this the subalgebra generated
by X and it is denoted hXi. The elements of this subalgebra are all finite linear combinations of
products of elements from X, so finite sums of things of the form

λx1 x2 · · · xr with λ ∈ K and xi ∈ X.

In particular, we say that X is a generating set for the algebra A provided that the only subalgebra
containing X is A itself.

Example 3.3.2. (1) C is generated as an R-algebra by i. (As i2 = −1.)

 3.3. INTERSECTIONS AND GENERATING SETS 20

(2) H is generated as an R-algebra by i, j. For, k = ij and i2 = −1 = j 2 .

(3) K is generated as a K-algebra by the empty set.
(4) M2 (K) is generated by E12 , E21 . For, E11 = E12 E21 and E22 = E21 E12 .
(5) The polynomial algebra K[X1 , . . . , Xn ] is generated by X1 , . . . , Xn .
(6) The Temperley-Lieb algebra T L3 (δ) is generated by u1 and u2 . For, it has basis {1, u1 , u2 , p, q},
and p = u1 u2 and q = u2 u1 .
In fact T Ln (δ) is generated by u1 , . . . , un−1 , though this is not obvious.
(7) The algebra of upper-triangular matrices Un ≤ Mn (K) is generated by the elementary
matrices Eii and Eii+1 for 1 ≤ i < n.

Example 3.3.3. What is the subalgebra A of M3 (K) generated by

 
1 1 1
 
0 1 1?
a :=  
0 0 1
It is the subspace of M3 (K) spanned by {1, a, a2 , a3 , . . .}. Setting b = a − 1 we see that A is
also spanned by {1, b, b2 , . . .}, so is generated by b. Now
     
1 0 0 0 1 1 0 0 1
2
b3 = 0.
     
1 = 0 1 0 , b = 0 0 1 , b = 0
     0 ,
0
0 0 1 0 0 0 0 0 0
Thus the subalgebra generated by a (or b) is
  
 x y z

 

 
A= 
 0 x  x, y, z ∈ K
y

.

 0 0 
x 
 CHAPTER 4

Ideals

4.1. Ideals — Definition and Example

Let θ : A → B be a homomorphism of algebras. We have just seen that the image Im(θ) is
a subalgebra of B, but what about its kernel? Provided that B is not the zero algebra we have
θ(1A ) = 1B 6= 0. Hence 1A 6∈ Ker(θ) and so Ker(θ) cannot be a subalgebra of A.
On the other hand, Ker(θ) is closed under multiplication by any element of A. For, if x ∈
Ker(θ) and a ∈ A, then

θ(ax) = θ(a)θ(x) = 0 and θ(xa) = θ(x)θ(a) = 0,

so ax, xa ∈ Ker(θ).

Definition 4.1.1. A (two-sided) ideal I of an algebra A, denoted I CA, is a subset I satisfying

(a) I is a vector subspace of A.
(b) ax ∈ I for all a ∈ A and x ∈ I.
(c) xa ∈ I for all a ∈ A and x ∈ I.

Remark 4.1.2. (1) I = {0} (the zero ideal) and I = A (the unit ideal) are always ideals
of A.
(2) There is also the notion of a left ideal, for which one only demands (a) and (b), and a
right ideal, satisfying only (a) and (c). If A is commutative, then these are all the same.
(3) If θ : A → B is an algebra homomorphism, then Ker(θ) is an ideal in A.

Example 4.1.3. (1) The set of polynomials having zero constant term forms an ideal of
K[X].
(2) Let A and B be algebras, and consider their direct product A × B. We saw earlier that
the projection map πA : A × B → A, (a, b) 7→ a, is a surjective algebra homomorphism.
Its kernel is {(0, b) | b ∈ B}, which is an ideal of A × B.
(3) More generally let I C A and J C B ideals. Then the set

I × J = {(x, y) | x ∈ I, y ∈ J}

is an ideal of A × B. For, it is easily seen to be a vector subspace, so we just need to

check that it is closed under multiplication by elements of A × B. Let (a, b) ∈ A × B and
(x, y) ∈ I × J. Since ax, xa ∈ I and by, yb ∈ J, we deduce that (a, b)(x, y) = (ax, by) and
(x, y)(a, b) = (xa, yb) are both in I × J as required.
We will see in the exercises that every ideal of A × B is of this form.
(4) Let A be an algebra and I C A an ideal. Then the set

Mn (I) = {(xij ) | xij ∈ I}

21
 4.2. SUMS AND INTERSECTIONS OF IDEALS 22

is an ideal of Mn (A), and in fact every ideal of Mn (A) is of this form.

( !)
x y
Example 4.1.4. Let U = be the algebra of upper-triangular matrices.
0 z
( !)
x 0
(1) Is I := an ideal?
0 0
! !
1 0 1 1
No, since ∈
/ I. (It is only a left ideal.)
0 0 0 1
( !)
0 y
(2) Is J := an ideal?
0 0
! ! !
a b 0 y 0 ay
Yes. It is a subspace, satisfies (b) since = ∈ Y , and
0 c 0 0 0 0
similarly satisfies (c).

Lemma 4.1.5. If an ideal contains 1, or any invertible element, then it is the unit ideal. In
particular, the only ideals in a division algebra are the zero ideal and the unit ideal.

Proof. If I contains an invertible element b, then it also contains b−1 b = 1, and hence also
a1 = a for any a ∈ A. 

4.2. Sums and intersections of ideals

Lemma 4.2.1. Let I and J be ideals of an algebra A. Then the vector space constructions

I +J and I ∩J

are again ideals of A.

Proof. We know that I + J and I ∩ J are both subspaces of A. Take x ∈ I, y ∈ J and a ∈ A.

Then
ax, xa ∈ I and ay, ya ∈ J,
so that
a(x + y) = ax + ay ∈ I + J and (x + y)a = xa + ya ∈ I + J.
Hence I + J is an ideal.
Similarly, if x ∈ I ∩ J and a ∈ A, then

ax, xa ∈ I and ay, ya ∈ J, so ax, xa ∈ I ∩ J.

Hence I ∩ J is an ideal. 
P T
More generally, if Ij is a collection of ideals, then j Ij and j Ij are again ideals. In particular,
it is now possible to define the smallest ideal containing any subset X ⊂ A — we just consider the
intersection of all such ideals containing X. We call this the ideal generated by X, and denote it
by (X). The elements of (X) are all possible finite linear combinations of the form

a1 x1 b1 + · · · + an xn bn with ai , bi ∈ A and xi ∈ X.

As a special case we have the principal ideal (x), an ideal generated by one element.
 4.3. PRODUCTS OF IDEALS 23

NB: please don’t get the two notations confused. We use angle brackets hXi for the subalgebra
generated from the set X and round brackets for the ideal generated by X. In general these two
sets are very different.
We remark that these constructions also work for left (and right) ideals.

Example 4.2.2. (1) Let U be the algebra of upper-triangular 3 × 3 matrices. What is

(E22 ), the ideal generated by E22 ?
Recall that Eij Epq = δjp Eiq . Since U has basis Eij for i ≤ j, we see that (E22 ) is
spanned by the elements Eij E22 Epq = δ2j Ei2 Epq = δ2j δ2p Eiq for all i ≤ j and p ≤ q.
The only non-zero elements occur when j = p = 2, and hence i ∈ {1, 2} and q ∈ {2, 3}.
Thus (E22 ) is spanned by the four elements E12 , E13 , E22 , E23 , so
  
 0 b c

 

 
(E22 ) = 0 d e b, c, d, e ∈ K .


 

 0 0 0 

(2) Let A = M3 (K). What is (E22 )?

This time A has basis Eij for all i, j. Thus (E22 ) contains Ei2 E22 E2j = Eij , so it
equals the whole of A.
In fact, the zero ideal and the unit ideal are the only ideals of Mn (K).
(The story is different for Mn (B) where B is an algebra.)
(3) It is an important fact that every ideal of K[X] is principal. In fact, each non-zero ideal
can be written uniquely as (f ) for some monic polynomial f (so having leading coefficient
1).

4.3. Products of ideals

Definition 4.3.1. Let I and J be ideals of A. We define the ideal IJ to be the smallest ideal
containing the set {xy | x ∈ I, y ∈ J}. Since ax ∈ I for all a ∈ A and x ∈ I, and similarly yb ∈ J
for all b ∈ A and y ∈ J, we see that every element of IJ can be written as a finite sum of products
xy for x ∈ I and y ∈ J.

Clearly IJ ⊆ I ∩ J, but in general this inclusion is strict.

We now observe that this product is associative. For, if I,J,L are ideals of A, then (IJ)L is the
smallest ideal containing all products (xy)z for x ∈ I, y ∈ J and z ∈ L. Since the multiplication
in A is associative we have (xy)z = x(yz), and hence (IJ)L = I(JL).
More generally, given a finite set of ideals I1 , . . . , In of A, we can define their product inductively
as I1 I2 · · · In = (I1 · · · In−1 )In .
As special cases we observe that AI = I = IA and 0I = 0 = I0 for all ideals I E A.

Lemma 4.3.2. Let I, J and L be ideals. Then I(J + L) = IJ + IL and (I + J)L = IL + JL.

Proof. Since J, L ⊆ J + L we must have IJ, IL ⊆ I(J + L), and hence IJ + IL ⊆ I(J + L).
Conversely, I(J + L) is the smallest ideal containing all elements of the form x(y + z) for x ∈ I,
y ∈ J and z ∈ L. Since x(y + z) = xy + xz ∈ IJ + IL we have I(J + L) ⊆ IJ + IL.
The proof for (I + J)L = IL + JL is analogous. 
 CHAPTER 5

Quotient Algebras

5.1. Definition of Quotient Algebras

Recall that if V is a vector space and U ≤ V a subspace, then we can form the quotient vector
space V /U . This has elements the cosets v + U := {v + u | u ∈ U } for v ∈ V , so that v + U = v 0 + U
if and only if v − v 0 ∈ U . The addition and scalar multiplication are given by

(v + U ) + (v 0 + U ) = (v + v 0 ) + U and λ(v + U ) = λv + U for v, v 0 ∈ V, λ ∈ K.

Moreover the natural map π : V → V /U , v 7→ v + U , is a surjective linear map with kernel U .

One important result is then the Factor Lemma, which states that if f : V → W is a linear
map of vector spaces such that U ⊂ Ker(f ), then there is a unique linear map f¯: V /U → W such
that f = f¯π. In other words, the map f factors through the quotient space V /U .
In this chapter we show how these ideas can be extended to algebras and algebra homomor-
phisms.

Lemma 5.1.1. If I is an ideal in an algebra A, then the vector space quotient A/I becomes an
algebra via the multiplication

(A/I) × (A/I) → A/I, (a + I, b + I) 7→ ab + I.

Clearly this has unit 1 + I. Moreover the natural map π : A → A/I is a surjective algebra homo-
morphism with kernel I.
In other words, ideals are the same as kernels of homomorphisms.

Proof. We have that the addition and scalar multiplication are well-defined as this is the
same as the quotient vector space construction — but as revision of this construction we will verify
the addition and scalar multiplication.
Suppose a + I = a0 + I and b + I = b0 + I. By definition, this implies that a − a0 and b − b0 ∈ I.
We check the addition is well defined. Now

a + I + b + I := (a + b) + I and a0 + I + b0 + I := (a0 + b0 ) + I.

Also
a + b − (a0 + b0 ) = a − a0 + b − b0 ∈ I

as a − a0 and b − b0 ∈ I and I is an ideal. Thus (a + b) + I = (a0 + b0 ) + I if a + I = a0 + I and

b + I = b0 + I.
We check scalar multiplication is well defined. Now let λ ∈ K.

λ(a + I) := (λa) + I and λ(a0 + I) := (λa0 ) + I.

24
 5.2. THE FACTOR LEMMA 25

Also
λa + λa0 = λ(a − a0 ) ∈ I
as a − a0 ∈ I and I is an ideal. Thus (λa) + I = (λa0 ) + I if a + I = a0 + I.
We now check that the multiplication is well-defined;

(a + I)(b + I) := (ab) + I and (a0 + I)(b0 + I) := (a0 b0 ) + I.

Also
ab − (a0 b0 ) = (a − a0 )b + a0 (b − b0 ) ∈ I
as (a − a0 )b ∈ I and a0 (b − b0 ) ∈ I as I is an ideal. Thus (ab) + I = (a0 b0 ) + I if a + I = a0 + I and
b + I = b0 + I.
The product is clearly associative and bilinear (since it is induced from the product in A), and
1 + I is a unit. Thus A/I is an algebra.
The natural map is easily seen to be an algebra homomorphism, since it is surjective π(ab) =
ab + I = (a + I)(b + I) = π(a)π(b) and π(1) = 1 + I is the identity element of A/I. 

5.2. The Factor Lemma

Lemma 5.2.1 (The Factor Lemma). Let f : A → B be a homomorphism of algebras, and

I ⊆ Ker(f ) an ideal of A. Then there is a unique algebra homomorphism f¯: A/I → B such that
f = f¯π, where π : A → A/I; a 7→ a + I is the natural quotient map.

Proof. Using the Factor Lemma for vector spaces B.8.5, we know that there is a unique linear
map f¯: A/I → B such that f = f¯π. Using this we have f¯(a + I) = f¯π(a) = f (a). We therefore
only need to check that f¯ is an algebra homomorphism.
Let a, a0 ∈ A then

f¯(a + I)f¯(a0 + I) = f (a)f (a0 ) = f (aa0 ) = f¯(aa0 + I),

and similarly f¯(1 + I) = f (1) = 1.

Although strictly speaking this is not required, we will also prove that f¯ is well defined so that
we see where the assumption I ⊆ Ker f is used.
We need to prove that defining f¯(a + I) = f (a) does not depend on the choice of coset
representative a.
Suppose a + I = a0 + I so a − a0 ∈ I. then

f (a) − f (a0 ) = f (a − a0 ) = 0

where the first equality follows as f is linear and the second follows as a − a0 ∈ I and I ⊆ Ker(f ).
Thus it is unambiguous to to define f¯(a + I) = f (a). 

Example 5.2.2. The quotient algebra R[x]/(x2 + 1) is isomorphic to C.

Let I = (x2 + 1). Consider

x3 + 3x2 + x − 3 + I = (x + 3)(x2 + 1) − 6 + I = −6 + I

as (x + 3)(x2 + 1) ∈ I. We effectively treat any element of I as zero in the quotient.

In general, using the division algorithm for p(x) ∈ R[x] we may write

p(x) = q(x)(x2 + 1) + ax + b
 5.2. THE FACTOR LEMMA 26

for some q(x) ∈ R[x] and a, b ∈ R. Thus

p(x) + I = ax + b + I,

as p(x) − (ax + b) = q(x)(x2 + 1) ∈ I.

By treating the elements of I as zero, we are effectively making x2 = −1. I.e. we are really
evaluating each polynomial at x = i where i ∈ C.
Formally: consider the evaluation map

θ = evi : R[x] → C, x 7→ i.

This is clearly surjective, since a + bx 7→ a + bi. Moreover θ(x2 + 1) = i2 + 1 = 0, so x2 + 1 ∈ Ker(θ).

Since Ker(θ) is an ideal, we deduce that (x2 + 1) ⊆ Ker(θ). By the Factor Lemma we now have
an induced algebra homomorphism θ̄ : R[x]/(x2 + 1) → C, and this is still surjective.
The easiest way to complete the proof is to use a dimension argument. We know that C is a
two-dimensional real vector space. Since θ̄ is surjective, we must have that R[x]/(x2 + 1) is at least
two dimensional. On the other hand, we’ve seen that every element of R[x]/(x2 + 1) is of the form
ax + b + (x2 + 1). Thus R[x]/(x2 + 1) is spanned by 1 + (x2 + 1) and x + (x2 + 1), so is at most
two dimensional.
We deduce that R[x]/(x2 + 1) is two dimensional, and hence that θ̄ is an isomorphism.

Example 5.2.3. Let U be the algebra of upper-triangular 3 × 3 matrices

 
x u w
 
0 y v 
 
0 0 z

and let I E U be the ideal of matrices of shape

 
0 0 w
 
.
0 0 0

0 0 0

Now U has basis Eij for 1 ≤ i ≤ j ≤ 3 and I is spanned by E13 . Thus U/I has basis

f1 := E11 + I, f2 = E22 + I, f3 = E33 + I, f4 = E12 + I, f5 = E23 + I,

and multiplication table

f1 f2 f3 f4 f5
f1 f1 0 0 f4 0
f2 0 f2 0 0 f5
f3 0 0 f3 0 0
f4 0 f4 0 0 0
f5 0 0 f5 0 0
For example,
f4 f5 = (E12 + I)(E23 + I) = E12 E23 + I = E13 + I = 0 + I,
since E13 ∈ I.
Note that E11 + E22 + E33 is the unit in U , so f1 + f2 + f3 is the unit in U/I.
 5.2. THE FACTOR LEMMA 27

Example 5.2.4. The augmentation ideal of the group algebra KG is the kernel I of the
homomorphism
 : KG → K, eg 7→ 1 for all g ∈ G.
P P
The ideal I consists of all elements g λg eg such that g λg = 0, so has basis eg −1 for g ∈ G\{1}.
Observe that KG/I ∼ = K.

Theorem 5.2.5 (First Isomorphism Theorem). Let θ : A → B be a homomorphism of algebras.

Then the induced algebra homomorphism

θ̄ : A/ Ker(θ) → Im(θ)

is an isomorphism.

Proof. The map θ̄ exists by the Factor Lemma for algebras, 5.2.1. On the other hand, we
know it is an isomorphism of vector spaces, so it is an algebra isomorphism.
To prove it’s an isomorphism of vector spaces (without using theorem B.8.6) observe that θ̄ is
onto. Also observe that it is injective as

θ̄(a+Ker θ) = θ̄(a0 +Ker θ) ⇔ θ(a) = θ(a0 ) ⇔ θ(a−a0 ) = 0 ⇔ a−a0 ∈ Ker θ ⇔ a+Ker θ = a0 +Ker θ.

Thus θ̄ is a bijection and hence an isomorphism. 

 CHAPTER 6

Presentations of Algebras

So far we have defined algebras in terms of a basis and structure coefficients. Similarly, to
define an algebra homomorphism, we need to give the images of the basis elements (so give a linear
map) and then check that this preserves the unit and respects the multiplication. This is fine for
small examples, but becomes very inefficient for larger algebras.
Compare with the situation for finite groups. You have probably only seen groups as subgroups
of symmetric groups, or else via their multiplication tables. It is then quite laborious to check
whether a given map defines a group homomorphism. Now imagine that you want to work with
the Monster Group, having roughly 8 × 1053 elements.
The purpose of this section is to introduce a more compact way of exhibiting an algebra A.
The basic idea is to write A as a quotient algebra B/I, where B is a larger algebra, but easier to
understand, and I C B is some ideal.
Consider for example the quotient algebra given in Example 5.2.3. This algebra is quite
awkward to work with, but if we regard it as the quotient U/I, then it becomes much easier to
think about.
The type of algebra B one chooses may depend on the particular situation, but one choice
which always works is when we take B to be a free algebra. In this case the description of A as a
quotient of a free algebra is called a presentation by generators and relations. Such a description
is particularly useful when constructing homomorphisms between algebras, since we only need to
give the images of the generators and then check that the relations still hold. This is the content
of the Proposition 6.2.5.
A word of warning: it is not always easy to transfer between these two descriptions, so it is
often best to know both a basis for the algebra as well as a presentation in terms of generators and
relations. Also, as for bases, generators are not unique.

6.1. Free Algebras

Example 6.1.1. There are two ways to think about the group algebra KG where G is the
cyclic group of order 4. If we set G = {1, g, g 2 , g 3 } where g 4 = 1 then KG has basis {1, eg , eg2 , eg3 }.
An algebra that is isomorphic to KG is the following algebra

KhXi/(X 4 − 1).

Here KhXi is the free algebra on a single letter X. It consists of all linear combinations of the
powers of X. I.e.
KhXi = Span{1, X, X 2 , X 3 , . . .}.
In this case, KhXi = K[X] the polynomial algebra in a single variable X. The word free is meant
to indicate that there are no relations on the elements.
28
 6.1. FREE ALGEBRAS 29

As we have seen quotienting out by the ideal (X 4 −1) effectively makes X 4 = 1 in the quotient,
the same relation that appears in the group algebra KG. We call KhXi/(X 4 − 1) a presentation
of KG.

Definition 6.1.2. Let S be a set. We think of S as an ‘alphabet’, so a ‘word’ in S is a finite

ordered list X1 X2 · · · Xr with Xi ∈ S, and the length of a word X1 X2 · · · Xr is r. We also include
the ‘empty word’, of length 0 and denote it by 1. Note that we can identify S with the set of words
of length 1.
The free algebra KhSi is the vector space with basis all the words in the alphabet S and
multiplication given by concatenation of words, so

X1 X2 · · · Xr · Y1 Y2 · · · Ys = X1 X2 · · · Xr Y1 Y2 · · · Ys .

This multiplication is clearly associative with unit the empty word 1. As usual, we often abbreviate
XX by X 2 , and XXX by X 3 , and so on.
If S = {Xi | i ∈ I}, then we also write KhXi | i ∈ Ii for KhSi.

Example 6.1.3. (1) If S = ∅, then there is only one possible word, so Kh∅i = K.
(2) If S = {X}, then KhXi has basis 1, X, X 2 , X 3 , . . . and multiplication X m · X n = X m+n .
Thus KhSi = K[X], the polynomial algebra.
(3) If S = {X, Y }, then KhX, Y i has basis (in order of increasing length)

1
X, Y
X 2 , XY, Y X, Y 2
X 3 , X 2 Y, XY X, Y X 2 , XY 2 , Y XY, Y 2 X, Y 3
...

and as an example of the multiplication we have

(2X + XY + 3Y 2 X)(4X − Y X)
= 2X(4X − Y X) + XY (4X − Y X) + 3Y 2 X(4X − Y X)
= 8X 2 − 2XY X + 4XY X − XY 2 X + 12Y 2 X 2 − 3Y 2 XY X
= 8X 2 + 2XY X − XY 2 X + 12Y 2 X 2 − 3Y 2 XY X.

Note that XY 6= Y X, so KhX, Y i is not isomorphic to the polynomial algebra K[X, Y ].

In fact we have
K[X, Y ] ∼
= KhX, Y i/(XY − Y X)

Proposition 6.1.4 (The Universal Property). Let S be a set and A an algebra. Given a map
f : S → A there exists a unique algebra homomorphism θ : KhSi → A extending f , so θ(X) = f (X)
for all X ∈ S.
In other words there is a bijection between algebra homomorphisms KhSi → A and maps
S → A.
Another random document with
no related content on Scribd:
The Project Gutenberg eBook of Satu sydämestä
ja auringosta
This ebook is for the use of anyone anywhere in the United States
and most other parts of the world at no cost and with almost no
restrictions whatsoever. You may copy it, give it away or re-use it
under the terms of the Project Gutenberg License included with this
ebook or online at www.gutenberg.org. If you are not located in the
United States, you will have to check the laws of the country where
you are located before using this eBook.

Title: Satu sydämestä ja auringosta

Author: Elina Vaara

Release date: October 27, 2023 [eBook #71967]

Language: Finnish

Original publication: Porvoo: WSOY, 1925

Credits: Tuula Temonen

*** START OF THE PROJECT GUTENBERG EBOOK SATU

SYDÄMESTÄ JA AURINGOSTA ***
SATU SYDÄMESTÄ JA AURINGOSTA

Kirj.

Elina Vaara

Porvoossa, Werner Söderström Oy, 1925.

SISÄLLYS:
SATU SYDÄMESTÄ JA AURINGOSTA

Kaukaiset metsät.
Satu sydämestä ja auringosta.
Mennyt suvi.
Lumikkosydän.
Järvellä.
Hiiden hovi.
Syysromanssi.
Saaren soittaja.

VILLIVIINI
 Karnevaalihuume.
Käsky — kielto..
Kuolleet.
Judithin tuska.
Villiviini.

HILJAISIA AKORDEJA

Puutumus.
Hiljainen huone.
Akordi.
Hiljaisimmat.
Hartaus.
Keväthämärä.

LAULU KAUKAISESTA RAKKAUDESTA

Blayn prinssi.
Trubaduurilaulu.
Aavelinna.
Suleika.
Yö keitaalla.
Oi Sulamith, on päivät hämärtyneet.
Netkron sadusta.

PUISTOKUJA

Kevät.
Kultaiset pallot.
Kellastuneesta vihkosta.
Katkenneiden pilarien kaupunki.
Tähtisumua.
Uneksijat.
Puistokuja.
SATU SYDÄMESTÄ JA AURINGOSTA.

KAUKAISET METSÄT.

Keskiyöllä ikävään ma havaan, sateen lauluun ikkunani

avaan. Sydän valittaa. Siellä sananjalat viherjäiset, sinikellot,
punakämmekkäiset vavisten nyt vartoo kuolemaa.

Siellä syksy mustan tulvan lailla hautaa metsät, joissa

huolta vailla häipyi suven kuut. Haavanlehdet maahan
varisevat, raskasmielisinä huokailevat tuuleen tummat,
rakkaat havupuut.

Kuule, sadeyö, kun kuiskaan sulle: katoovaisuudesta voitko

mulle laulun virittää, jok' on sinipunainen ja musta, jok' on
täynnä murheen huumausta — laulun, jonka voi vain
nyyhkyttää!

SATU SYDÄMESTÄ JA AURINGOSTA.

Oli kerran kuninkaantalo, missä yö oli päivinkin. Ei tulvinut
koskaan valo sen synkkiin saleihin. Ah, tummat uutimet häilyi
siell' edessä ikkunain; mut auringonikävä säilyi yhä sielussa
prinsessain.

Pian hämyyn hautakuorin kaks heistä kaipaus vei, mut

prinsessa hennoin ja nuorin hän kuolla tahtonut ei. Miten
verhojen takaa ehtoin surunsairain kuuntelikaan hän huminaa
syreenilehtoin, jotka varisti kukkasiaan!

Kuu venheellään kun sousi yli torninhuippujen, kuva

nuorukaisen nousi hänen uniinsa ihmeellinen. Sen
silkinkuivaa tukkaa ois nauraen hyväillyt ja tuhannen
syreeninkukkaa hänen tielleen kylvänyt…

Ja prinsessa pimeän linnan ei enää viipyä voi: unen kutsu

pohjalla rinnan kuin kiihkeä viulu soi. Hän pakeni linnasta
salaa kuin pääsky, pesästään joka paisuvin siivin halaa sini-
ilmaan värisevään.

Näyn ihanan eessä aivan sydän nuori hurmaantui: laill'

yrtteihin peittyvän laivan maa säteiden virrassa ui. Utupilvissä
taivahalla lepäs aurinko hehkuen kuin valkeiden vuorten alla
ois järvi kultainen.

Ja kun nukahti matkalainen puun himmeän varjoihin, niin

untensa nuorukainen hänet herätti suudelmin. Käden
prinsessan käteen hän liitti, vei kauas maailmaan, ja hymyä
auringon riitti, he kunne kulkivatkaan.

Moni kaupunki kaunis, suuri ilon, loistonsa tarjolle toi. Oli

puutarhat kukassa juuri ja suihkulähteet soi! Mut prinsessa
 päivien mennen pian tunsi sydämessään, ett' uupuu valohon
ennen kuin yöhön pimeään.

Yhä prinsessa kulkee, kulkee all' ihanan auringon… Mut

illoin, kun silmät hän sulkee, sydän luotaan kaukana on. Nyt
raukein siivin se halaa taa tummain uudinten, ja se pimeään
linnahan palaa kuin pesäänsä pääskynen.

MENNYT SUVI.

Kuin käsi lämmin, suven muisto mun sydäntäni hyväilee,

vaikk' auringoton on jo puisto ja kaikki ilo pakenee, pois ilo
pakenee.

Ma lehtimetsäin huminahan taas haaveissani unohdun ja

heinään silkinruskeahan, mi peitti kukkulani mun, ah,
kukkulani mun.

Ja kaislarannat, vedet tummat on laulelmiksi muuttuneet:

ne on niin kaukaiset ja kummat kuin paimenhuilun säveleet,
kuin huilun säveleet.

En konsaan lämpöä ma vailla nyt lähde syksyn kylmyyteen,

kun vaeltaa sain päivän mailla ja nähdä onnen sydämeen, ah,
onnen sydämeen.

LUMIKKO-SYDÄN.
 Sun nuoruutesi kuninkaallinen on kristalliseen arkkuun
suljettuna, ja parvi kääpiöiden itkee, palvoo sua vuorten yössä
vahakynttilöin.

Kuu läpikuultavalle haudalles

niin säälivänä kylvää säteitänsä,
ja huhuilussa juron huuhkajankin
on outo, murheellinen vienous.

Kentiesi löytää lumon-alaisen

hän, jonka kyyneleet on kyllin kuumat
ja kyllin hartaat vangin lunnahiksi:
tuo tulimieli poika kuninkaan.

Mut ehkä valitsee hän toisen tien puutarhojen ja linnain

houkuttaissa. Ja silloin arkussas, oi sydän, sydän, sa
iankaikkisesti maata saat.

JÄRVELLÄ.

Oi järvi, laineitten koti ja kalain kimmelsuomuisten, ma

kuuntelen läikyntääsi ja laulua hyräilen.

On järven pohjalla linna,

niin kaunis simpukkalinna, ai!
Sen ammoin mahtava Ahti
on rakentanut kai.
 Mut Ahdin maineikas suku
— kuin moni ylhäinen suku muu,
joka liian on vanha ja hieno —
jo sammuu, rappeutuu.

Vain pieni, kalpea prinssi

nyt linnanpuistossa leikkiä lyö,
rapukarjoja paimentaapi
ja itkee, kun tummuu yö.

Oi järvi, laineitten koti ja kalain kimmelsuomuisten, ma

kuuntelen läikyntääsi ja laulua hyräilen.

HIIDEN HOVI.

Vuorivirta kuohuu alla ikkunain, lyövät tornin seinään laineet

loiskahtain takana tunturin viiden. Nuku, nuori kukka ripsin
kasteisin! Sull' on kultasänky silkkiuutimin linnassa mahtavan
hiiden.

Peikko lemmenkade nukkua ei voi; kiilusilmin väijyy,

niinkuin vartioi saituri timanttijyvää. Ryöstö-impi huokaa
taljavuoteellaan, unohtaa ei saata edes unissaan
onnettomuuttansa syvää.

Vuorivirta kuohuu alla ikkunain, lyövät tornin seinään laineet

loiskahtain takana tunturin viiden. Nuku, nuori kukka. ripsin
kasteisin! Sull' on kultasänky silkkiuutimin linnassa mahtavan
hiiden.
SYYSROMANSSI.

Sinimustia astereita syksyn tummuvat päivät on. Sinimustia

astereita kasvoi puistossa kartanon.

Pieni kreivitär unelmoiden

niitä poimi, ne murheen toi.
Alla vanhojen vaahteroiden
vielä nyyhkytys illoin soi.

Pieni kreivitär kuninkaalle

antoi lempensä ainoan.
Vieras juhlittu kaukomaalle
läks, ei kuulunut palaavan.

Läpi sydämes kävi miekka,

pieni kreivitär, tiennyt et:
luvuttomat kuin rannan hiekka
kuninkaiden on rakkaudet!

Sinimustia astereita syksyn tummuvat päivät on. Sinimustia

astereita kasvoi puistossa kartanon.

SAAREN SOITTAJA.

Kun vetten sini vaaleana väikkyi,

kun kukat pihlajan ja syreenin
tän saaren hautas lumeen tuoksuvaiseen,
ma tänne silloin matkasin.
 Nyt täällä viipynyt oon liian kauan.
Mun viuluni on mykäks vaiennut.
Siks poveen pusertunut tuska, hurma
on järkeni mun sokaissut.

Kuin syksyn liekehtivä koivu tuolla

veen miilunmustaan syliin kumartuu,
niin hullu rakkaus ja synkkä murhe
mun sydämessäin sekaantuu.

Ja halveksivat katseet kylätieltä

mua peikkoin lailla seuraa uniin yön.
Mut vannon, että kerran vielä soitan
tai viulun sirpaleiksi lyön!

———

Nyt kielet viritän, nyt kostan teille. Ja tää on kosto halvan

soittajan: te mielet ynseät, kuin lapset pahat ma teidät
soinnuin taivutan.

Ma sallin tähtein sataa ylitsenne

kuin sädehtiväin tuliperhosten.
Ma kutsun ihastuksen huulillenne
ja sytän kaipuin sydämen.

Sa tyttö kaunis, kylmyytesi hautaan

ma sävelien villiruusuihin!
Jääkirkkaat silmäs pysähtyvät minuun
ja hämärtyvät kyynelin.
 Mun viuluni, ah, itkee, nauraa, kuohuu,
kun outo, kuuma onni siinä soi.
— Mut riemusta käy raskahaksi pääni.
En viipyä ma enää voi.

Ma syöksyn venhevalkamahan: vihdoin

pois olen vapaa täältä lähtemään.
Jää hyvästi, oi haavojeni saari!
Nyt viime kerran sinut nään.
VILLIVIINI.

KARNEVAALIHUUME.

On murhetta maailma täynnä, tie musta ja kuolleet puut. Minä

sentään nauraen kuljen, minä niinkuin ihmiset muut. Ja
hehkuvat keltaiset lyhdyt.

Mut sydämeni ei naura,

sitä turhaan pyytelen.
Tänä yönä mun kuolema saartaa,
vaikk' elää himoitsen.
Ja hehkuvat keltaiset lyhdyt.

Oi, ruusuja, okaita halaan,

en rauhaa hautausmaan!
Sydän mulla on ihanan nuori,
kun toipuu se tuskastaan.
Ja hehkuvat keltaiset lyhdyt.

Minä lingota liekkejä tahdon

ja häätää kuoleman pois.
Minä tahdon riemuita, laulaa,
 kuin onnekas rintani ois.
Ja hehkuvat keltaiset lyhdyt.

On maailma murhetta täynnä, tie musta ja kuolleet puut.

Minä sentään nauraen kuljen kuin ihmisnaamiot muut. Ja
hehkuvat keltaiset lyhdyt.

KÄSKY — KIELTO.

— Tuskan, tuskan hinnalla on riemus ostetut.

Kyyneleitä, verta oon ma niistä maksanut.
Siispä syöksy nautintoon, sen mehu kuuma juo!
— Ah, en voi, en voi, en voi! on turha käsky tuo.

— Sydän, sydän, heikkousko siitä estää sun?

— En ma pelkää muuta kuin ett' ilo tahraa mun.
Kyyneleet ja veri ovat kiirastulenain:
Jos ma tulta pakenen, oon maata, maata Vain!

KUOLLEET.

Onnelliset, onnelliset kuolleet, jotka alla mätäneväin lehtein

itse verkkaan maaksi lahoatte! Leponne on niinkuin musta
rauha saderaskaan, marraskuisen illan.

Onnelliset, onnelliset kuolleet!

Tärisytä intohimo ykskään
 untanne ei majesteetillista,
piikit elon orjantappuroiden
enää otsaanne ei viillä veriin.

Onnelliset, onnelliset kuolleet! Keihäänhaava syvä

kyljessänne voideltu on unhon narduksella, kädet kuumat,
lävistämät naulain, kiedotut on unhon käärinliinaan.

Onnelliset, onnelliset kuolleet, jotka alla mätäneväin lehtein

itse verkkaan maaksi lahoatte! Tuhatkerroin autuaammat
sentään ootte, jollei ylösnousemuksen valo koskaan kohtaa
silmiänne.

JUDITHIN TUSKA.

Olen kamppaillut kuoleman painin kera tuskan enkelin. Ma

syömeni rinnasta revin ja maahan tallasin. Ah, silmin
mustenevin sun povees painoin tikarin ja pakenin yön
pimeihin niin kauas kuin ma jaksoin voimin murtuvin.

Nyt ammottava, avoin verihaava mun rintani on. Miks

lähteelle laahautuisin? Ma tunnen kohtalon: vaikk' aalloissa
Jordanin uisin, ois janoni lieska sammumaton, ah,
sammumaton kuin auringon, joka sadevedestä kuiviks juo ojat
aavikon!

Käy raju tuskannyyhke kuin puistatus ruumiissain. Maan

tomussa ma ryömin nyt rukoillen rakkauttain. Ah, kasvoin
kyynelten syömin sun eessäs, armas, haavoissain ma
 polvillain nyt makaan vain ja kuolen, nimes kuumeisilla
huulillain.

VILLIVIINI.

Ruskeata, purppuraa, hiilen hehku kaiken yllä! Huumaukses

tunnen kyllä, sydämen se sairaaks saa.

Joka syksy köynnöstyt

sydämeni ympärille
tummin liekkein, tuhlaat sille
hiuduttavat hyväilyt.

Ruskeata, purppuraa, hiilen hehku kaiken yllä! Kerran tules,

tiedän kyllä, tuhkaksi mun kuluttaa.
HILJAISIA AKORDEJA

PUUTUMUS.

Valkea pumpuliköynnös välissä ikkunaruutuin.

Kaksi kättä pöydän verkaa vasten puutuin.

Harmaita sauhulintuja suhisee vaiheilla taivaan.

Kaksi valjua kättä, vihityt huoleen ja vaivaan.

Mustaa liejua kadulla vettä valuvalla.

Ihmisen kädet vihreän varjostimen alla.

HILJAINEN HUONE.

Valko-ompeleiset uutimet unteloina nuokkuu renkaissaan.

Vihreätä, hämynpehmeätä iltalamppu heittää valoaan.

Kuva kehyksessään hymyilee.

Hellät, uskolliset silmät nään.