SINGULAR VALUE APPROXIMATIONS IN CRYPTOGRAPHY

by
Joshua Boone
Bachelor of Science in Mathematics, Southern Illinois University, 2006

A Research Report
Submitted in Partial Fulfillment of the Requirements for the
Master of Science Degree

Department of Mathematics
in the Graduate School
Southern Illinois University Carbondale
July, 2008

RESEARCH REPORT APPROVAL

SINGULAR VALUE APPROXIMATION IN CRYPTOGRAMS

By
Joshua Boone

A Research Report Submitted in Partial

Fulfillment of the Requirements
for the Degree of
Master of Science
in the field of Mathematics

Approved by:
Dr. Robert Fitzgerald, Chair
Dr. Jerzy Kocik
Dr. David W. Kammler

Graduate School
Southern Illinois University Carbondale
June 30, 2008

AN ABSTRACT OF THE RESEARCH REPORT OF

Joshua Boone, for the Master Of Science degree in Mathematics, presented on

August 2, 2008, at Southern Illinois University Carbondale.
Singular Value Approximations in Cryptography
MAJOR PROFESSOR: Dr. Robert Fitzgerald
The English Language is a vfc language, that is, vowels follow consonants
a majority of the time. We will use this fact, along with some creative linear
algebra, to separate the letters of a cryptogram into three major groups: vowels,
consonants, and neuter. First, an adjacency matrix is created, which shows how
many times a given letter follows another. Then the Singular Value Analysis of
this matrix begins, which will use the Perron-Frobenius Theorem. After some sign
comparisons of certain singular vectors are done, the letters are separated into the
aforementioned groups. This process is easily implemented as a computer program.
After partitioning, it is relatively simple to decode the message with some elementary
frequency analysis.

ii

TABLE OF CONTENTS

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

ii

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

iv

1 Prerequisites

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.1

Elementary Cryptography . . . . . . . . . . . . . . . . . . . . . . .

1.2

The Singular Value Decompostion . . . . . . . . . . . . . . . . . . .

1.3

The Perron-Frobenius Theorem . . . . . . . . . . . . . . . . . . . .

2 Consonant/Vowel Partitioning via Singular Value Analysis . . . . . . . .

2.1

The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2.2

The SVD and its rank-two approximation . . . . . . . . . . . . . .

3 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

12

4 Java and MATLAB Code used for computation . . . . . . . . . . . . . .

19

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

26

Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

26

Vita . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

30

iii

LIST OF FIGURES

3.1

The Adjacency Matrix A . . . . . . . . . . . . . . . . . . . . . . . . . .

12

3.2

The second singular vectors. . . . . . . . . . . . . . . . . . . . . . . . .

13

3.3

The partitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

14

3.4

The Adjacency Matrix A . . . . . . . . . . . . . . . . . . . . . . . . . .

15

3.5

The second singular vectors. . . . . . . . . . . . . . . . . . . . . . . . .

16

3.6

The partitions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

16

3.7

The Letter Frequency of the English Language. [7] . . . . . . . . . . .

18

iv

CHAPTER 1
PREREQUISITES

1.1

ELEMENTARY CRYPTOGRAPHY
In simplest terms, cryptography can be thought of as the study of hiding

information from unwanted viewers. One of the most ancient applications of cryptography is transmission of secret messages. A cryptogram is an example of this
application.
Definition. A substitution cryptogram is a message in which each letter is replaced
by another in a one-to-one manner. The method by which this is done is called the
substitution cipher. Whitespace and punctuation are removed entirely.
Definition. A shift cipher is a substitution cipher Cs in which each letter is mapped
to another letter by a constant shift 1 s 26. If we assign a number to each
letter, beginning with 1 for A, 2 for B, all the way to 26 for Z, then we can write
Cs () = + s (mod 26), where 1 26 corresponds to a letter in our alphabet.

Example 1.1.1. Austin wants to send Bob the message Kenny is great using
the C2 shift cipher. Notice that the letter Y will be mapped to A. The encrypted
message would be MGPPAKUITGCV. To decrypt, Bob would simply use the
inverse (here, C24 ) cipher on the message.

Breaking the substitution cipher of a cryptogram is usually done by analysis

of the relative frequency of certain letters, as well as certain combinations of letters.
This process can be simplified by first separating the letters into vowels and consonants, which we explore in this paper. In order to understand the mathematics
behind this idea, we must first introduce a special matrix.
1

Definition. The adjacency matrix of a substitution cryptogram is a 26 26 matrix

with integer entries such that the i, j entry is the number of times the j-th letter
follows the i-th letter in the cryptogram.
Notice that the sum of the components of the i-th row equals the sum of the
components of the i-th column, which equals the number of times the i-th letter
occurs in the cryptogram. Since a combination of two letters is called a digraph,
this matrix is also sometimes called the digraph frequency matrix.

1.2

THE SINGULAR VALUE DECOMPOSTION

In this section, we introduce some definitions and results from linear algebra.

This material is taken from [2].

Definition. The hermitian conjugate, or adjoint, of an m n matrix A (written
A ) is the n m matrix whose i, j entry is the complex conjugate of the j, i entry
of A. If A = A , A is called hermitian.
Definition. A square matrix Q Cnn is unitary if Q = Q1 .
Definition. Let A Cmn . A singular value decomposition (SVD) of A is a factorization
A = U W
where
U Cmm is unitary,
W Cnn is unitary, and
Rmn is diagonal with entries i such that 1 2 . . . k 0, where
k = min(m, n)

The components of this factorization have special names. The columns of U

and W are called the left and right singular vectors, respectively. The i are called
the singular values.
Of course, this factorization is only useful when it exists. The following theorem proves that every complex matrix has a singular value decomposition. In order
to understand the theorem, we must first understand the vector and matrix norms.
Definition. The 2-norm of a column vector v Cm is the length of v, given by
kvk2 =

vv

where v is the adjoint of v.

Definition. The induced 2-norm of a matrix A Cmn is the smallest number
C C such that the inequality
kAvk2 C kvk2
holds for every v Cn .
Theorem 1.2.1. Every matrix A Cmn has an SVD.
Proof. Set 1 = kAk2 . We can assume 1 > 0 since otherwise A = 0 and the
decompostion 0I0 (where 0 is the zero matrix) is an SVD. Since {Az : kzk2 = 1}
is compact, kAzk2 achieves its maximum. So, there must be vectors w1 Cn and
u1 Cm with kw1 k2 = ku1 k2 = 1 and Aw1 = 1 u1 . Consider any extensions of w1 to
an orthonormal basis {wj } of Cn and of u1 to an orthonormal basis {uj } of Cm , and
let U1 and W1 denote the unitary matrices with columns uj and wj , respectively.
Then

U1 AW1

1 x
,
=S=

0 B

(1.1)

where 0 is a column vector of all zeros with dimension m 1, x is a row vector of

dimension n 1, and B has dimensions (m 1) (n 1). Furthermore,







x

1
1
1

12 + x x = (12 + x x)1/2 ,




0 B

x
x



2

(1.2)

which implies kSk2 (12 + x x)1/2 . Since U1 and W1 are unitary, we can see that
kSk2 = kAk2 = 1 , so x = 0.

If n = 1 and m = 1, were done. Otherwise, B

describes the action of A on the subspace orthogonal to w1 . By induction, B has

an SVD: B = U2 2 W2 . Since the product of diagonal matrices is diagonal, it is
obvious that

0
1 0 1 0 1
W1

A = U1

0 W2
0 2
0 U2

(1.3)

is an SVD of A.
An important piece of information is the rank of the matrix A, which can be
determined as follows:
Theorem 1.2.2. The rank of A is r, the number of nonzero singular values.
Proof. The rank of a diagonal matrix is the number of nonzero entries. Since, in
A = U W , U and W are of full rank, the rank of A is the rank of , which is
r.
For the sake of this particular paper, it is more useful to consider the SVD as
a sum of rank-one matrices, as follows:
Theorem 1.2.3. The matrix A Cmn can be written as the sum of r rank-one
matrices, where r is the rank of A. In other words,
A=

r
X

j uj wj .

j=1

(1.4)

Proof. If we consider the matrix as the sum of r matrices i , where i =

diag(0, ..., 0, i , 0, ..., 0), then we can see this follows directly from (1.1).
We will use a rank two approximation of the SVD, by taking only the first two
terms of the sum.

1.3

THE PERRON-FROBENIUS THEOREM

One last piece of information we need is the behavior of positive matrices. The

Perron-Frobenius theorem gives us all the information we need. We will assume the
following topological fact in order to prove this theorem. This material is taken
from [3].
Theorem 1.3.1. (Brouwers Fixed Point Theorem). Let S be a nonempty, closed,
bounded, convex set in Rn . Let f : S S be a continuous map. Then there exists
an x S such that f (x) = x.
The following theorem is one of the most important theorems in matrix theory. There are many versions of this theorem, but the following simple version is
sufficient. Here, the notation A > 0 means the matrix A has only positive entries.
Theorem 1.3.2. (Perron-Frobenius Theorem). Let A > 0 be an n n matrix.
Then:
(i) Ay = 0 y for some 0 > 0, y > 0
(ii) The eigenvalue 0 is maximal in modulus among all the eigenvalues of A. In
other words, for any eigenvaue e of A, |e| 0 .
(iii) The eigenvalue 0 is geometrically simple, so that any two eigenvectors corresponding to 0 are linearly dependent.
(iv) Any positive eigenvector of A (of any eigenvalue) is a scalar multiple of y.

Proof.
(i) Let
(
S=

(x1 , x2 , ..., xn )T :

xi = 1 and xi 0 for all i .

Define the map f : S S as follows:

(
)1
X
f (x) =
(Ax)i
Ax
i

Here (Ax)i is the i-th coordinate of Ax. Since A > 0, x S implies that
the vector Ax is nonzero and the map f is well-defined. It is easy to check
that f is continuous and maps S into S. By Brouwers Fixed Point Theorem,
f (y) = y for some y S. Thus
(
X

)1
(Ay)i

Ay = y.

If we set

i (Ay)i

= 0 , then 0 > 0 and Ay = 0 y. Since A > 0, it follows

that y > 0.
(ii) If we apply (i) to AT , we see that AT z = 1 z for some 1 > 0, z > 0. Now

1 y T z = y T AT z = 0 y T z,
and since y T z > 0, we have 0 = 1 . Thus AT z = 0 z. Let Au = eu for
some eigenvalue e C. Let u+ be defined by u+ = (|u1 |, |u2 |, ..., |un |)T , where
u = (u1 , u2 , ..., un )T . Without loss of generality, let u+ be a probability vector.
We have

X
j



X



aij |uj | 
aij uj  = |eui | = |e||ui |.


j

Thus Au+ |e|u+ . If we premultiply this by z T , we get |e| 0

6

(iii) Suppose Av = 0 v for some real, nonzero vector v. We want to show that v is
a scalar multiple of y. If v and y are linearly independent, then R such
that y v is a nonnegative, nonzero vector with at least one zero coordinate.
Since
A(y v) = 0 (y v),
we see that y v is an eigenvector of A. However, since A > 0, any nonnegative eigenvector must be postive. This is a contradiction. Thus, v is a
scalar multiple of y. By considering the real and imaginary part separately,
we can show that any complex eigenvector of A corresponding to 0 is a scalar
multiple of y.
(iv) Suppose Au = eu for u > 0. Then ez T u = z T Au = z T u. Since z T u > 0, we
have e = 0 . The rest follows from (iii).

CHAPTER 2
CONSONANT/VOWEL PARTITIONING VIA SINGULAR VALUE
ANALYSIS

2.1

THE PROBLEM
Now we are ready to discuss the problem of breaking a given cryptogram. The

idea that follows is from [4]. Our entire scheme relies on the fact that, in the English
language, vowels usually follow consonants and vice versa. This is why English is
referred to as a vfc language. Formally, if
number of consonant-vowel pairs
number of vowel-vowel pairs
<
number of vowels
number of consonants

(2.1)

for a given language, it is referred to as vfc.

First, we will create two column vectors, v and c, where

vi = 1 if the i-th letter of the alphabet is a vowel, and 0 otherwise

ci = 1 if the i-th letter of the alphabet is a consonant, and 0 otherwise.

If A is the adjacency matrix for the cryptogram, we can write (2.1) as

v T Av
cT Av
<
v T A(v + c)
cT A(v + c)
which can be simplified to
(v T Av)(cT Ac) (v T Ac)(cT Av) < 0

(2.2)

So, our problem is simply this: Given A, find a v and c so that (2.2) holds.
8

2.2

THE SVD AND ITS RANK-TWO APPROXIMATION

We can provide the SVD of a matrix as the sum of rank-one matrices, as seen

in (1.6). Here again is the equation:

A=

r
X

j uj wjT .

j=1

Since we are considering only real matrices, the hermitian conjugate is simply
the transpose, as shown above. By the normalization of uj and wj , we see that the
matrices uj wjT have the same norm. In other words, if the j decrease rather rapidly,
we can find a decent approximation to A by taking only a few terms. A truncation
after 1 term is a rank-one approximation, after 2 is a rank-two approximation, et
cetera. The rank-two approximation is the first approximation that has information
about corresponding pairs of letters (rank-one assumes A is symmetric). We will
take this approximation:
A A2 = 1 u1 w1T + 2 u2 w2T

(2.3)

and attempt to solve our problem assuming A = A2 , that is, that A has rank equal
to 2. In an attempt to solve this problem, it is necessary to create a third partition,
that we will refer as neuter, which holds letters that are neither vowels or consonants.
Although there are no neuter letters in the English alphabet, consonants like h that
follow consonants regularly in normal language would fall into this category (as a
consonant exhibiting vowel-like behavior).
We will use the partitions on the following page, where the ui2 and wi2 are
the entries of the u2 and w2 , respectively. Theorem 2.2.1 will show that these
partitions are the solution to our problem.

vi =

1 if ui2 > 0 and wi2 < 0

ci =

1 if ui2 < 0 and wi2 > 0

ni =

otherwise

otherwise

1 if sign(ui2 ) = sign(wi2 )

otherwise

Theorem 2.2.1. (Moler and Morrison, 1983). Let A = A2 be a nonnegative ranktwo matrix with SVD expansion as in (2.3). Let v and c be defined as above. Then

D = (v T Av)(cT Ac) (v T Ac)(cT Av) < 0

is satisfied.
Proof. Since A is nonnegative, it follows from Theorem 1.3.2 that u1 and w1 have
nonnegative components. Substituting (2.3) into (2.2), we get
D = (v T (1 u1 w1T + 2 u2 w2T )v)(cT (1 u1 w1T + 2 u2 w2T )c)
(v T (1 u1 w1T + 2 u2 w2T )c)(cT (1 u1 w1T + 2 u2 w2T )v)
= 1 2 (v T u1 w1T vcT u2 w2T c + v T u2 w2T vcT u1 w1T c
v T u1 w1T ccT u2 w2T v v T u2 w2T ccT u1 w1T v).
Of all these inner products, only w2T v and cT u2 are negative. Since each of the first
two terms in the parentheses contain exactly one of these, the third contains both,
and the fourth contains neither, and since 1 2 is positive, it follows that D < 0.

10

In the next chapter, two examples of the process are shown. Since hand
calculation would be tedious, two computer programs were written to calculate the
adjacency matrix A, the SVD of A, and the resulting partitions. The code can be
found in Chapter 4. The text from the two examples can be found in the Appendix.

11

CHAPTER 3
EXAMPLES

Rather than encrypt our first example, we will do the analysis on the verbatim
text.
Example 3.0.1. For our first example, we will use George W. Bushs 9/11 Address
after the September 11, 2001 attacks, a 2765-character piece of text taken from [5].
Using the Java program written for this purpose, we see that the adjacency matrix
A has the form:
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
W
X
Y
Z

A
1
0
20
21
33
5
4
22
6
0
0
9
7
11
1
1
0
12
18
17
2
2
11
0
5
0

B
1
0
0
2
6
0
0
2
3
0
1
4
1
2
3
0
0
4
7
3
0
0
0
0
0
0

C
16
0
0
1
13
1
0
0
16
0
1
0
1
9
4
0
0
6
4
2
3
0
1
0
4
0

D
7
0
0
4
27
2
0
1
0
0
0
13
0
52
10
0
0
2
1
1
1
0
0
0
0
0

E
0
18
10
19
15
10
13
47
12
0
7
13
27
17
1
8
0
45
24
29
3
23
15
0
4
2

F
1
0
0
5
14
3
2
0
3
0
1
2
2
5
27
0
0
7
5
1
0
0
0
0
3
0

G
5
0
0
0
1
0
1
0
9
0
0
0
0
23
4
0
0
7
1
0
4
0
0
0
0
0

H
1
0
6
1
3
1
12
1
0
0
0
0
0
0
5
0
0
0
8
72
0
0
13
0
2
0

I
5
1
6
10
9
7
3
12
1
0
3
11
5
12
2
3
0
30
10
25
3
5
17
0
7
0

J
0
0
0
1
1
0
0
0
0
0
0
0
0
1
1
0
0
1
1
0
0
0
0
0
0
0

K
3
0
6
0
0
0
0
0
0
0
0
3
0
1
1
0
0
4
1
0
0
0
0
0
0
0

L
17
4
0
3
13
4
1
1
18
0
0
22
0
1
5
6
0
6
0
1
3
0
0
0
0
0

M
12
0
0
1
12
1
0
1
4
0
0
2
3
4
14
0
0
1
4
0
1
0
0
0
1
0

N
56
0
0
4
42
0
0
0
50
0
0
0
1
5
30
0
0
6
0
1
11
0
1
0
0
0

O
1
3
12
13
8
27
6
21
11
2
1
5
3
11
6
7
0
17
19
31
0
0
13
0
8
0

P
3
0
0
1
4
0
0
0
1
0
0
4
1
2
7
2
0
3
7
3
1
0
0
0
1
0

Figure 3.1. The Adjacency Matrix A

12

Q
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

R
17
2
1
2
56
9
5
3
8
0
0
2
0
1
39
7
0
10
1
7
33
0
0
0
1
0

S
11
0
0
8
48
1
2
0
16
0
3
0
4
17
10
2
0
13
13
10
11
0
0
0
0
0

T
28
0
14
16
15
4
7
12
23
0
2
6
2
24
6
3
0
12
33
16
6
0
0
0
4
0

U
2
9
5
1
2
6
0
2
0
4
0
0
3
1
36
0
1
2
2
6
0
0
0
0
0
0

V
6
0
0
0
14
0
0
0
3
0
0
2
0
0
4
0
0
1
0
0
0
0
0
0
0
0

W
4
0
0
7
20
0
0
0
1
0
0
2
0
4
9
1
0
5
10
2
0
0
0
0
7
0

X
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

Y
10
2
2
0
6
0
0
0
0
0
0
5
1
4
0
0
0
10
0
6
0
0
1
0
1
0

Z
0
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

By summing rows (or columns), we can see from the matrix that the letter X
does not occur, the letter Q occurs exactly once, and the letter Z occurs twice.
Also, the two-letter combination (or digraph) TH is the most common, occuring
72 times. Furthermore, since nearly half of the occurences of U in this document
exist in the OU digraph, there is a good chance that U will be seen as a neuter
element.
In order to determine the vowel/consonant partitions, we will take the SVD of
A and compare signs of the second singular vectors (second columns of U and W ).
This is consistent with Theorem 2.2.1, which assumes A is rank-two.
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
W
X
Y
Z

-0.36724192219341
0.07486813752176
0.13608384875342
0.12471268293278
-0.37151887148156
0.10065093619693
0.10415482681094
0.31538079432540
-0.22966398618755
0.00335511305787
0.03463788608883
0.06254095871470
0.13116502453518
0.02321115037834
-0.31529595108696
0.03145633741803
-0.00050067671627
0.22679560289022
0.19701568282605
0.47316940903535
-0.15515208162008
0.11554777987678
0.19052577389345
-0.00000000000000
0.05506079683650
0.00881517222704

0.15759583437214
0.00746555169787
-0.10964210237106
-0.10735461932021
0.48243204003143
-0.09033451147643
-0.03411358646954
0.34516586763380
0.24722916881021
-0.00105138892359
0.00652792414376
-0.11266008969954
-0.11083096452985
-0.51206052989804
0.29321990869506
-0.00941252098165
-0.00335518842500
-0.33980484123968
-0.16237591061664
0.02949728921350
-0.05480153612514
-0.08225315918080
-0.05854831435437
0.00000000000000
0.00374259054637
-0.00419650318511

Figure 3.2. The second singular vectors.

13

The letter X is absent in the cryptogram, but due to roundoff, the values in
the X row are nonzero. This problem (which would cause us to accidently consider
X a vowel at first glance) can be avoided by using rounding. Five decimal places
is sufficient. Comparing signs, we see that the following partitions are created.
Vowels: A E I O
Consonants: C D F G J L M N P R S V W Z
Neuter Elements: B H K Q T U X Y

Figure 3.3. The partitions.

Normally, some simple frequency analysis can be done to determine to which

letter each of the encrypted letters corresponds. For this example, however, no
encryption was done, so this step is unnecessary.

An example on an actual encrypted piece of text is helpful to fully understand

the method. The following example uses a simple shift cipher.
Example 3.0.2. For this example, we will use the shift cipher C2 on Dr. Martin
Luther Kings famous I Have a Dream speech, also from [5]. In other words,
we will replace each letter with the letter two places after it in the alphabet, then
replace each Y with A and each Z with B.
The 6,805-character text has the following adjacency matrix A.

14

15

A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
W
X
Y
Z

A
1
0
21
7
3
4
7
4
0
2
0
0
0
7
5
4
4
0
0
16
3
22
0
1
2
0

B
0
0
0
0
0
0
0
0
0
0
4
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

C
5
0
0
14
30
30
86
19
15
70
12
2
3
36
27
34
9
8
0
27
36
32
8
7
10
1

D
3
0
17
0
0
11
8
4
0
0
3
0
2
19
2
3
14
0
0
3
4
7
3
0
1
0

E
3
0
25
0
0
2
36
7
2
0
43
0
0
2
0
11
10
0
0
10
13
1
1
0
2
0

F
2
0
24
0
0
2
80
4
2
0
7
0
0
10
1
79
19
0
0
7
5
3
2
0
0
0

G
7
4
0
42
29
35
46
6
23
162
23
2
8
50
52
58
2
15
0
106
28
56
6
56
36
0

H
3
0
0
0
1
3
18
12
12
1
11
0
0
8
0
1
100
1
0
5
14
21
4
0
0
0

I
0
0
14
0
0
5
30
8
3
0
24
0
2
2
0
53
10
0
0
8
2
1
4
0
0
0

K
7
0
23
2
13
20
25
18
18
52
0
0
4
42
21
21
8
10
0
53
37
88
7
8
47
1

L
0
0
2
0
0
1
2
4
0
0
0
0
0
2
0
4
3
0
0
0
0
0
0
0
0
0

M
0
0
6
0
17
0
1
0
0
0
2
0
0
3
0
4
3
0
0
7
3
0
0
0
0
0

N
2
0
54
17
3
4
26
4
5
2
55
0
1
66
3
6
21
12
0
4
16
9
13
0
0
0

O
8
0
33
0
1
1
28
4
1
0
10
0
2
4
2
2
62
0
0
9
4
3
3
0
2
0

P
3
0
102
0
0
2
71
4
5
0
111
0
2
5
4
12
74
0
0
15
11
6
27
0
4
0

Q
27
0
1
5
29
32
27
25
17
24
33
3
3
26
13
46
13
8
0
59
62
102
0
4
10
0

Figure 3.4. The Adjacency Matrix A

J
4
0
3
0
21
3
20
4
20
2
8
0
0
3
0
3
3
3
0
4
21
202
4
0
29
0

R
1
0
7
0
0
1
13
4
1
2
10
0
0
2
3
1
15
12
0
2
11
2
7
0
0
0

S
0
0
0
0
0
0
3
0
0
0
0
0
0
0
0
2
0
0
0
1
0
0
0
0
0
0

T
1
0
29
8
9
26
100
48
25
2
15
0
1
2
11
1
45
15
0
4
1
23
29
0
2
0

U
8
0
35
0
2
26
70
3
1
3
78
0
2
12
5
31
19
3
0
21
28
14
36
0
1
0

V
18
0
93
0
4
19
89
24
6
24
64
0
12
16
19
63
50
4
0
20
73
21
18
0
3
2

W
0
0
5
8
5
7
4
8
5
4
0
11
0
2
10
4
55
2
6
11
11
14
0
0
0
1

X
2
0
25
1
0
0
26
0
0
0
11
0
0
2
0
1
4
0
0
2
0
2
0
0
0
0

Y
8
0
2
0
1
13
33
2
5
7
0
0
4
2
1
14
26
1
0
3
15
12
0
0
1
0

Z
0
0
0
0
0
0
3
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0

If we find the SVD of this matrix, we will get the following second singular
vectors (second columns of U and W ).
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
Y
U
V
W
X
Y
Z

0.00630771169603
0.00668519833513
-0.40400561554923
0.06541054530923
0.11678821960198
0.04843764908191
-0.29588468069332
-0.02559213553157
0.06227384136410
0.34438456003104
-0.34789041641652
0.00559967974476
-0.00129449686530
0.07284166722039
0.08602639860957
-0.02057750604381
-0.33313521273343
0.02266626545267
-0.00068397986930
0.20964621203852
0.02801489859255
0.51821656473647
-0.10836633438194
0.10856300934520
0.15248129307646
-0.00090887719170

0.02238194455629
-0.00535406661745
0.10366270317444
-0.03388032866720
-0.13807465649987
-0.15338425321626
0.43438317055276
-0.11129465254638
-0.09664453130702
0.39961406173167
0.27531989983701
-0.00919459618040
-0.00285456237121
-0.16666612273773
-0.16014004482120
-0.46345133351373
0.23507363726993
-0.04919654544566
-0.00276699614581
-0.17813573010846
-0.21673111412097
-0.29278897939887
-0.02962857438459
-0.08192771022439
-0.03225975525502
-0.00517072875066

Figure 3.5. The second singular vectors.

Comparing signs, we get the following partitions for the encrypted text:
Vowels: C G K Q
Consonants: B D E F I L N O R T U V X Y
Neuter Elements: A H J M P S W Z

Figure 3.6. The partitions.

16

To determine which of these vowels correspond to A, E, I, and O, we use

frequency analysis. Summing the columns of A, we see that the letter C occurs 521
times, the letter G occurs 852 times, K occurs 525 times, and Q occurs 569 times.
This suggests that G corresponds to E, the most common letter in the English
language (see Figure 3.7). We will use G
= e to denote this, where a lowercase letter
corresponds to a plaintext letter. The other values are all similar, so there is not
enough information yet present to decrypt the rest of the letters. However, looking
at the matrix A will help immensely. Notice first that the digraph GC occurs most
often out of the vowel digraphs beginning with G. This likely corresponds to the
vowel combination EA. Similarly, if C
= a, then likely K
= i, since the ai is the
most common vowel digraph beginning with a[3]. This leaves Q
= o, which makes
sense because it is the only vowel that occurs regularly as a repeated digraph oo
(besides ee of course). This is also a common way to differentiate e from other
vowels, since ee is the most common vowel digraph.
Once the vowels are found (u is easy to find because of the high occurence
of ou) the rest of the decryption process is straightforward. Once again we will
look at the diagonal of A, which gives repeated digraph frequency. Here, the highest
values belong, respectively, to GG, NN, UU and VV. We already assume that
G
= e, and a little knowledge in repeated digraph frequency (see [3] and [6]) tells
us that it is very likely that N
= l, U
= s, and V
= t. Once t is found, it is
easy to find h. Simple frequency analysis tells us that n, r, and d are the
most commonly-occurring letters left, so we can feasibly assume that P
= n, T
=
r, and F
= d. At this point, we have 13 of the 26 letters decrypted. The remaining
letters can be found by simply substituting known letters and creatively attempting
a trial-and-error method with a bit of known letter frequency.

17

Figure 3.7. The Letter Frequency of the English Language. [7]

18

CHAPTER 4
JAVA AND MATLAB CODE USED FOR COMPUTATION
This chapter includes all of the code used to create the adjacency matrix,
compute the SVD, and compare signs of the second left and right singular vectors
to give the resulting partitions. This Java program creates the adjacency matrix A.
package readmatrix;
import java.io.*;
//******************************************************************************
// ReadMatrix.java
// Author: Josh Boone
// Department of Mathematics
// Southern Illinois University Carbondale
//******************************************************************************
// This program creates a 26x26 adjacency matrix for a given input file. The
// matrix is printed in an easy-to read format, as well as Matlab/Maple form, to
// .txt files that are saved in the same location as the input file. The
// formatted matrix can then be directly pasted into Matlab/Maple for SVD
// analysis. This program works for raw data (plaintext files) as well as
// formatted cryptograms, due to the robustness of the removeSpace method.
//******************************************************************************
public class readMatrix {
public static void main(String[] args) {
BufferedReader input = null;
BufferedReader input2 = null;
PrintWriter output = null;
PrintWriter output_matlab = null;
PrintWriter output_maple = null;
PrintWriter fixed = null;
String s = null;
String s2 = null;
boolean spacer = false;
int[][] matrix = new int[27][27];
try {
input = new BufferedReader(new FileReader("in.txt"));
fixed = new PrintWriter(new FileWriter("fixed.txt"));
output_maple = new PrintWriter(new FileWriter("out_maple.txt"));
output_matlab = new PrintWriter(new FileWriter("out_matlab.txt"));
output = new PrintWriter(new FileWriter("out.txt"));
}
catch (IOException e) {
e.printStackTrace();
}
// This block of code removes the non-letter characters from the input
// file and saves the resulting string as "fixed.txt" in the same
// directory as the input file.
try{
String str = null;
while((str = input.readLine()) != null){
s = removeSpace(str); // see method at bottom of file.
s2 = encrypt(s, 0); // to encrypt a plaintext with a shift
19

// cipher, insert a number between 1 and 25

fixed.print(s2);
}
input.close();
fixed.close();
}
catch(IOException e){
e.printStackTrace();
}
// Now we read in the "fixed" version of the text.
try{
input2 = new BufferedReader(new FileReader("fixed.txt"));
}
catch(IOException e){
e.printStackTrace();
}
// This will determine the adjacency matrix.
try {
String str = null;
while ((str = input2.readLine()) != null) {
s = str;
for (int i = 0; i < s.length() - 1; i++) {
// The location of A is 1, B is 2, etc. in our 27x27 matrix.
// The 0th row and column are used only for instantiation
int r1 = 0;
int r2 = 0;
// The next two blocks of code determine the adjacency matrix
// by comparing every two successive letters.
if (s.charAt(i) == A) {
r1 = 1;
} else if (s.charAt(i) == B) {
r1 = 2;
} else if (s.charAt(i) == C) {
r1 = 3;
} else if (s.charAt(i) == D) {
r1 = 4;
} else if (s.charAt(i) == E) {
r1 = 5;
} else if (s.charAt(i) == F) {
r1 = 6;
} else if (s.charAt(i) == G) {
r1 = 7;
} else if (s.charAt(i) == H) {
r1 = 8;
} else if (s.charAt(i) == I) {
r1 = 9;
} else if (s.charAt(i) == J) {
r1 = 10;
} else if (s.charAt(i) == K) {
r1 = 11;
} else if (s.charAt(i) == L) {
r1 = 12;
} else if (s.charAt(i) == M) {
r1 = 13;
} else if (s.charAt(i) == N) {
r1 = 14;
} else if (s.charAt(i) == O) {
20

r1 = 15;
} else if (s.charAt(i) == P) {
r1 = 16;
} else if (s.charAt(i) == Q) {
r1 = 17;
} else if (s.charAt(i) == R) {
r1 = 18;
} else if (s.charAt(i) == S) {
r1 = 19;
} else if (s.charAt(i) == T) {
r1 = 20;
} else if (s.charAt(i) == U) {
r1 = 21;
} else if (s.charAt(i) == V) {
r1 = 22;
} else if (s.charAt(i) == W) {
r1 = 23;
} else if (s.charAt(i) == X) {
r1 = 24;
} else if (s.charAt(i) == Y) {
r1 = 25;
} else if (s.charAt(i) == Z) {
r1 = 26;
}
if (s.charAt(i + 1) == A) {
r2 = 1;
} else if (s.charAt(i + 1) == B)
r2 = 2;
} else if (s.charAt(i + 1) == C)
r2 = 3;
} else if (s.charAt(i + 1) == D)
r2 = 4;
} else if (s.charAt(i + 1) == E)
r2 = 5;
} else if (s.charAt(i + 1) == F)
r2 = 6;
} else if (s.charAt(i + 1) == G)
r2 = 7;
} else if (s.charAt(i + 1) == H)
r2 = 8;
} else if (s.charAt(i + 1) == I)
r2 = 9;
} else if (s.charAt(i + 1) == J)
r2 = 10;
} else if (s.charAt(i + 1) == K)
r2 = 11;
} else if (s.charAt(i + 1) == L)
r2 = 12;
} else if (s.charAt(i + 1) == M)
r2 = 13;
} else if (s.charAt(i + 1) == N)
r2 = 14;
} else if (s.charAt(i + 1) == O)
r2 = 15;
} else if (s.charAt(i + 1) == P)
r2 = 16;
} else if (s.charAt(i + 1) == Q)
r2 = 17;
} else if (s.charAt(i + 1) == R)
r2 = 18;
} else if (s.charAt(i + 1) == S)
r2 = 19;
21

{
{
{
{
{
{
{
{
{
{
{
{
{
{
{
{
{
{

} else if (s.charAt(i + 1) == T) {
r2 = 20;
} else if (s.charAt(i + 1) == U) {
r2 = 21;
} else if (s.charAt(i + 1) == V) {
r2 = 22;
} else if (s.charAt(i + 1) == W) {
r2 = 23;
} else if (s.charAt(i + 1) == X) {
r2 = 24;
} else if (s.charAt(i + 1) == Y) {
r2 = 25;
} else if (s.charAt(i + 1) == Z) {
r2 = 26;
}
matrix[r1][r2] = matrix[r1][r2] + 1; // increment matrix
}
}
// This prints to the out_matlab.txt file in Matlab matrix form.
output_matlab.print([);
for (int j = 1; j <= 26; j++) {
for (int k = 1; k <= 26; k++) {
output_matlab.print(matrix[j][k]);
if (k != 26) {
output_matlab.print( );
}
}
if (j != 26) {
output_matlab.print(;);
}
}
output_matlab.print(]);
// This prints to the out_maple.txt file in Maple matrix form.
output_maple.print([);
for (int j = 1; j <= 26; j++) {
for (int k = 1; k <= 26; k++) {
if (k == 1) {
output_maple.print([);
}
output_maple.print(matrix[j][k]);
if (k != 26) {
output_maple.print(,);
} else {
output_maple.print(]);
if (j != 26)
output_maple.print(,);
}
}
}
output_maple.print(]);
// This prints to the out.txt file, putting in "easy-to-read" form.
// First, we will determine how we need to format the matrix (with
// either two or three spaces between single-digit values).
for (int j = 1; j <= 26; j++) {
for (int k = 1; k <= 26; k++) {
if (matrix[j][k] > 99)
22

spacer = true;
}
}
// If any value of the matrix is greater than 99, we use the
// following block of code.
if (spacer == true) {
// First, we print the labels for the columns:
output.println(" A
B
C
D
E
F
G
H
I
J
K
"
+ "L
M
N
O
P
Q
R
S
T
U
V
"
+ "W
X
Y
Z");
for (int j = 1; j <= 26; j++) {
output.print((char) (j + 64)); // print labels for rows,
output.print(" ");
// using ASCII codes
for (int k = 1; k <= 26; k++) {
output.print(matrix[j][k]);
if (matrix[j][k] <= 9) {
output.print("
");
} else if (matrix[j][k] <= 99){
output.print(" ");
} else{
output.print( );
}
}
output.println();
output.println();
}
}
// Otherwise, we use the following piece of code.
if (spacer == false) {
// First, we print the labels for the columns:
output.println(" A B C D E F G H I J K L M N O "
+ "P Q R S T U V W X Y Z");
for (int j = 1; j <= 26; j++) {
output.print((char) (j + 64)); // print labels for rows,
output.print(" "); // using ASCII codes for the letters
for (int k = 1; k <= 26; k++) {
output.print(matrix[j][k]);
if (matrix[j][k] <= 9) {
output.print(" ");
} else
output.print( );
}
output.println();
}
}
output_matlab.close(); // close matlab output stream
output_maple.close(); // close maple output stream
output.close(); // close regular output stream
input2.close(); // close input stream (from fixed text)
}
catch (IOException e) {
e.printStackTrace();
}
23

}
// This method removes all non-letter characters from a raw text file and
// returns a new string containing only capital letters. For a cryptogram,
// it simply removes the spaces.
public static String removeSpace(String s) {
String str = "";
for (int i = 0; i < s.length(); i++) {
char ch = s.charAt(i);
if ((ch >= A && ch <= Z) || (ch >= a && ch <= z))
str = str + Character.toUpperCase(ch);
}
return str;
}
// This method is only used as an accessory method to provide shift-cipher
// encryption for examples.
public static String encrypt(String s, int i) {
String str = "";
if (i < 0)
System.err.print("Please input a value between 0 and 25.");
for (int j = 0; j < s.length(); j++) {
if ( (int)s.charAt(j) + i <= Z ){
int char_int = (int) s.charAt(j) + i;
char c = (char) char_int;
str = str + c;
}
else{
int char_int = (int) s.charAt(j) + i - 26;
char c = (char) char_int;
str = str + c;
}
}
return str;
}
}

24

This MATLAB program returns the singular values, second left and right
singular vectors, the frequency vector, and the partitions of the letters into vowels,
consonants, and neuter elements.
%thesis.m
%Josh Boone, Department of Mathematics
%Southern Illinois University Carbondale
%Coded for use in MATLAB, Mathworks Inc.
%
%Returns the singular values, the second right singular vector, the
%second left singular vector, the vowel, consonant, and neuter
%elements of the cryptogram, and the letter frequency vector
function thesis(A);
[U, S, W] = svd(A);
singular_values = diag(S)
second_left_singular_vector = U(:,2)
second_right_singular_vector = W(:,2)
letter_frequency_vector = sum(A)
l = roundn(second_left_singular_vector, -5);
r = roundn(second_right_singular_vector, -5);
k = max(size(A));
alpha = ABCDEFGHIJKLMNOPQRSTUVWXYZ;
disp(Vowels:);
for i=1:k
if sign(r(i)) ~= sign(l(i))
if r(i) > 0
disp(alpha(i));
end;
end;
end;
disp(Consonants:);
for i=1:k
if sign(r(i)) ~= sign(l(i))
if r(i) < 0
disp(alpha(i));
end;
end;
end;
disp(Neuter Elements:);
for i=1:k
if sign(r(i)) == sign(l(i))
disp(alpha(i));
end;
end;
Java is a trademark of Sun Microsystems, Inc.
MATLAB is a trademark of The MathWorks, Inc.
Maple is a trademark of Maplesoft, Inc.

25

REFERENCES

[1] Bapat, R.B. and Raghavan, T.E.S., Nonnegative Matrices and Applications,
Cambridge University Press, 1997.
[2] Bau, D. and Trefethen, L., Numerical Linear Algebra, Society for Industrial and
Applied Mathematics, Philadelphia, PA, 1997.
[3] Gaines, H.F., Cryptanalysis, Dover, New York, 1956.
[4] Moler, C. and Morrison, D., Singular Value Analysis of Cryptograms, The
American Mathematical Monthly, Vol. 90, No. 2, (Feb., 1983), pp. 78-87.
[5] Famous
Martin

Speeches
Luther

by
King.

Presidents,
Famous

Prime
Quotes.

Ministers
May-June

and
2008

<http://www.famousquotes.me.uk/speeches/index.htm>.
[6] Sinkov, A., Elementary Cryptanalysis, A Mathematical Approach, New Mathematical Library, vol 22, Mathematical Association of America, Washington,
D.C., 1968.
[7] Frequency

Analysis.

Wikipedia.

29

Apr.

2008.

May-June

<http://en.wikipedia.org/wiki/Frequency analysis (cryptanalysis)>.

26

2008

APPENDIX

GEORGE W. BUSHS 9/11 ADDRESS

Good evening. Today, our fellow citizens, our way of life, our very freedom
came under attack in a series of deliberate and deadly terrorist acts. The victims
were in airplanes, or in their offices; secretaries, businessmen and women, military
and federal workers; moms and dads, friends and neighbours. Thousands of lives
were suddenly ended by evil, despicable acts of terror. The pictures of airplanes
flying into buildings, fires burning, huge structures collapsing, have filled us with
disbelief, terrible sadness, and a quiet, unyielding anger. These acts of mass murder
were intended to frighten our nation into chaos and retreat. But they have failed; our
country is strong. A great people has been moved to defend a great nation. Terrorist
attacks can shake the foundations of our biggest buildings, but they cannot touch
the foundation of America. These acts shattered steel, but they cannot dent the steel
of American resolve. America was targeted for attack because were the brightest
beacon for freedom and opportunity in the world. And no one will keep that light
from shining. Today, our nation saw evil, the very worst of human nature. And
we responded with the best of America with the daring of our rescue workers,
with the caring for strangers and neighbours who came to give blood and help
in any way they could. Immediately following the first attack, I implemented our
governments emergency response plans. Our military is powerful, and its prepared.
Our emergency teams are working in New York City and Washington, D.C. to help
with local rescue efforts. Our first priority is to get help to those who have been
injured, and to take every precaution to protect our citizens at home and around
the world from further attacks. The functions of our government continue without
interruption. Federal agencies in Washington which had to be evacuated today are
reopening for essential personnel tonight, and will be open for business tomorrow.
Our financial institutions remain strong, and the American economy will be open for
business, as well. The search is underway for those who are behind these evil acts.
Ive directed the full resources of our intelligence and law enforcement communities
to find those responsible and to bring them to justice. We will make no distinction
between the terrorists who committed these acts and those who harbour them. I
appreciate so very much the members of Congress who have joined me in strongly
condemning these attacks. And on behalf of the American people, I thank the many
world leaders who have called to offer their condolences and assistance. America and
our friends and allies join with all those who want peace and security in the world,
and we stand together to win the war against terrorism. Tonight, I ask for your
prayers for all those who grieve, for the children whose worlds have been shattered,
for all whose sense of safety and security has been threatened. And I pray they will
be comforted by a power greater than any of us, spoken through the ages in Psalm
23: Even though I walk through the valley of the shadow of death, I fear no evil,
for You are with me. This is a day when all Americans from every walk of life unite
in our resolve for justice and peace. America has stood down enemies before, and
we will do so this time. None of us will ever forget this day. Yet, we go forward to
defend freedom and all that is good and just in our world.

27

DR. MARTIN LUTHER KINGS I HAVE A DREAM SPEECH

Five score years ago, a great American, in whose symbolic shadow we stand
signed the Emancipation Proclamation. This momentous decree came as a great
beacon light of hope to millions of Negro slaves who had been seared in the flames of
withering injustice. It came as a joyous daybreak to end the long night of captivity.
But one hundred years later, we must face the tragic fact that the Negro is still
not free. One hundred years later, the life of the Negro is still sadly crippled by
the manacles of segregation and the chains of discrimination. One hundred years
later, the Negro lives on a lonely island of poverty in the midst of a vast ocean of
material prosperity. One hundred years later, the Negro is still languishing in the
corners of American society and finds himself an exile in his own land. So we have
come here today to dramatize an appalling condition. In a sense we have come to
our nations capital to cash a check. When the architects of our republic wrote the
magnificent words of the Constitution and the declaration of Independence, they
were signing a promissory note to which every American was to fall heir. This
note was a promise that all men would be guaranteed the inalienable rights of life,
liberty, and the pursuit of happiness. It is obvious today that America has defaulted
on this promissory note insofar as her citizens of colour are concerned. Instead of
honouring this sacred obligation, America has given the Negro people a bad check
which has come back marked insufficient funds. But we refuse to believe that
the bank of justice is bankrupt. We refuse to believe that there are insufficient
funds in the great vaults of opportunity of this nation. So we have come to cash
this check a check that will give us upon demand the riches of freedom and the
security of justice. We have also come to this hallowed spot to remind America of
the fierce urgency of now. This is no time to engage in the luxury of cooling off
or to take the tranquillising drug of gradualism. Now is the time to rise from the
dark and desolate valley of segregation to the sunlit path of racial justice. Now
is the time to open the doors of opportunity to all of Gods children. Now is the
time to lift our nation from the quick-sands of racial injustice to the solid rock of
brotherhood. It would be fatal for the nation to overlook the urgency of the moment
and to underestimate the determination of the Negro. This sweltering summer of
the Negros legitimate discontent will not pass until there is an invigorating autumn
of freedom and equality. Nineteen sixty-three is not an end, but a beginning. Those
who hope that the Negro needed to blow off steam and will now be content will have
a rude awakening if the nation returns to business as usual. There will be neither
rest nor tranquillity in America until the Negro is granted his citizenship rights. The
whirlwinds of revolt will continue to shake the foundations of our nation until the
bright day of justice emerges. But there is something that I must say to my people
who stand on the warm threshold which leads into the palace of justice. In the
process of gaining our rightful place we must not be guilty of wrongful deeds. Let
us not seek to satisfy our thirst for freedom by drinking from the cup of bitterness
and hatred. We must forever conduct our struggle on the high plane of dignity
and discipline. We must not allow our creative protest to degenerate into physical
violence. Again and again we must rise to the majestic heights of meeting physical
force with soul force. The marvellous new militancy which has engulfed the Negro
community must not lead us to distrust of all white people, for many of our white
brothers, as evidenced by their presence here today, have come to realize that their
destiny is tied up with our destiny and their freedom is inextricably bound to our
freedom. We cannot walk alone. And as we walk, we must make the pledge that
we shall march ahead. We cannot turn back. There are those who are asking the
devotees of civil rights, When will you be satisfied? We can never be satisfied
as long as our bodies, heavy with the fatigue of travel, cannot gain lodging in the
28

motels of the highways and the hotels of the cities. We cannot be satisfied as long
as the Negros basic mobility is from a smaller ghetto to a larger one. We can
never be satisfied as long as a Negro in Mississippi cannot vote and a Negro in New
York believes he has nothing for which to vote. No, no, we are not satisfied, and
we will not be satisfied until justice rolls down like waters and righteousness like a
mighty stream. I am not unmindful that some of you have come here out of great
trials and tribulations. Some of you have come fresh from narrow cells. Some of
you have come from areas where your quest for freedom left you battered by the
storms of persecution and staggered by the winds of police brutality. You have been
the veterans of creative suffering. Continue to work with the faith that unearned
suffering is redemptive. Go back to Mississippi, go back to Alabama, go back to
Georgia, go back to Louisiana, go back to the slums and ghettos of our northern
cities, knowing that somehow this situation can and will be changed. Let us not
wallow in the valley of despair. I say to you today, my friends, that in spite of
the difficulties and frustrations of the moment, I still have a dream. It is a dream
deeply rooted in the American dream. I have a dream that one day this nation
will rise up and live out the true meaning of its creed: We hold these truths to be
self-evident: that all men are created equal. I have a dream that one day on the
red hills of Georgia the sons of former slaves and the sons of former slave-owners
will be able to sit down together at a table of brotherhood. I have a dream that one
day even the state of Mississippi, a desert state, sweltering with the heat of injustice
and oppression, will be transformed into an oasis of freedom and justice. I have a
dream that my four children will one day live in a nation where they will not be
judged by the colour of their skin but by the content of their character. I have a
dream today. I have a dream that one day the state of Alabama, whose governors
lips are presently dripping with the words of interposition and nullification, will be
transformed into a situation where little black boys and black girls will be able to
join hands with little white boys and white girls and walk together as sisters and
brothers. I have a dream today. I have a dream that one day every valley shall be
exalted, every hill and mountain shall be made low, the rough places will be made
plain, and the crooked places will be made straight, and the glory of the Lord shall
be revealed, and all flesh shall see it together. This is our hope. This is the faith
with which I return to the South. With this faith we will be able to hew out of the
mountain of despair a stone of hope. With this faith we will be able to transform the
jangling discords of our nation into a beautiful symphony of brotherhood. With this
faith we will be able to work together, to pray together, to struggle together, to go
to jail together, to stand up for freedom together, knowing that we will be free one
day. This will be the day when all of Gods children will be able to sing with a new
meaning, My country, tis of thee, sweet land of liberty, of thee I sing. Land where
my fathers died, land of the pilgrims pride, from every mountainside, let freedom
ring. And if America is to be a great nation this must become true. So let freedom
ring from the prodigious hilltops of New Hampshire. Let freedom ring from the
mighty mountains of New York. Let freedom ring from the heightening Alleghenies
of Pennsylvania! Let freedom ring from the snow-capped Rockies of Colorado! Let
freedom ring from the curvaceous peaks of California! But not only that; let freedom
ring from Stone Mountain of Georgia! Let freedom ring from Lookout Mountain of
Tennessee! Let freedom ring from every hill and every molehill of Mississippi. From
every mountainside, let freedom ring. When we let freedom ring, when we let it ring
from every village and every hamlet, from every state and every city, we will be able
to speed up that day when all of Gods children, black men and white men, Jews
and Gentiles, Protestants and Catholics, will be able to join hands and sing in the
words of the old Negro spiritual,Free at last! free at last! thank God Almighty, we
are free at last!
29

VITA
Graduate School
Southern Illinois University
JOSHUA BOONE

Date of Birth: February 6, 1984

6685 OLD HIGHWAY 13, APT. C2, CARBONDALE, ILLINOIS 62901

Southern Illinois University at Carbondale
Bachelor of Science, Mathematics, May 2006

Special Honors and Awards:

SIUC Masters Fellowship

Research Paper Title:

Singular Value Approximations in Cryptography

Major Professor: Dr. Robert Fitzgerald

30