Constructible Polygons

MATH214 – Spring 2023 Team member names

Constructible Polygons
2 Table of contents

Compass
Construc-
and
tion of ωn
Straightedge

Con-
structible
Polygons

Usage in
reality
1. Compass and Straightedge
4 Compass and Straightedge

Definition
• Compass: An idealized tool for drawing circle with given radius.
• Straightedge: An idealized tool for drawing a straight line.

Remark:
1 The radius of circle compass can draw has no limitations.
2 The line straightedge draw is of infinite length. However, scale
marks cannot be drawn merely by straightedge.
5 Euclidean Construction

Definition
Euclidean Construction: The construction of lengths, angles, and other
geometric figures in a plane space using only an idealized straightedge
and a compass with finite operations.

Remark:
1 Euclidean Construction is also known as
Compass-and-straightedge Construction
2 Things is much more different with infinite operations. As you can
”approach” a geometric figures with infinite operations, like what
you have done in mathematical analysis.
6 Euclidean Construction

The only constructions we can make is a sequence of the basic ones:

1 Connecting two given points with a straight line
2 Drawing the circle centered at a given point passing through
another given point
3 Constructing a point as the intersection of two lines
4 Constructing a point (or points) as the intersection of a line and a
circle
5 Constructing a point (or points) as the intersection of two circles
7 Intersection with algebra

Compass can do more than just ”drawing a circle”. The compass

preserves some ”information”, i.e. some given lengths. With compass,
we can ”move” a line segment without changing its length all over the
space. The core of algebra is ”identity”, i.e. stableness and
unchangeableness of a given object. With initial inspiration, we can
relate algebra with Euclidean Construction.
Randomly pick up two points in the space. We set the length of the
line segment to be 1 and either point to be origin. Draw an orthogonal
line with length one, and we can construct a cartesian coordinate
system, thus a complex plane.
8 Intersection with algebra

Definition
Constructible Number: Given the unit length, the real number
representing the length of a line segment that can be drawn with
Euclidean construction is called Constructible Number.

Lemma
A point is reachable by Euclidean construction if and only if the
imaginary and real part of the imaginary it represents are both
constructible numbers.
Proof:
(⇒) Assume P(x, y ), where x and y are constructible numbers. Then
points Px (x, 0) and Py (0, y ) can be Euclidean constructed. Using
perpendiculars and translations, P can thus be reached.
9 Intersection with algebra

(⇐) Since the coordinate of point P(x, y ), as well as coordinates

(0, 0), (1, 0), (0, 1) are constructed, the vertical and horizontal axis
can be constructed. Then the projection point on x-axis Px (x, 0) and
on y-axis Py (0, y ) can also be constructed. Hence, x and y are
constructible numbers.
From now, we have introduced Euclidean construction into algebra.
10 Field Extension
Definition
Field Extension: A field extension is a pair of fields K ⊆ L, such that
the operations of K are those of L restricted to K. In this case, L is an
extension field of K and K is a subfield of L. We can also denote L as
L/K, i.e. L over K

Example: C can be viewed as the extension field of R

Remark:
1 L should preserve all operations that belong to the field K

2 If K is the subfield of L, L can be viewed as a K-vector space.

3 For a field F,viewing F as a F-vector space, we denote
L = span(α1 , α2 , · · · , αn ) + F as F (α1 , α2 , · · · , αn )
Definition
Degree of the extension:For a field extension K ⊆ L, if L as a K-vector
space is a finite-dimensional space, then the degree of the extension is
the dimension of L, denoted by [L : K ]
11 The Field of Constructible Numbers

Theorem
The set C of constructible numers is a field extension of Q. Apart from
operations in Q, i.e. plus, multiply, C is also closed under taking square
root.
Proof: You can do plus, subtract, multiply and divide through
Euclidean construction in high school. With plus, subtract and a unit
element we can construct Z. With division we can construct Q. Taking
square roots with compass and straightedge is also taught in high
school.
Remark: You can only do these five operations in Euclidean
construction. Viewing lines and circles as linear and quadratic
equations in the cartesian coordinate system, their solutions of their
intersections only contain five operations. However, this gives no
implication of elements in C. We need to show that only certain
numbers are constructible.
12 Properties of Field Extension
Properties:
1 Given three field K , L, M, withK ⊆ L ⊆ M, and [M : L], [L : K ]
are both finite, then [M : K ] is also finite and
[M : K ] = [M : L][L : K ]
√ √ √
2 If r1 , r2 , · · · , rn ∈ Q, while r1 , r2 , · · · , rn ∈
/ Q and
√ √ √
{ r1 , r2 , · · · , rn } are linearly independent under
√ √ √ √ √ √
Q( r1 , r2 , · · · , rn ), then [Q( r1 , r2 , · · · , rn ) : Q] = 2n
3 [Q(ω, r1 , · · · , rn ) : Q(r1 , · · · , rn )] = 2 ⇔ ∃r ∈ Q(r1 , · · · , rn ), such
√ √
that ω = r , with r ∈ / Q(r1 , · · · , rn ) .
Remark:
Euclidean construction only allows for plus, subtract, multiply, division
and taking square roots. The former four operations are close under Q
and any its extension fields. Property 2 can directly lead to the
following result: if ω is constructible, then [Q(ω) : Q] = 2t , with
t ∈ N. While the converse has not yet been justified, we have the
feeling that we can show the constructibility of some numbers by
studying the degree of extension field constructed from these numbers
13 Table of contents

Compass
Construc-
and
tion of ωn
Straightedge

Con-
structible
Polygons

Usage in
reality
2. Construction of ωn
15 ωn and Polygon

We know that the root of x n = 1 on the complex plane are the points
of an n regular polygon, and all the root can be represented as
ωnk , 1 ≤ k ≤ n, ωn = cos 2π 2π
n + isin n
Then the problem of constructing a polygon can be transformed into
an algebra problem, i.e. whether ωn is constructable.
16 Field Chain

Theorem
n + isin n . If [Q(ω) : Q] = 2 , t ∈ N, then
Suppose ω = cos 2π 2π t

∃Q = F0 ⊂ F1 ⊂ F2 ... ⊂ Ft ⊂ Ft+1 = Q(ω), [Fi+1 : Fi ] = 2, 0 ≤ i ≤ t

Remark: This theorem indicates that if [Q(ω):Q]=2t , then there is a

field chain to reach Q(ω) from Q, and each extension has degree 2.
When [Fi+1 : Fi ]=2, we can use square root and plus to represent all
the element in Fi+1 with the element in Fi . Therefore, the element in
Q(ω) can be represented by using the element in Q with several times
operations of plus, multiply and square root, which means ω is
constructible.
Conversely, if ω is constructible, using the Remark in 1.12 we have
[Q(ω):Q]=2t
Therefore, we have ω is constructible ⇔ [Q(ω):Q]=2t , t ∈ N
17 Galois Group

Definition
The Galois Group: Let E be a finite extension of a field F (i.e., E /F ,
with E over F ). {Aut(E /F ) is the set of field automorphisms
} of E over
F . Gal(E /F ) = σ|σ ∈ Aut(E /F ), a = a, ∀a ∈ F , (Gal(E /F ), ◦) is
σ

a group.

Remark: It’s clear that σ ∈ Gal(E /F ) is also a automorphism on the E

vector space over F.
Theorem
[Q(ω) : Q] = |Gal(Q(ω) : Q)|

Remark: With this theorem we get ω is constructible ⇔

|Gal(Q(ω) : Q)|=2t , t ∈ N , so we only need to figure out how many
elements are in Gal(Q(ω) : Q)
18 Galois Group

Lemma
There is a isomorphism between Gal(Q(ω) : Q) and (Z/nZ)×

Proof:
It’s clear that σ(ω) will determine the{ σ }
We can prove that Gal(Q(ω) : Q) = σ|σ(ω) = ω r , r ∈ (Z/nZ)×
Suppose σr (ω) = ω r , as σr (σs (ω)) = σrs (ω), then σr σs = σrs , so
f : σ(ω) = ω r → r is a isomorphism from Gal(Q(ω) : Q) to (Z/nZ)×
■
19 Galois Group

According to the this lemma, we can know that Gal(Q(ω) : Q) and

Z/nZ have the same number of elements. We know that Z/nZ
consists of all {
positive integers that are less
} than n and coprime with n
(i.e., Z/nZ = a ∈ J1, n − 1K|(a, n) = 1 ). The number of its
elements is also known as the Euler function value of n, denoted φ(n).
So we got the following results:

[Q(ω) : Q] = |Gal(Q(ω) : Q)| = φ(n)

Therefore, if φ(n) is a power of two, then the nth roots of unity is

constructible, which means we can construct a regular n-gon by using
straightedges and compasses. And note that the lemma on page eight
is an iff proposition, which means this is also a sufficient and necessary
condition.
So what numbers meet the requirements?
20 Constructible Polygons

Theorem
For any the positive integer n, φ(n) is a power of two if and only if
n = 2s p1 p2 ... pt (s, t ∈ N, pi is the Fermat Prime).
p.s. Fermat prime refers to all prime numbers that can be expressed in
k
the form 22 + 1 with k ∈ N.
Proof.
Firstly, based on some simple knowledge of number theory, we can
know that
If (m, n) = 1, φ(mn) = φ(m)φ(n) (1)
For any positive integer n, we can express it as n = 2k0 p1 k1 p2 k2 ... pt kt
(pi are all the odd prime factor of n, ki ∈ N). And we know that
∀i, j ∈ J1, tK, (pi , pj ) = (pi , 2) = 1, and φ(2k0 ) = 2k0 −1 is obviously a
power of two. So φ(n) is a power of two if and only if ∀i ∈ J1, tK,
φ(pi ki ) is a power of two.
21 Constructible Polygons

Based on some simple knowledge of number theory, we know that

φ(p k ) = p k − p k−1 = p k−1 (p − 1). So if k ≥ 2, then p|φ(p k ) can’t be
a power of two. And it’s easy to find that φ(p) = p − 1. So p should
be 2m + 1(m ∈ N). If m have a odd factor x that m = x × mx (x > 1),
then 2m + 1 = (2mx )x + 1x = (2mx + 1)(... ). This contradicts the fact
that p is a prime number. So m can’t have any odd factors, which
means m is a power of two. So all pi should be Fermat prime. ■

Summary

Now we have completed the entire process of proof, the final

conclusion is: We can construct a regular n-sided polygon if and
only if n = 2s p1 p2 ... pt (s, t ∈ N, pi is the Fermat Prime).
Note: Currently, there are only five known Fermat prime numbers:
2 5 17 257 65537
3. Usage in reality
23 The Role of Constructions in Modern Science

Indeed, compass-and-straightedge constructions are no longer practical

tools in modern engineering and science, they still have contributions
in theoretical and historical importance.
The question of constructing regular polygons led to the development
of Galois Theory which not only solved the construction problem but
also provided deep insights into soling polynomial equations, which is
fundamental in fields like cryptography, coding theory and quantum
mechanics.
Definition
A Galois field is a field that contains a finite number of elements. Foe
every prime power p n , where p is a prime number and n is a positive
integer, there exist a unique Galois field with p n elements, denoted by
GF (p n )

Example:
GF (p n ) = {a(x)|a(x) = ap x p + ... + a1 x + a0 ; ai ∈ {0, 1, 2, ..., p − 1}}
24 Cryptography

We have shown that constructing a regular n-gons is equivalent to

determining whether [Q(ω) : Q] has a degree of two. This idea is
similar to the cryptography.
Example: In the AES(Advanced Encryption Standard) encryption
algorithm, each byte is treated as an element in the finite field GF (28 ).
While constructing this field is similar to constructing ωn .
Remark: When constructing GF (28 ) from GF (2), we first construct a
polynomial ring GF (2)[x] and a irreducible polynomial
f (x) = x 8 + x 4 + x 3 + x + 1 which cannot be divisible by all
polynomials with less degree than 8 on GF (2). Then GF (28 ) is defined
by:
GF (28 ) = (GF (2)[x])mod(f (x))
So GF (28 ) is a finite extension of GF(2), whose dimension is 8, and
Gal(GF (28 ) : GF (2)) is isomorphic to (Z/8Z)× .
Thank you!