Geometry Extended
Our purpose in compiling and presenting this simple geometry is to
provide an example of how higher dimensional geometry can successfully be
integrated into secondary school level mathematics courses, without
sacrificing mathematical rigor. Towards that end, our geometry is a highly
derivative work, drawing many postulates, theorems, and sometimes even
proofs from Houghton-Mifflin's outstanding School Mathematics
Geometry . This book is not mathematically naive, but is nevertheless
acessible to the beginner, and so provided a comfortable starting point
for our work on this project.
Our geometry adheres as much as conveniently possible to the
traditional form for geometries, deriving theorems about sets of undefined
primitives from unproven postulates. In addition, in order to keep this
geometry as user-friendly as possible, it is synthetic and Euclidean. The
only unfamiliar element of it should be the fact that it applies to sets of
points in four dimensions.
Included is a list of all postulates and
theorems.
1. Undefined Terms
It is necessary in any logical system to start with some undefined
terms, sometimes called primitves. All other terms may be defined in terms of
these few primitives. While presenting the undefined terms below, we
will offer some intuitive examples of physical representations of them, but
these intuitive explanations are not to be confused with rigorous
definitions. They are simply to provide a more gut-level understanding of
what the theorems and postulates to follow are saying.
Our first undefined term is point. Euclid described the point
as "that which has no part." Euclid meant that the point has no extension,
in neither breadth, nor width, nor height. It is usually represented in
geometric figures as a tiny dot, but keep in mind that this tiny dot is
just a convenient representation of a point; if it truly had no width or
breadth or depth, it would be impossible to see. The abstract points which
concern geometry are without extension at all. The point is an abstraction
of location; it corresponds perfectly to a precise place.
The next undefined term is line. A line, as will be made clearer later
on, is in fact a set of points, but it is nonetheless not definable
entirely in terms of points. When we talk about lines, we are talking about
straight lines, such as the edge of a ruler. The line only extends in one
dimension, and extends infinitely in both directions along that dimension. A line
is also thought of as infinitely dense with points, such that if you grab it
at any place along the infinite line, you are guaranteed to touch it at a
point-- another way of saying this is that there are no holes in the line.
Next we have the plane. A plane, like a line, is a set of
points, but one which cannot be defined solely in terms of points, and so
is treated as an undefined term. The plane has extension in width and
breadth, but not depth. It is analogous to the line, in that as the line is
straight, the plane is flat; as the line is infintely dense with points,
so is the plane; and as the line extends infinitely in one dimension, the
plane extends infinitely in two dimensions.
Finally, we will also be using the term space, or sometimes
hyperplane to denote a set of points which extends in three
mutually perpendicular directions. A space is the logical extension of the
progression from point (0 dimensions) to line (1 dimension) to plane (2
dimensions). Since our geometry is treating points in four dimensions,
however, the notion of space as used in it may be quite different from the
ordinary understanding of a space being the set of all points. In fact,
there are an infinite number of spaces in our geometry, and we will often
make reference to two spaces intersecting, for example. In ordinary
experience, one is unlikely to encounter spaces intersecting, but for the
purposes of our geometry, they will.
2. Distance
2.1 Initial Postulates
Before continuing, we must make some unproven assumptions on which to
base our proofs in this chapter.
Postulate 1
Given any two points, there is exactly one line which contains both of them.
Another way of putting this is to say that two points determine a line.
When talking about lines, it is often handy to use this postulate to only
refer to two points when speaking about the line; since the two points
determine a line, there is no possibility of confusion as to which line one
is referring to. When we write, "A<->B", we mean, "the line determined by
points A and B," which is the same as "B<->A."
Postulate 2: The Distance Postulate
Given any pair of distinct points, there corresponds a unique positive
real number called the distance between the two points.
Notice that we aren't using any units to measure distance in. Geometry is
just an abstraction; if one wanted to apply it to real world problems, one
would have to be concerned with units of measurement, but for our
abstraction generic real numbers will suffice. When referring to the
distance between points A and B, we will write, "AB," which is equivalent
to "BA."
Postulate 3: The Ruler Postulate
The points of a line can be placed in correspondence in such a way that:
- there is a one to one correspondence between the set of points on
the line and the set of real numbers, and
- the distance between two points equals the absolute value of the
difference of the corresponding numbers
Such a correspondence is called a coordinate system. Given a
coordinate system for any line, the number corresponding to a given point
is called the coordinate of the point.
The Ruler Postulate tells us that these coordinate systems exist, but
doesn't give us any way of quickly creating one. For that purpose, we will
use the Ruler Postulate to prove our first theorem, the Ruler Placement
Theorem.
Thm 2.1: The Ruler Placement Theorem
Given any two points P and Q of a line, there is a coordinate system for
the line such that the coordinate of P is zero and the coordinate of Q is
positive.
P R O O F
The Ruler Postulate tells us that some coordinate system exists for P<->Q.
Let x denote the coordinate of P in this system. We can create a new
correspondence like that described in the Ruler Postulate by adding -x to
every coordinate of the line. In this new system, the coordinate of P is 0.
If the coordinate of Q happens to be positive, we are done. If it is
negative, then we can simply create a new coordinate system by reversing
the sign of each coordinate. Thus P has coordinate zero and Q has a
positive coordinate.
2.2 Betweenness
Betweenness is an excellent example of how intuitive concepts can be
codified mathematically. Nearly everyone has an intuitive understanding of
what it means to say that an object is between two other objects. This
intuitive understanding is not mathematically rigorous enough for our
purposes, however. Thus the following definition:
B is between A and C if
- A, B, and C are distinct points on the same line, and
- AB + BC = AC.
We sometimes will write A-B-C, which means "B is between A and C."
The definition uses distance, and applying the concept of a coordinate
system we have the following theorem.
Thm 2.2
Let A, B, and C be three distinct points of a line, with the respective
coordinates x, y, and z. If x < y < z, then A-B-C.
P R O O F
By the ruler postulate:
- AB = |y-x|
- BC = |z-y|
- AC = |z-x|.
But y-x is positive because x < y; similarly, z-y and z-x are positive. So:
- AB = y-x
- BC = z-y
- AC = z-x
Finally:
AB + BC = (y - x) + (z - y)
= z-x
=AC.
Since AB + BC = AC, A-B-C by the definition of between.
Thm 2.3
Given any three distinct points on a line, exactly one is between the
other two.
P R O O F
Let A, B, and C be three distinct collinear points. We will first prove
that at least one of the following is true: A-B-C, A-C-B, or B-A-C. We will
then show that at most one of these is true.
Let the respective coordinates of A, B and C be x, y and z. Either x < y
or y < x. If x < y, then there are three possibilites:
- z < x < y,
- x < z < y,
- x < y < z.
If x < y, then there are also three possibilities:
- y < x < z,
- y < z < x,
- z < y < x.
In each of these cases, Thm 2.2 tells us that one of the points A, B, and
C is between the other two, so we know that at least one point is between
the other two. We will now show that at most one point is between the
other two.
If A-B-C, then by definition of between, AB + BC = AC, and by the Distance
postulate, AB, BC and AC are all positive numbers. This means that AC > BC
and AC > AB. If A-C-B, similarly, AC + BC = AB, so that AB > AC and AB >
CB. Similarly, if B-A-C, then BA + AC = BC, so that BC > AB and BC > AC.
Now, it is clear that at most one of the possible cases can be true: if
A-B-C, (meaning that AC > AB and AC > BC) then it is impossible for C to be
between A and B, as AB would be simultaneously less than and greater than AC.
It is also impossible for A to be between B and C, because BC would be both
less than and greater than AC. Similar logic applies to the other two cases,
so that at most one point can be between the other two points.
2.3 Segments, Rays and Midpoints
Obviously, one rarely encounters lines of infinite length in problems
involving geometric truths. It is much more likely that one is concerned
with problems concerning finite, straight lengths. For this reason, geometry
provides an abstractions of a portion of a straight line. These abstractions
are called segments and rays.
Given any two points A and B. The set containing A, B, and all points
between A and B, is a segment, and is denoted by A--B or B--A. The
points A and B are the endpoints of the segment. The segment is
sometimes said to join its endpoints. The length of a
segment is simply the distance between its endpoints, and as such is
written as AB.
The ray A->B is the union of
- the segment joining A and B, and
- the set of all points C such that A-B-C.
The point A is called the endpoint of the ray. The interior of
the ray is the set of all points of A->B different from A. Notice that
the ray A->B is different from the ray B->A, whereas the segment A--B is
identical to the segment B--A. If A-B-C, the rays B->A and B->C are called
opposite rays. The point B is called the common endpoint of the
two rays.
LEMMA
If A-B-C, and the respective coordinates of A, B and C are x, y and z, then
either x < y < z or z < x < y.
P R O O F
By the definition of between, we know that AB + BC = AC. We also know that AB,
BC, and AC are positive numbers from the distance formula. Thus AC > AB,
and AC > BC. Since A, B and C are distinct, their coordinates must be
different. Therefore, either x < y or x > y.
Let us first examine the case where x < y. Now, either y < z or y > z.
Suppose for the time being that y > z. We do not know whether x < z < y or
z < x < y, but it is immaterial. If either one is true, then we have
violated Thm 2.3, because we are given that A-B-C, whereas the inequalities
above would assert respectively that A-C-B and C-A-B. Therefore, if x <
y, x < y < z.
The other case, where x > y, works out similarly.
Thm 2.4
Given any coordinate system for a line. Of any two opposite rays on the
line, one ray is the set of all points having coodinates greater than or
equal to the coordinate of the common endpoint, and the other ray is the
set of all points having coordinates less than or equal to the coordinate
of the common endpoint.
P R O O F
Let A, B and C be three points on the given line such that A-B-C, and let
x, y, and z be the respective coordinates of A, B and C. We wish to show
that either B->A or B->C is the set of all points having coordinates
greater than or equal to y (the coordinate of B), and that the opposite
ray is the set of all point with coordinates less than or equal to y.
By the lemma, either x < y < z, or z < x < y. Let us first examine the
case where x < y < z. Let P be any point on B->C, and let p denote its
coordinate. We wish to show that p >= y.
By the definition of ray, either P is B, B-P-C, or B-C-P. In the first
case, p=y and p >=y. In the second case, by the lemma, either y < b < z or
z < b < y. Since we are examining the case where y < z, we know that b > y.
If, on the other hand, B-C-P, the lemma tells us that either y < z < p or
p < z < y, and again we know that y < z < p because we are examining the
case where y < z. So indeed, for any point P on ray B->C, its coordinate is
greater than or equal to the coordinate of B.
What of the opposite ray? Again, let P be any point on B->A, and let p be
P's coordinate. Either P is B, or A-P-B, or P-A-B. We wish to show that in
all three cases, p <= y. If P is B, we have the case where p = y. If A-P-B,
the lemma indicates that either x < p < y, or y < p < x, and just as above
we know that x < p < y because we are examining the case where x < y < z.
So we know that p < y. Again, if P-A-B, by the lemma, either p < x < y or
y < x < p, and we know that p < x < y because we know that x < y. So
again, p < y.
To sum up: we first observed that either x < y < z or z < y < x. We
decided to examine the first case, and have proven that the
demonstrandum holds for it, because for any point on B->C, its
coordinate is greater than or equal to B's coordinate, and for any point
on B->A (which is, by definition of opposite ray, B->C's opposite ray), its
coordinate is less than or equal to B's. The same logic would apply for
the case where z < y < x, and it would be unnecessarily repetitive to
include the same arguments again here.
Thm 2.5: The Point-Plotting Theorem
Let A->B be a ray and let x be a positive number. Then there is exactly one
point P of A->B such that AP = x.
P R O O F
By the Ruler Placement Theorem, a coodinate system for A<->B can be chosen
so that the coordinate of A is zero and the coordinate of B is positive.
Having chosen this coordinate system, by Theorem 2.4A->B is the set of all
points of A<->B having nonnegative coordinates. Let P be the point of
A<->B whose coordinate is the positive number x. Then, since x is positive,
P is on A->B and AP = |x-0| = x. So P is a point of A->B such that AP = x.
If Q is any other point of A->B, then Q has a nonnegative coordinate y
that is different from x. Then AQ = |y-0| = y and so AQ != x. Therefore P
is the only point of A->B such that AP=x.
The next theorem pertains to the midpoints of segments. A point B is
called a midpoint of segment A--C if A-B-C and AB = BC.
Thm 2.6: The Midpoint Theorem
Every segment has exactly one midpoint.
P R O O F
Let A and C be the endpoints of a segment. We wish to show that there is
exactly one point B such that AB = BC and A-B-C. (Notice that the
stipulation of A-B-C is necessary for any geometry of dimension greater
than 2, as there are infinitely many points in three or more dimensions
that are equidistant from two points.)
We first designate a coordinate system for A<->C such that the coordinate
of A is zero and the coordinate of C is positive. By the Point-Plotting
Theorem, there is exactly one point B of A->C such that AB =1/2*AC. Since
0 < 1/2*AC < AC, B is between A amd C by Thm 2.2, so AB + BC = AC. Since
AB = 1/2*AC, we can see by sybtraction that BC = 1/2*AC as well. Since
A-B-C and AB = BC, B, is a midpoint of A--C.
For any other point D between A and C, AD is the coordinate of D and AD !=
1/2*AC. We know that AD + DC = AC. Then AD != DC, because if AD = DC, then
AD + AD = AC, so AD = 1/2*AC, which is impossible. So, D is not a midpoint
of A--C, and B is the only midpoint of A--C.
3. Lines, Planes, Spaces and Separation
The previous sections have been preparatory to this one. The content of
them has, by and large, not been spectacularly relevant to the topic of
higher dimensions, but they are necessary to provide firm footing for this
section. We will often point out when theorems are dimensionally analogous to one
another, and sometimes will even omit proofs if the proof is sufficiently
obvious from an analogous theorem.
Before proceeding, a simple definition: Hyperspace is
the set of all points.
Theorem 3.1
If two lines intersect, then their intersection contains exactly one point.
P R O O F
The proof follows directly from postulate 1: the intersection cannot
contain more than one point, because any two points are contained in
exactly one line. If the lines intersected in two or more points, the
points of the intersection would lie in at least two lines.
Some Postulates and Definitions
First, when we refer to a set of points as collinear, we mean that
some line contains all the points of the set. More than one set is
collinear if their union is collinear. Similarly, a set of points is
coplanar if some plane contains all the points of the set, and more
than one set are coplanar if their union is coplanar. Finally, a set of
points is cospatial if some space contains them all, and more than
one set of points is cospatial if their union is cospatial.
Postulate 4 consists of three parts, all of which are analogous.
Postulate 4
- Every plane contains at least three noncollinear points.
- Every space contains at least four noncoplanar points.
- Hyperspace contains at least five noncospatial points.
Postulate 5
If a plane contains two points of a line, then the plane contains the whole
line.
The next postulate is analogous to Postulate 5.
Postulate 6
If a space contains three noncollinear points of a plane, then the space contains the
whole plane.
Theorem 3.2
If a line intersects a plane not containing it, then the intersection
contains exactly one point.
P R O O F
The proof follows directly from postulate 5; if the intersection of the
line and the plane contained more than one point, the plane would contain
the line, which violates the given. Therefore, no more than one point can
be in the intersection. Since it is given that the line intersects the
plane, the intersection can contain no more than one point.
3.2 Sets Determining a Plane
Postulate 7
Any three points lie in at least one plane, and any three noncollinear
points lie in exactly one plane.
Thm 3.4
Let l be a line and P be a point not on l. Then there is exactly one plane
containing l and P.
P R O O F
By the Ruler Postulate, l contains at least two different points, which we
call A and B. The points, A, B and P are noncollinear. By Postulate 7,
there is exactly one plane containing A, B and P, which we will call E.
Postulate 5 insures that E contains l, because it contains A and B. No
other plane can contain l and P because no other plane contains A, B and P.
Thm 3.5
Let l1 and l2 be any two intersecting lines. Then there is exactly one
plane containing them.
P R O O F
Let P be a point on l2, but not on l1, and let Q be the point o
fintersection of l1 and l2. By Thm 3.4, there is a plane E containing P and
l1. Since P and Q are in E and on l2, Postulate 5 insures that E contains
l2. Thus the plane E contians both l1 and l2. No other plane can contain l2
and l1 because no other plane can contain l1 and P.
Postulate 7
If two distinct planes intersect, then their intersection is a line.
3.3 Sets Determining a Space
This section is developed in parallel with 3.2. The theorems and
postulates in it are the higher dimensional analogues of the facts about
planes in 3.2
Postulate 8
Any four points lie in at least one space, and any four noncoplanar points
lie in exactly one space.
Thm 3.6
Let E be a plane and P a point not on it. Then there is exactly one space
containing E and P.
P R O O F
By Postulate 4, E contains at least three noncollinear points. Let these
three points be called A, B and C. The set of points {A,B,C,P} is
noncoplanar because A, B and C lie in exactly one plane, E, by postulate
six, and so, by postulate 6x, the four points lie in exactly one space.
This space contains E by Postulate 6 because it contains three points of
E. No other space can contain E and P because no other space can contain
A, B, C and P.
Thm 3.7
Let E and F be two intersecting planes. Then there is exactly one space
containing them.
P R O O F
Let l be the line of intersection of E and F. Let P be a point on E that is
not on F, and let A and B be distinct points on l. By Thm 3.5, there
exists a space S containing F and P. It also contains E by postulate 6, because
it contains three noncollinear points of E. No other space can contain E
and F because no other space can contain F and P.
Postulate 9
If two distinct spaces intersect, then their intersection is a plane.
3.4 Convexity and Separation
In this section we examine mathematically the intuitively obvious ideas of
convexity and separation. First a definition: a set of points is
convex
if for any two points P and Q of the set, the entire segment P--Q is in
the set. Using this definition, we will state the following important
postulate:
Postulate 10: The Plane Separation Postulate
Given any line and any plane containing it. The points of the plane that do
not lie on the line form two disjoint sets such that
- each of the sets is convex and non-empty and
- if P is any point of one set and Q is any point in the other set,
then the segment P--Q intersects the line.
Given a line l and a plane E containing l, each of the two sets
established by postulate 8 is called a half-plane with edge
l. N.B: neither half-plane contains its edge. We say that l
separates E into two half-planes, called opposite half-planes.
The next
Thm 3.8
Given any plane E, any line l in E, and any two points A and B in E but not
on l. If no point of l is between A and B, then A and B lie in the same
half-plane with edge l.
P R O O F
Suppose that A and B were in opposite half-planes with edge l, then by the
Plane Separation Postulate, A--B would intersect l, and hence there would
be a point of l between A and B. But this contradicts our hypothesis.
Therefore A and B are in the same half-plane with edge l.
Thm 3.9
Given any plane E, any line l in E, and any two points A and B that are in
E but not on l. If some point of l is between A and B, then A and B are in
opposite half-planes with edge l.
P R O O F
Suppose that A and B were in the same half-plane with edge l. Then that
half-plane, which is convex, would contain A--B, and thus no point of l
would be between A and B. But this would contradict the given data.
Therefore, A and B are not in the same half-plane with edge l.
3.5 More on Separation
Postulate 11: The Space Separation Postulate
Given any plane and any space containing it. The points of space that do
not lie on the plane E form two disjoint sets such that
- each of the sets is convex and non-empty and
- if P is any point in one set and Q is any point in the other setm then
the segment P--Q intersects the plane.
Each of the sets described in the Space Separation Postulate is called
a half-space with face E. A plane is said to separate
a space into two half-spaces that are called opposite
half-spaces.
Postulate 12: The Hyperspace Separation Postulate
The points of hyperspace that do not lie in a given space form two
disjoint sets such that
- each of the sets is convex and non-empty, and
- if P is any point in one set and Q is any point in the other set, then
the segment P--Q intersects the space.
Similarly, these two disjoint sets are called opposite
half-hyperspaces with boundary E.
Thm 3.10
Given any space S and any plane E in S and any two points A and B that are
in S but not on E. If no point of E is between A and B, then A and B are
in the same half-space with face E.
P R O O F
Suppose that A and B were in opposite half-spaces with face E. Then, by the
Space Separation Postulate, A--B would intersect E. Then, however, a point
of E would be between A and B. This would violate the hypothesis. Thus, A
and B must be in opposite half-spaces with face E.
Thm 3.11
Given any space S and any plane E in S, and any two ponts A and B that
arein S but not on E. If some point of E is between A and B, then A and B
are in opposite half-spaces with face E.
The proof is analogous to that of Thm 3.9
Thm 3.12
If a ray contains a point of a half-plane, and if the endpoint of the ray is
on the edge of the half-plane, then the interior of the ray lies in the
half-plane.
P R O O F
Let H be a half-plane with edge l. Let A->B be a ray with its endpoint on
l and B in H. We wish to show that the interior of A->B lies in half-plane
H.
By the Plane Separation Postulate, H is convex, so A--B is contained in (H
U l) (by def. of convexity). For any C such that A-B-C, B--C is contained in H,
also by definition of convexity. A->B's interior consists of the union of the
interior of A--B with all such C's, so A->B's interior is containted in
half-plane H.
The final two theorems are offered without proof, as their proof runs
along lines quite similar to that of thm 3.12.
Thm 3.13
If a ray contains a point of a half-space, and if the endpoint of the ray
is on the face of the half-space, then the interior of the ray lies in the
half-space.
Thm 3.14
If a ray contains a point of a half-hperspace, and if the endpoint of the
ray is on the face of the half-hyperspace, then theinterior of the ray lies
in the half-hyperspace.