April 14, 2009 Version 1.0 Page 1
Non-Euclidean Geometry Rick Roesler
I can think of three ways to talk about non-Euclidean geometry. I’m pretty sure they
are all equivalent, but I can’t prove it.
1. The Parallel Postulate
Euclidean geometry is called ‚Euclidean‛ because the Greek mathematician Euclid
developed a number of postulates about geometry.
One of these was the parallel postulate:
Given a line, L, and a point, P, not on L: there is exactly one line through P that’s
parallel to L.
For a really long time (2000 years?) people tried to prove that the parallel postulate
could be derived from the other postulates; in fact, it’s independent. And because it’s
independent, you can invent a geometry that has a different postulate - these are the
non-Euclidean geometries.
Euclidean geometry is the geometry of a ‘flat’ space - like this piece of paper or
computer screen (a plane) -- or Newtonian space-time. There are two archetypal non-
Euclidean geometries: spherical geometry and hyperbolic geometry. I’ll mostly talk
about spherical geometry because it’s easier to picture, and I found some good graphics
on the web. In spherical geometry, there are no parallel lines – not even one! So what’s
a ‚line‛on the surface of sphere? In Euclidean geometry (a plane), when we have a way
to measure distances between points, one way to define a line (segment) is that it’s the
curve between two points that has the minimum length. In a plane, the curve isn’t
curved, it’s a straight line, but it satisfies the ‘minimum distance’ principle.
April 14, 2009 Version 1.0 Page 2
In math, these curves of minimum distance are called geodesics. Get a ball (whose
surface is a sphere) and a piece of string; mark two points on the ball and stretch the
string tightly along the surface of the ball between the two points. You’ll find that the
string is part of a great circle; a great circle is the intersection of the sphere with a plane
passing through the center of the sphere. Because we’ve stretched the string as tightly
as possible, the curve that the string defines must have the shortest length between the
two points; if it didn’t, we could stretch it even tighter. So a great circle path is a
geodesic on a sphere.
Figure 1. Three examples of great circles on a sphere. Each great circle is in a plane that includes the center of the sphere. The center of the sphere is also the center of each great circle.
On the earth, the equator is a great circle, as are the lines of constant longitude. (But not
the lines of constant latitude. Why?) Now you can convince yourself that on a sphere
there are no ‚parallel lines‛. Why?
The surface of a sphere satisfies all the other Euclidean axioms, but not the parallel
postulate. So it’s non-Euclidean. By the way, you now understand why a flight from
Dallas to Tokyo goes through Alaska. Why? (And this is a great example of an
‘everyday use’ of non-Euclidean geometry. But gravity is better and more exciting, and
I’ll get to that later!)
April 14, 2009 Version 1.0 Page 3
2. A Triangle’s Angles Don’t Have to Sum to 180˚
In a plane (Euclidean geometry), if you draw a triangle and measure the three included
angles, you’ll find that the sum always add up to exactly 180˚. Now draw a triangle on
a globe (spherical, non-Euclidean, geometry). The first side goes from the North Pole to
the equator via the prime meridian (0˚ longitude, near London). For the second side,
start at the North Pole and go to the equator via New Orleans (90˚W longitude). The
third side goes along the equator, connecting the first two sides. Now look at the
angles - what’s the measure of each one? What’s the total?
You should have found that the total is greater than 180˚; in fact, for this example, it
should have been 270˚. This will always be true in a spherical geometry, but how far it
deviates from 180˚ depends on the area of the triangle. (In hyperbolic geometry, by the
way, the sum of the angles is always less than 180˚.) So the second definition of non-
Euclidean geometry is something like: ‚if you draw a triangle, the sum of the three
included angles will not equal 180˚.‛
April 14, 2009 Version 1.0 Page 4
Figure 2. Example of a spherical triangle. In this illustration the angle at the North Pole is
50˚ rather than the 90˚ angle we constructed in the text; here the sum of the angles is 230˚. When we construct smaller triangles on the sphere (see the inset), we end up with angles
that sum to almost exactly 180˚.
Remember, I said that the deviation from 180˚ depended on the size of the triangle. In
the picture above, the inset shows that if you draw a triangle in your back yard and
measure the angles, you’ll get something very, very close to 180˚; your back yard is
locally flat and approximates a Euclidean plane - even though your back yard is really
curved because it’s part of the Earth’s surface1. That’s why for a very, very long time,
people thought the earth was flat. Because it is – locally.
1 The spherical triangle that we constructed with three 90˚ angles has an area 1/8 that of the earth’s surface
(Aearth ~ 5 x 1014
m2). If you drew an isosceles right triangle in your back yard with each side being 1m, the
triangle’s area would be 0.5 m2 which is 10
15 times smaller – a million billion times smaller -- than the surface of
the earth. And that’s why the earth seems flat to us.
April 14, 2009 Version 1.0 Page 5
3. PARALLEL TRANSPORT - THE KEY TO UNDERSTANDING EINSTEIN’S GENERAL
THEORY OF RELATIVITY (GRAVITY)!
Imagine that you’re at the North Pole, and you’re carrying a spear. Hold the spear
horizontally and point it south (hint: every direction is south!). Start walking along a
constant-longitude line until you reach the equator. Note that as you move, the spear
never changes direction, it’s always pointing south; each position of the spear is parallel
to its previous position. Now, when you’ve reached the equator, keep facing south, but
move sideways toward the east along the equator. Your spear is still pointing south; it’s
still being parallel transported. After you’ve gone a fair way, maybe 1/8 the
circumference, around the earth along the equator, start walking backwards toward the
North Pole along a line of constant longitude. Your spear is still facing south.
When you reach the North Pole (your starting point), you’ll find that the spear is NOT
pointing in the same direction it started in, even though we never changed its direction
while we walked! The result is a global rotation without local rotation.
Figure 3. Parallel transport of a spear. The red arrows represent the spear; the closed path that the spear traverses is in blue. Note that the spear never changes direction; it always points South. Yet when we return it to the North Pole, it’s been rotated from its initial direction.
April 14, 2009 Version 1.0 Page 6
You should take some time now to convince yourself that if you’re in a Euclidean (flat)
space - like a flat piece of paper, a plane – a parallel transport of a spear will not change
its direction when you go in a closed loop; it always ends up pointing in the same
direction it started. So another definition of non-Euclidean geometry is something like:
if I parallel-transport a spear around a closed loop, it will end up pointing in a different
direction than it started.
Now you’re ready to understand gravity! Imagine that you and a friend are ants that
are standing at two different locations on the equator. Since you’re ants, you don’t
understand ‚up‛ and ‚down‛; you only understand two dimensions: ‚forward and
backward‛ and ‚left and right‛. Begin by measuring the distance between you - you’re
very sophisticated ants. Now you both start walking north – your initial directions are
parallel – toward the North Pole. As you walk, you continue to measure the distance
between you, and you find that it’s getting less and less!
You know that Newton, a very famous ant-physicist, said that objects will move in a
straight line with constant velocity unless acted upon by an external force; therefore,
two objects moving parallel to each other will always remain the same distance apart
unless acted upon by an external force. Newton was a Euclidean thinker. And in
Euclidean geometry if you and your friend start off walking parallel to each other, the
distance between you should always be constant - that’s what parallel lines do in
Euclidean space! But your relative distance is decreasing! Therefore, there must be
some force pulling you together because you eventually collide once you reach the
North Pole. Being good students of Newtonian physics, you decide to call that force
‚gravity‛.2
Einstein modified Newton and said that, instead of moving in straight lines, objects
move along geodesics (shortest path curves) in curved (non-Euclidean) space-times
unless acted on by an external force. In addition, Einstein said that it’s mass that curves
space-time. So gravity really isn’t a force at all. What we call gravity is simply the
result of us (or the earth) trying to move along the shortest space-time path when the
earth (or the sun) is curving the space-time we’re moving in. Mass here curves
2 It’s interesting that the ants can’t feel anything pulling them together. They just drift closer and closer together
as they move North. They infer the existence of the force because they think they’re in a flat space and the distance between parallel lines in a flat space should always remain the same. If you go skydiving or if you’re orbiting the earth, you don’t feel anything pulling you down either - you’re just in free fall. Remember, it’s not the fall that hurts; it’s the sudden stop when you hit the ground!
April 14, 2009 Version 1.0 Page 7
spacetime; that spacetime curvature causes the motion of mass there to deviate from a
straight line.
And, in fact, gravity is mostly due to the curving of time. Here’s a simple
demonstration you can do at home. The last page of this article is a picture that you can
print out. It’s what I call ‘cone world’. Cut the page along the line marked ‚floor‛.
Notice that, because the paper is flat, the shortest distance between two points is a
straight line. Roll the paper into a cone and tape the sides together; space (distance
from the floor) goes up and down the cone in straight lines, while time goes around the
cone in a circle. In cone world, space is ‘straight’, but time is ‘curved’. Near the vertex
of the cone is a tabletop three feet off the ground (the base of the cone). Now imagine
that a cup of coffee is knocked off the table. There is no ‚gravity‛ in cone world; objects
just move along the straightest-possible path (a geodesic) per Einstein.
I’ve drawn the path of the falling cup. (Remember, it’s a straight line when the paper is
unrolled laying flat on the table; you can convince yourself that this is still the shortest
possible path by taking a string and stretching it tightly between two points on the
cone.) So what happens? Well, one second after it leaves the table, the coffee cup is
about 2.75 feet above the floor; after another second, it’s about 2.25 feet above the floor;
after another second, it’s about 1 foot above the floor; and it finally hits the floor after
about 3.4 seconds. This is really important: the coffee cup ‚fell‛ to the floor NOT
because there’s gravity in cone world, but because, in cone world, time is curved!
That’s all we did: we curved time, and, voila, we have gravity!
April 14, 2009 Version 1.0 Page 8
Figure 4. Path of a falling coffee cup in Cone World. The cup ‘accelerates’ as it approaches the floor, just as if it were under the influence of gravity. Yet there is no gravity in Cone World. Like the ants on the sphere, the cup is simply following the shortest path in this world where time is curved. Note: this is not a parabola like a real-world path would be, but it looks similar.
Now because matter curves spacetime, it turns out that clocks run faster at higher
altitudes. Your GPS (global positioning system) uses a system of satellites to triangulate
your position. To do this, they need to synchronize the time between the satellite and
your navigation system; but the satellite clocks tick faster than the clock in your
navigation system. So the GPS system has to compensate for this asymmetry which is a
direct result of the non-Euclidean geometry of spacetime. If we didn’t correct for this,
your navigation system would be off by hundreds of miles after just a couple of days.
That’s it: three different ways to think about non-Euclidean geometry. One really
important example: gravity! And another important example: the Global Positioning
System.
0
0.5
1
1.5
2
2.5
3
3.5
0 1 2 3 4
Heig
ht
(feet)
Time (sec)
April 14, 2009 Version 1.0 Page 9
Figure 5. Cone World. Cut this out, roll it into a cone, and tape the edges together. Convince yourself that a geodesic (shortest distance curve) on the cone is the same as a straight line on this piece of paper before it’s rolled up. The coffee cup follows the shortest path in Cone World; the result is that it ‘accelerates’ toward the floor.