The Meaning of Einstein’s Equation

John C. Baez∗ and Emory F. Bunn†

January 4, 2006

Abstract
This is a brief introduction to general relativity, designed for both students
and teachers of the subject. While there are many excellent expositions of
general relativity, few adequately explain the geometrical meaning of the
basic equation of the theory: Einstein’s equation. Here we give a simple
formulation of this equation in terms of the motion of freely falling test
particles. We also sketch some of the consequences of this formulation and
explain how it is equivalent to the usual one in terms of tensors. Finally, we
include an annotated bibliography of books, articles and websites suitable
for the student of relativity.

1 Introduction
General relativity explains gravity as the curvature of spacetime. It’s all about
geometry. The basic equation of general relativity is called Einstein’s equation.
In units where c = 8πG = 1, it says

Gαβ = Tαβ . (1)

It looks simple, but what does it mean? Unfortunately, the beautiful geometri-
cal meaning of this equation is a bit hard to find in most treatments of relativity.
There are many nice popularizations that explain the philosophy behind rela-
tivity and the idea of curved spacetime, but most of them don’t get around to
explaining Einstein’s equation and showing how to work out its consequences.
There are also more technical introductions which explain Einstein’s equation
in detail — but here the geometry is often hidden under piles of tensor calculus.
This is a pity, because in fact there is an easy way to express the whole
content of Einstein’s equation in plain English. In fact, after a suitable prelude,
one can summarize it in a single sentence! One needs a lot of mathematics to
derive all the consequences of this sentence, but it is still worth seeing — and
we can work out some of its consequences quite easily.
∗ Department of Mathematics, University of California, Riverside, California 92521, USA.

email: baez@math.ucr.edu
† Physics Department, University of Richmond, Richmond, VA 23173, USA. email:

ebunn@richmond.edu

1
 In what follows, we start by outlining some differences between special and
general relativity. Next we give a verbal formulation of Einstein’s equation.
Then we derive a few of its consequences concerning tidal forces, gravitational
waves, gravitational collapse, and the big bang cosmology. In the last section
we explain why our verbal formulation is equivalent to the usual one in terms of
tensors. This article is mainly aimed at those who teach relativity, but except
for the last section, we have tried to make it accessible to students, as a sketch
of how the subject might be introduced. We conclude with a bibliography of
sources to help teach the subject.

2 Preliminaries
Before stating Einstein’s equation, we need a little preparation. We assume the
reader is somewhat familiar with special relativity — otherwise general relativity
will be too hard. But there are some big differences between special and general
relativity, which can cause immense confusion if neglected.
In special relativity, we cannot talk about absolute velocities, but only rel-
ative velocities. For example, we cannot sensibly ask if a particle is at rest,
only whether it is at rest relative to another. The reason is that in this theory,
velocities are described as vectors in 4-dimensional spacetime. Switching to a
different inertial coordinate system can change which way these vectors point
relative to our coordinate axes, but not whether two of them point the same
way.
In general relativity, we cannot even talk about relative velocities, except for
two particles at the same point of spacetime — that is, at the same place at the
same instant. The reason is that in general relativity, we take very seriously the
notion that a vector is a little arrow sitting at a particular point in spacetime.
To compare vectors at different points of spacetime, we must carry one over to
the other. The process of carrying a vector along a path without turning or
stretching it is called ‘parallel transport’. When spacetime is curved, the result
of parallel transport from one point to another depends on the path taken! In
fact, this is the very definition of what it means for spacetime to be curved.
Thus it is ambiguous to ask whether two particles have the same velocity vector
unless they are at the same point of spacetime.
It is hard to imagine the curvature of 4-dimensional spacetime, but it is easy
to see it in a 2-dimensional surface, like a sphere. The sphere fits nicely in
3-dimensional flat Euclidean space, so we can visualize vectors on the sphere as
‘tangent vectors’. If we parallel transport a tangent vector from the north pole
to the equator by going straight down a meridian, we get a different result than
if we go down another meridian and then along the equator:

2
Because of this analogy, in general relativity vectors are usually called ‘tangent
vectors’. However, it is important not to take this analogy too seriously. Our
curved spacetime need not be embedded in some higher-dimensional flat space-
time for us to understand its curvature, or the concept of tangent vector. The
mathematics of tensor calculus is designed to let us handle these concepts ‘in-
trinsically’ — i.e., working solely within the 4-dimensional spacetime in which
we find ourselves. This is one reason tensor calculus is so important in general
relativity.
Now, in special relativity we can think of an inertial coordinate system, or
‘inertial frame’, as being defined by a field of clocks, all at rest relative to each
other. In general relativity this makes no sense, since we can only unambiguously
define the relative velocity of two clocks if they are at the same location. Thus
the concept of inertial frame, so important in special relativity, is banned from
general relativity!
If we are willing to put up with limited accuracy, we can still talk about the
relative velocity of two particles in the limit where they are very close, since
curvature effects will then be very small. In this approximate sense, we can talk
about a ‘local’ inertial coordinate system. However, we must remember that
this notion makes perfect sense only in the limit where the region of spacetime
covered by the coordinate system goes to zero in size.
Einstein’s equation can be expressed as a statement about the relative ac-
celeration of very close test particles in free fall. Let us clarify these terms a
bit. A ‘test particle’ is an idealized point particle with energy and momentum
so small that its effects on spacetime curvature are negligible. A particle is said
to be in ‘free fall’ when its motion is affected by no forces except gravity. In
general relativity, a test particle in free fall will trace out a ‘geodesic’. This
means that its velocity vector is parallel transported along the curve it traces
out in spacetime. A geodesic is the closest thing there is to a straight line in
curved spacetime.
Again, all this is easier to visualize in 2d space rather than 4d spacetime. A
person walking on a sphere ‘following their nose’ will trace out a geodesic — that
is, a great circle. Suppose two people stand side-by-side on the equator and start
walking north, both following geodesics. Though they start out walking parallel

3
to each other, the distance between them will gradually start to shrink, until
finally they bump into each other at the north pole. If they didn’t understand
the curved geometry of the sphere, they might think a ‘force’ was pulling them
together.
Similarly, in general relativity gravity is not really a ‘force’, but just a man-
ifestation of the curvature of spacetime. Note: not the curvature of space, but
of spacetime. The distinction is crucial. If you toss a ball, it follows a parabolic
path. This is far from being a geodesic in space: space is curved by the Earth’s
gravitational field, but it is certainly not so curved as all that! The point is that
while the ball moves a short distance in space, it moves an enormous distance
in time, since one second equals about 300,000 kilometers in units where c = 1.
This allows a slight amount of spacetime curvature to have a noticeable effect.

3 Einstein’s Equation
To state Einstein’s equation in simple English, we need to consider a round ball
of test particles that are all initially at rest relative to each other. As we have
seen, this is a sensible notion only in the limit where the ball is very small. If
we start with such a ball of particles, it will, to second order in time, become
an ellipsoid as time passes. This should not be too surprising, because any
linear transformation applied to a ball gives an ellipsoid, and as the saying goes,
“everything is linear to first order”. Here we get a bit more: the relative velocity
of the particles starts out being zero, so to first order in time the ball does not
change shape at all: the change is a second-order effect.
Let V (t) be the volume of the ball after a proper time t has elapsed, as
measured by the particle at the center of the ball. Then Einstein’s equation
says:  
flow of t−momentum in t direction +
V̈  1  flow of x−momentum in x direction + 
 =−  
V t=0 2  flow of y−momentum in y direction + 
flow of z−momentum in z direction
where these flows are measured at the center of the ball at time zero, using local
inertial coordinates. These flows are the diagonal components of a 4 × 4 matrix
T called the ‘stress-energy tensor’. The components Tαβ of this matrix say how
much momentum in the α direction is flowing in the β direction through a given
point of spacetime, where α, β = t, x, y, z. The flow of t-momentum in the t-
direction is just the energy density, often denoted ρ. The flow of x-momentum
in the x-direction is the ‘pressure in the x direction’ denoted Px , and similarly
for y and z. It takes a while to figure out why pressure is really the flow of
momentum, but it is eminently worth doing. Most texts explain this fact by
considering the example of an ideal gas.
In any event, we may summarize Einstein’s equation as follows:

V̈  1
 = − (ρ + Px + Py + Pz ). (2)
V t=0 2

4
This equation says that positive energy density and positive pressure curve
spacetime in a way that makes a freely falling ball of point particles tend to
shrink. Since E = mc2 and we are working in units where c = 1, ordinary mass
density counts as a form of energy density. Thus a massive object will make
a swarm of freely falling particles at rest around it start to shrink. In short:
gravity attracts.
We promised to state Einstein’s equation in plain English, but have not done
so yet. Here it is:

Given a small ball of freely falling test particles initially at rest

with respect to each other, the rate at which it begins to shrink is
proportional to its volume times: the energy density at the center
of the ball, plus the pressure in the x direction at that point, plus
the pressure in the y direction, plus the pressure in the z direction.

One way to prove this is by using the Raychaudhuri equation, discussions

of which can be found in the textbooks by Wald and by Ciufolini and Wheeler
cited in the bibliography. But an elementary proof can also be given starting
from first principles, as we will show in the final section of this article.
The reader who already knows some general relativity may be somewhat
skeptical that all of Einstein’s equation is encapsulated in this formulation.
After all, Einstein’s equation in its usual tensorial form is really a bunch of
equations: the left and right sides of equation (1) are 4 × 4 matrices. It is hard
to believe that the single equation (2) captures all that information. It does,
though, as long as we include one bit of fine print: in order to get the full content
of the Einstein equation from equation (2), we must consider small balls with
all possible initial velocities — i.e., balls that begin at rest in all possible local
inertial reference frames.
Before we begin, it is worth noting an even simpler formulation of Einstein’s
equation that applies when the pressure is the same in every direction:

Given a small ball of freely falling test particles initially at rest

with respect to each other, the rate at which it begins to shrink is
proportional to its volume times: the energy density at the center
of the ball plus three times the pressure at that point.

This version is only sufficient for ‘isotropic’ situations: that is, those in which
all directions look the same in some local inertial reference frame. But, since
the simplest models of cosmology treat the universe as isotropic — at least
approximately, on large enough distance scales — this is all we shall need to
derive an equation describing the big bang!

4 Some Consequences
The formulation of Einstein’s equation we have given is certainly not the best
for most applications of general relativity. For example, in 1915 Einstein used

5
general relativity to correctly compute the anomalous precession of the orbit
of Mercury and also the deflection of starlight by the Sun’s gravitational field.
Both these calculations would be very hard starting from equation (2); they
really call for the full apparatus of tensor calculus.
However, we can easily use our formulation of Einstein’s equation to get a
qualitative — and sometimes even quantitative — understanding of some conse-
quences of general relativity. We have already seen that it explains how gravity
attracts. We sketch a few other consequences below. These include Newton’s
inverse-square force law, which holds in the limit of weak gravitational fields
and small velocities, and also the equations governing the big bang cosmology.

Tidal Forces, Gravitational Waves

We begin with some qualitative consequences of Einstein’s equation. Let V (t)
be the volume of a small ball of test particles in free fall that are initially at rest
relative to each other. In the vacuum there is no energy density or pressure,
so V̈ |t=0 = 0, but the curvature of spacetime can still distort the ball. For
example, suppose you drop a small ball of instant coffee when making coffee
in the morning. The grains of coffee closer to the earth accelerate towards it a
bit more, causing the ball to start stretching in the vertical direction. However,
as the grains all accelerate towards the center of the earth, the ball also starts
being squashed in the two horizontal directions. Einstein’s equation says that
if we treat the coffee grains as test particles, these two effects cancel each other
when we calculate the second derivative of the ball’s volume, leaving us with
V̈ |t=0 = 0. It is a fun exercise to check this using Newton’s theory of gravity!
This stretching/squashing of a ball of falling coffee grains is an example of
what people call ‘tidal forces’. As the name suggests, another example is the
tendency for the ocean to be stretched in one direction and squashed in the
other two by the gravitational pull of the moon.
Gravitational waves are another example of how spacetime can be curved
even in the vacuum. General relativity predicts that when any heavy object
wiggles, it sends out ripples of spacetime curvature which propagate at the speed
of light. This is far from obvious starting from our formulation of Einstein’s
equation! It also predicts that as one of these ripples of curvature passes by, our
small ball of initially test particles will be stretched in one transverse direction
while being squashed in the other transverse direction. From what we have
already said, these effects must precisely cancel when we compute V̈ |t=0 .
Hulse and Taylor won the Nobel prize in 1993 for careful observations of
a binary neutron star which is slowly spiraling down, just as general relativ-
ity predicts it should, as it loses energy by emitting gravitational radiation.
Gravitational waves have not been directly observed, but there are a number of
projects underway to detect them. For example, the LIGO project will bounce
a laser between hanging mirrors in an L-shaped detector, to see how one leg of
the detector is stretched while the other is squashed. Both legs are 4 kilometers
long, and the detector is designed to be sensitive to a 10−18 -meter change in
length of the arms.

6
Gravitational Collapse
Another remarkable feature of Einstein’s equation is the pressure term: it says
that not only energy density but also pressure causes gravitational attraction.
This may seem to violate our intuition that pressure makes matter want to
expand! Here, however, we are talking about gravitational effects of pressure,
which are undetectably small in everyday circumstances. To see this, let’s re-
store the factors of c. Also, let’s remember that in ordinary circumstances most
of the energy is in the form of rest energy, so we can write the energy density ρ
as ρm c2 , where ρm is the ordinary mass density:

V̈  1
 = −4πG(ρm + 2 (Px + Py + Pz )).
V t=0 c
On the human scale all of the terms on the right are small, since G is very small.
(Gravity is a weak force!) Furthermore, the pressure terms are much smaller
than the mass density term, since they are divided by an extra factor of c2 .
There are a number of important situations in which ρ does not dominate
over P . In a neutron star, for example, which is held up by the degeneracy
pressure of the neutronium it consists of, pressure and energy density contribute
comparably to the right-hand side of Einstein’s equation. Moreover, above a
mass of about 2 solar masses a nonrotating neutron star will inevitably collapse
to form a black hole, thanks in part to the gravitational attraction caused by
pressure. In fact, any object of mass M will form a black hole if it is compressed
to a radius smaller than its Schwarzschild radius, R = 2GM/c2 .

Newton’s Inverse-Square Force Law

A basic test of general relativity is to check that it reduces to good old Newtonian
gravity in the limit where gravitational effects are weak and velocities are small
compared to the speed of light. To do this, we can use our formulation of
Einstein’s equation to derive Newton’s inverse-square force law for a planet
with mass M and radius R. Since we can only do this when gravitational effects
are weak, we must assume that the planet’s radius is much greater than its
Schwarzschild radius: R  M in units where c = 8πG = 1. Then the curvature
of space – as opposed to spacetime – is small. To keep things simple, we make
a couple of additional assumptions: the planet has uniform density ρ, and the
pressure is negligible.
We want to derive the familiar Newtonian result
GM
a=−
r2
giving the radial gravitational acceleration a of a test particle at distance r from
the planet’s center, with r > R of course.
To do this, let S be a sphere of radius r centered on the planet. Fill the
interior of the sphere with test particles, all of which are initially at rest relative
to the planet. At first, this might seem like an illegal thing to do: we know that

7
notions like ‘at rest’ only make sense in an infinitesimal neighborhood, and r is
not infinitesimal. But because space is nearly flat for weak gravitational fields,
we can get away with this.
We can’t apply our formulation of Einstein’s equation directly to S, but we
can apply it to any infinitesimal sphere within S. In the picture below, the solid
black circle represents the planet, and the dashed circle is S. The interior of
S has been divided up into many tiny spheres filled with test particles. Green
spheres are initially inside the planet, and red spheres are outside.

Suppose we pick a tiny green sphere that lies within the planet’s volume, a
distance less than R from the center. Our formulation of Einstein’s equation
tells us that the fractional change in volume of this sphere will be

V̈  1
 =− ρ (inside the planet).
V 2
t=0

On the other hand, as we saw in our discussion of tidal effects, spheres outside
the planet’s volume will be distorted in shape by tidal effects, but remain un-
changed in volume. So, any little red sphere that lies outside the planet will
undergo no change in volume at all:

V̈ 
 =0 (outside the planet).
V
t=0

Thus, after a short time δt has elapsed, the test particles will be distributed like
this:

8
 Because the volume occupied by the spheres of particles within the planet
went down, the whole big sphere of test particles had to shrink. Now consider
the following question: what is the change in volume of the large sphere of test
particles? All the little green spheres shrank by the same fractional amount:
!
δV 1 V̈ 1
= (δt)2 = − ρ(δt)2 ,
V 2 V 4

to leading order in δt, while all the little red spheres underwent no change in
volume. So, if VP stands for the volume occupied by the test particles that were
initially inside the planet, and VS stands for the volume of the big sphere S,
then the overall change in volume is

δVS = δVP
 
δV
= VP
V
1
= − ρ(δt)2 VP
4
1
= − M (δt)2 .
4
We can also express δVS in terms of the change in the sphere’s radius, δr:

δVS = 4πr2 δr.

9
Setting the two expressions for δVS equal, we see
M
δr = − (δt)2 .
16πr2
But the change in radius δr is just the radial displacement of a test particle on
the outer edge of the sphere, so δr = 12 a(δt)2 where a is the radial acceleration
of this particle. Plugging this into the above equation, we obtain
M
a=− .
8πr2
We’ve been working all along in units in which 8πG = 1. Switching back to
conventional units, we obtain
GM
a=− 2
r
just as Newton said.

The Big Bang

We can also derive some basic facts about the big bang cosmology. Let us
assume the universe is not only expanding but also homogeneous and isotropic.
The expansion of the universe is vouched for by the redshifts of distant galaxies.
The other assumptions also seem to be approximately correct, at least when
we average over small-scale inhomogeneities such as stars and galaxies. For
simplicity, we will imagine the universe is homogeneous and isotropic even on
small scales.
An observer at any point in such a universe would see all objects receding
from her. Suppose that, at some time t = 0, she identifies a small ball B of
test particles centered on her. Suppose this ball expands with the universe,
remaining spherical as time passes because the universe is isotropic. Let R(t)
stand for the radius of this ball as a function of time. The Einstein equation
will give us an equation of motion for R(t). In other words, it will say how the
expansion rate of the universe changes with time.
It is tempting to apply equation (2) to the ball B, but we must take care.
This equation applies to a ball of particles that are initially at rest relative to
one another — that is, one whose radius is not changing at t = 0. However,
the ball B is expanding at t = 0. Thus, to apply our formulation of Einstein’s
equation, we must introduce a second small ball of test particles that are at rest
relative to each other at t = 0.
Let us call this second ball B 0 , and call its radius as a function of time r(t).
Since the particles in this ball begin at rest relative to one another, we have

ṙ(0) = 0.

To keep things simple, let us also assume that at t = 0 both balls have the exact
same size:
r(0) = R(0).

10
 Equation (2) applies to the ball B 0 , since the particles in this ball are initially
at rest relative to each other. Since the volume of this ball is proportional to
r3 , and since ṙ = 0 at t = 0, the left-hand side of equation (2) is simply
 
V̈  3r̈ 
 = .
V r t=0
t=0

Since we are assuming the universe is isotropic, we know that the various com-
ponents of pressure are equal: Px = Py = Pz = P . Einstein’s equation (2) thus
says that 
3r̈  1
 = − (ρ + 3P ).
r t=0 2
We would much prefer to rewrite this expression in terms of R rather than
r. Fortunately, we can do this. At t = 0, the two spheres have the same
radius: r(0) = R(0). Furthermore, the second derivatives are the same: r̈(0) =
R̈(0). This follows from the equivalence principle, which says that, at any given
location, particles in free fall do not accelerate with respect to each other. At
the moment t = 0, each test particle on the surface of the ball B is right next to
a corresponding test particle in B 0 . Since they are not accelerating with respect
to each other, the observer at the origin must see both particles accelerating in
the same way, so r̈(0) = R̈(0). It follows that we can replace r with R in the
above equation, obtaining

3R̈  1
 = − (ρ + 3P ).
R 2
t=0

We derived this equation for a very small ball, but in fact it applies to a ball
of any size. This is because, in a homogeneous expanding universe, the balls of
all radii must be expanding at the same fractional rate. In other words, R̈/R
is independent of the radius R, although it can depend on time. Also, there is
nothing special in this equation about the moment t = 0, so the equation must
apply at all times. In summary, therefore, the basic equation describing the big
bang cosmology is
3R̈ 1
= − (ρ + 3P ), (3)
R 2
where the density ρ and pressure P can depend on time but not on position.
Here we can imagine R to be the separation between any two ‘galaxies’.
To go further, we must make more assumptions about the nature of the
matter filling the universe. One simple model is a universe filled with pressure-
less matter. Until recently, this was thought to be an accurate model of our
universe. Setting P = 0, we obtain

3R̈ ρ
=− .
R 2

11
If the energy density of the universe is mainly due to the mass in galaxies,
‘conservation of galaxies’ implies that ρR3 = k for some constant k. This gives
3R̈ k
=− 3
R 2R
or
k
R̈ = − .
6R2
Amusingly, this is the same as the equation of motion for a particle in an at-
tractive 1/R2 force field. In other words, the equation governing this simplified
cosmology is the same as the Newtonian equation for what happens when you
throw a ball vertically upwards from the earth! This is a nice example of the
unity of physics. Since “whatever goes up must come down — unless it exceeds
escape velocity,” the solutions of this equation look roughly like this:
open

critical
R

closed

t
So, the universe started out with a big bang! It will expand forever if its current
rate of expansion is sufficiently high compared to its current density, but it will
recollapse in a ‘big crunch’ otherwise.

The Cosmological Constant

The simplified big bang model just described is inaccurate for the very early
history of the universe, when the pressure of radiation was important. Moreover,
recent observations seem to indicate that it is seriously inaccurate even in the
present epoch. First of all, it seems that much of the energy density is not
accounted for by known forms of matter. Still more shocking, it seems that
the expansion of the universe may be accelerating rather than slowing down!
One possibility is that the energy density and pressure are nonzero even for the
vacuum. For the vacuum to not pick out a preferred notion of ‘rest’, its stress-
energy tensor must be proportional to the metric. In local inertial coordinates
this means that the stress-energy tensor of the vacuum must be
 
Λ 0 0 0
 0 −Λ 0 0 
T =  0

0 −Λ 0 
0 0 0 −Λ

12
where Λ is called the ‘cosmological constant’. This amounts to giving empty
space an energy density equal to Λ and pressure equal to −Λ, so that ρ + 3P
for the vacuum is −2Λ. Here pressure effects dominate because there are more
dimensions of space than of time! If we add this cosmological constant term to
equation (3), we get
3R̈ 1
= − (ρ + 3P − 2Λ), (4)
R 2
where ρ and P are the energy density and pressure due to matter. If we treat
matter as we did before, this gives

3R̈ k
= − 3 + Λ.
R 2R
Thus, once the universe expands sufficiently, the cosmological constant becomes
more important than the energy density of matter in determining the fate of
the universe. If Λ > 0, a roughly exponential expansion will then ensue. This
seems to be happening in our universe now.

Spatial Curvature
We have emphasized that gravity is due not just to the curvature of space, but
of spacetime. In our verbal formulation of Einstein’s equation, this shows up
in the fact that we consider particles moving forwards in time and study how
their paths deviate in the space directions. However, Einstein’s equation also
gives information about the curvature of space. To illustrate this, it is easiest
to consider not an expanding universe but a static one.
When Einstein first tried to use general relativity to construct a model of
the entire universe, he assumed that the universe must be static — although
he is said to have later described this as “his greatest blunder”. As we did in
the previous section, Einstein considered a universe containing ordinary matter
with density ρ, no pressure, and a cosmological constant Λ. Such a universe can
be static — the galaxies can remain at rest with respect to each other — only if
the right-hand side of equation (4) is zero. In such a universe, the cosmological
constant and the density must be carefully ‘tuned’ so that ρ = 2Λ. It is tempting
to conclude that spacetime in this model is just the good old flat Minkowski
spacetime of special relativity. In other words, one might guess that there are no
gravitational effects at all. After all, the right-hand side of Einstein’s equation
was tuned to be zero. This would be a mistake, however. It is instructive to see
why.
Remember that equation (2) contains all the information in Einstein’s equa-
tion only if we consider all possible small balls. In all of the cosmological ap-
plications so far, we have applied the equation only to balls whose centers were
at rest with respect to the local matter. It turns out that only for such balls is
the right-hand side of equation (2) zero in the Einstein static universe.
To see this, consider a small ball of test particles, initially at rest relative to
each other, that is moving with respect to the matter in the universe. In the

13
local rest frame of such a ball, the right-hand side of equation (2) is nonzero. For
one thing, the pressure due to the matter no longer vanishes. Remember that
pressure is the flux of momentum. In the frame of our moving sphere, matter
is flowing by. Also, the energy density goes up, both because the matter has
kinetic energy in this frame and because of Lorentz contraction. The end result,
as the reader can verify, is that the right-hand side of equation (2) is negative
for such a moving sphere. In short, although a stationary ball of test particles
remains unchanged in the Einstein static universe, our moving ball shrinks!
This has a nice geometric interpretation: the geometry in this model has
spatial curvature. As we noted in section 2, on a positively curved surface such
as a sphere, initially parallel lines converge towards one another. The same
thing happens in the three-dimensional space of the Einstein static universe. In
fact, the geometry of space in this model is that of a 3-sphere. This picture
illustrates what happens:

One dimension is suppressed in this picture, so the two-dimensional spherical

surface shown represents the three-dimensional universe. The small shaded
circle on the surface represents our tiny sphere of test particles, which starts
at the equator and moves north. The sides of the sphere approach each other
along the dashed geodesics, so the sphere shrinks in the transverse direction,
although its diameter in the direction of motion does not change.
As an exercise, the reader who wants to test his understanding can fill in the
mathematical details in this picture and determine the radius of the Einstein
static universe in terms of the density. Here are step-by-step instructions:
• Imagine an observer moving at speed v through a cloud of stationary par-
ticles of density ρ. Use special relativity to determine the energy density
and pressure in the observer’s rest frame. Assume for simplicity that the
observer is moving fairly slowly, so keep only the lowest-order nonvanishing
term in a power series in v.
• Apply equation (2) to a sphere in this frame, including the contribu-
tion due to the cosmological constant (which is the same in all reference
frames). You should find that the volume of the sphere decreases with
V̈ /V ∝ −ρv 2 to leading order in v.

14
 • Suppose that space in this universe has the geometry of a large 3-sphere of
radius RU . Show that the radii in the directions
 transverse to the motion

start to shrink at a rate given by (R̈/R) = −v 2 /RU
2
. (If, like most
t=0
people, you are better at visualizing 2-spheres than 3-spheres, do this step
by considering a small circle moving on a 2-sphere, as shown above, rather
than a small sphere moving on a 3-sphere. The result is the same.)
• Since our little sphere is shrinking in two dimensions, its volume changes
at a rate V̈ /V = 2R̈/R. Use Einstein’s equation to relate the radius RU
of the universe to the density ρ.
p
The final answer is RU = 2/ρ. As you can confirm in standard textbooks,
this is correct.
Spatial curvature like this shows up in the expanding cosmological models
described earlier in this section as well. In principle, the curvature radius can be
found from our formulation of Einstein’s equation by similar reasoning in these
expanding models. In fact, however, such a calculation is extremely messy. Here
the apparatus of tensor calculus comes to our rescue.

5 The Mathematical Details

To see why equation (2) is equivalent to the usual formulation of Einstein’s
equation, we need a bit of tensor calculus. In particular, we need to understand
the Riemann curvature tensor and the geodesic deviation equation. For a de-
tailed explanation of these, the reader must turn to some of the texts in the
bibliography. Here we briefly sketch the main ideas.
When spacetime is curved, the result of parallel transport depends on the
path taken. To quantify this notion, pick two vectors u and v at a point p
in spacetime. In the limit where  → 0, we can approximately speak of a
‘parallelogram’ with sides u and v. Take another vector w at p and parallel
transport it first along v and then along u to the opposite corner of this
parallelogram. The result is some vector w1 . Alternatively, parallel transport
w first along u and then along v. The result is a slightly different vector, w2 :
w1 w2

εv

p
εu

15
The limit
w2 − w 1
= R(u, v)w
lim (5)
2 →0
is well-defined, and it measures the curvature of spacetime at the point p. In
local coordinates we can write it as
α
R(u, v)w = Rβγδ uβ v γ w δ ,
α
where as usual we sum over repeated indices. The quantity Rβγδ is called the
‘Riemann curvature tensor’.
We can use this tensor to compute the relative acceleration of nearby par-
ticles in free fall if they are initially at rest relative to one another. Consider
two freely falling particles at nearby points p and q. Let v be the velocity of the
particle at p, and let u be the vector from p to q. Since the two particles start
out at rest relative to one other, the velocity of the particle at q is obtained by
parallel transporting v along u.
Now let us wait a short while. Both particles trace out geodesics as time
passes, and at time  they will be at new points, say p0 and q 0 . The point p0 is
displaced from p by an amount v, so we get a little parallelogram, exactly as
in the definition of the Riemann curvature:
p q

εv

p q
εu

Next let us compute the new relative velocity of the two particles. To compare
vectors we must carry one to another using parallel transport. Let v1 be the
vector we get by taking the velocity vector of the particle at p0 and parallel
transporting it to q 0 along the top edge of our parallelogram. Let v2 be the
velocity of the particle at q 0 . The difference v2 − v1 is the new relative velocity.
Here is a picture of the whole situation:
v1 v2

p q

εv

p q
εu

16
The vector v is depicted as shorter than v for purely artistic reasons.
It follows that over this passage of time, the average relative acceleration of
the two particles is a = (v2 − v1 )/. By equation (5),
v2 − v 1
lim = R(u, v)v,
→0 2
so
a
lim = R(u, v)v.
→0 
This is called the ‘geodesic deviation equation’. From the definition of the
Riemann curvature it is easy to see that R(u, v)w = −R(v, u)w, so we can also
write this equation as
aα α
lim = −Rβγδ v β uγ v δ . (6)
→0 
Using this equation we can work out the second time derivative of the volume
V (t) of a small ball of test particles that start out at rest relative to each other.
As we mentioned earlier, to second order in time the ball changes to an ellipsoid.
Furthermore, since the ball starts out at rest, the principal axes of this ellipsoid
don’t rotate initially. We can therefore adopt local inertial coordinates in which,
to second order in t, the center of the ball is at rest and the three principal axes
of the ellipsoid are aligned with the three spatial coordinates. Let r j (t) represent
the radius of the jth axis of the ellipsoid as a function of time. If the ball’s initial
radius is , then
1
rj (t) =  + aj t2 + O(t3 ),
2
or in other words,
r̈j aj
lim j = lim .
t→0 r t→0 
j
Here the acceleration a is given by equation (6), with u being a vector of length
 in the jth coordinate direction and v being the velocity of the ball, which is a
unit vector in the time direction. In other words,
r̈j (t) j j
lim = −Rβjδ v β v δ = −Rtjt .
t→0 r j (t)

No sum over j is implied in the above expression.

SincePthe volume of our ball is proportional to the product of the radii,
V̈ /V → j r̈j /rj as t → 0, so

V̈  α
lim  = −Rtαt ,
V →0 V t=0

where now a sum over α is implied. The sum over α can range over all four
coordinates, not just the three spatial ones, since the symmetries of the Riemann
t
tensor demand that Rttt = 0.
The right-hand side is minus the time-time component of the ‘Ricci tensor’
α
Rβδ = Rβαδ .

17
That is,
V̈ 
lim  = −Rtt (7)
V →0 V t=0

in local inertial coordinates where the ball starts out at rest.

In short, the Ricci tensor says how our ball of freely falling test particles
starts changing in volume. The Ricci tensor only captures some of the informa-
tion in the Riemann curvature tensor. The rest is captured by something called
the ‘Weyl tensor’, which says how any such ball starts changing in shape. The
Weyl tensor describes tidal forces, gravitational waves and the like.
Now, Einstein’s equation in its usual form says

Gαβ = Tαβ .

Here the right side is the stress-energy tensor, while the left side, the ‘Einstein
tensor’, is just an abbreviation for a quantity constructed from the Ricci tensor:
1
Gαβ = Rαβ − gαβ Rγγ .
2
Thus Einstein’s equation really says
1
Rαβ − gαβ Rγγ = Tαβ . (8)
2
This implies
α 1
Rα − gαα Rγγ = Tαα ,
2
but gαα = 4, so
α
−Rα = Tαα .
Plugging this in equation (8), we get
1
Rαβ = Tαβ − gαβ Tγγ . (9)
2
This is equivalent version of Einstein’s equation, but with the roles of R and T
switched! The good thing about this version is that it gives a formula for the
Ricci tensor, which has a simple geometrical meaning.
Equation (9) will be true if any one component holds in all local inertial
coordinate systems. This is a bit like the observation that all of Maxwell’s
equations are contained in Gauss’s law and ∇ · B = 0. Of course, this is only
true if we know how the fields transform under change of coordinates. Here we
assume that the transformation laws are known. Given this, Einstein’s equation
is equivalent to the fact that
1
Rtt = Ttt − gtt Tγγ (10)
2

18
in every local inertial coordinate system about every point. In such coordinates
we have  
−1 0 0 0
 0 1 0 0 
g=   (11)
0 0 1 0 
0 0 0 1
so gtt = −1 and
Tγγ = −Ttt + Txx + Tyy + Tzz .
Equation (10) thus says that
1
Rtt = (Ttt + Txx + Tyy + Tzz ).
2
By equation (7), this is equivalent to

V̈  1
lim  = − (Ttt + Txx + Tyy + Tzz ).
V →0 V t=0 2
As promised, this is the simple, tensor-calculus-free formulation of Einstein’s
equation.

6 Bibliography
Websites
There is a lot of material on general relativity available online, including many
books, such as this:
Lecture Notes on General Relativity, S. M. Carroll,
http://arxiv.org/abs/gr-qc/9712019
The free online journal Living Reviews in Relativity is an excellent way to learn
more about many aspects of relativity. One can access it at:
Living Reviews in Relativity, http://www.livingreviews.org
Part of learning relativity is working one’s way through certain classic confu-
sions. The most common are dealt with in the “Relativity and Cosmology”
section of this site:
Frequently Asked Questions in Physics, edited by D. Koks
http://math.ucr.edu/home/baez/physics/

Nontechnical Books
Before diving into the details of general relativity, it is good to get oriented
by reading some less technical books. Here are four excellent ones written by
leading experts on the subject:

19
General Relativity from A to B, R. Geroch (University of Chicago Press, Chicago,
1981).
Black Holes and Time Warps: Einstein’s Outrageous Legacy, K. S. Thorne (W.
W. Norton, New York, 1995).
Gravity from the Ground Up: An Introductory Guide to Gravity and General
Relativity, B. F. Schutz (Cambridge University Press, Cambridge 2003).
Space, Time, and Gravity: the Theory of the Big Bang and Black Holes, R. M.
Wald (University of Chicago Press, Chicago, 1992).

Special Relativity
Before delving into general relativity in a more technical way, one must get up
to speed on special relativity. Here are two excellent texts for this:
Introduction to Special Relativity, W. Rindler (Oxford University Press, Oxford,
1991).
Spacetime Physics: Introduction to Special Relativity, E. F. Taylor and J. A.
Wheeler (W. H. Freeman, New York, 1992).

Introductory Texts
When one is ready to tackle the details of general relativity, it is probably good
to start with one of these textbooks:
Introducing Einstein’s Relativity, R. A. D’Inverno (Oxford University Press,
Oxford, 1992).
Gravity: An Introduction to Einstein’s General Relativity, J. B. Hartle (Addison-
Wesley, New York, 2002).
Introduction to General Relativity, L. Hughston and K. P. Tod (Cambridge
University Press, Cambridge, 1991).
A First Course in General Relativity, B. F. Schutz (Cambridge University Press,
Cambridge, 1985).
General Relativity: An Introduction to the Theory of the Gravitational Field, H.
Stephani (Cambridge University Press, Cambridge, 1990).

More Comprehensive Texts

To become a expert on general relativity, one really must tackle these classic
texts:
Gravitation, C. W. Misner, K. S. Thorne, and J. A. Wheeler (W. H. Freeman,
New York, 1973).
General Relativity, R. M. Wald (University of Chicago Press, Chicago, 1984).

20
Along with these textbooks, you’ll want to do lots of problems! This book is a
useful supplement:
Problem Book in Relativity and Gravitation, A. Lightman and R. H. Price
(Princeton University Press, Princeton, 1975).

Experimental Tests
The experimental support for general relativity up to the early 1990s is sum-
marized in:
Theory and Experiment in Gravitational Physics, Revised Edition, C. M. Will
(Cambridge University Press, Cambridge, 1993).
A more up-to-date treatment of the subject can be found in:
“The Confrontation between General Relativity and Experiment,” C. M. Will,
Living Reviews in Relativity 4, 4 (2001). Available online at:
http://www.livingreviews.org/lrr-2001-4

Differential Geometry
The serious student of general relativity will experience a constant need to
learn more tensor calculus — or in modern terminology, ‘differential geometry’.
Some of this can be found in the texts listed above, but it is also good to read
mathematics texts. Here are a few:
Gauge Fields, Knots and Gravity, J. C. Baez and J. P. Muniain (World Scientific,
Singapore, 1994).
An Introduction to Differentiable Manifolds and Riemannian Geometry, W. M.
Boothby (Academic Press, New York, 1986).
Semi-Riemannian Geometry with Applications to Relativity, B. O’Neill (Aca-
demic Press, New York, 1983).

Specific Topics
The references above are about general relativity as a whole. Here are some
suggested starting points for some of the particular topics touched on in this
article.

The meaning of Einstein’s equation. Feynman gives a quite different ap-

proach to this in:
The Feynman Lectures on Gravitation, R. P. Feynman et al. (Westview Press,
Boulder, Colorado, 2002).
His approach focuses on the curvature of space rather than the curvature of
spacetime.

21
The Raychaudhuri equation. This equation, which is closely related to
our formulation of Einstein’s equation, is treated in some standard textbooks,
including the one by Wald mentioned above. A detailed discussion can be found
in
Gravitation and Inertia, I. Ciufolini and J. A. Wheeler (Princeton University
Press, Princeton, 1995).

Gravitational waves. Here are two nontechnical descriptions of the binary

pulsar work for which Hulse and Taylor won the Nobel Prize:
“The Binary Pulsar: Gravity Waves Exist,” C. M. Will, Mercury, Nov-Dec
1987, p. 162.
“Gravitational Waves from an Orbiting Pulsar,” J. M. Weisberg, J. H. Taylor,
and L. A. Fowler, Scientific American, Oct 1981, p. 74.
Here is a review article on the ongoing efforts to directly detect gravitational
waves:
“Detection of Gravitational Waves,” J. Lu, D. G. Blair, and C. Zhao, Reports
on Progress in Physics, 63, 1317-1427 (2000).
Some present and future experiments to detect gravitational radiation are de-
scribed here:
LIGO Laboratory Home Page, http://www.ligo.caltech.edu/
The Virgo Project, http://www.virgo.infn.it/
Laser Interferometer Space Antenna, http://lisa.jpl.nasa.gov/

Black holes. Astrophysical evidence that black holes exist is summarized in:
“Evidence for Black Holes,” M. C. Begelman, Science, 300, 1898-1903 (2003).
A less technical discussion of the particular case of the supermassive black hole
at the center of our Milky Way Galaxy can be found here:
The Black Hole at the Center of Our Galaxy, F. Melia (Princeton University
Press, Princeton, 2003).

Cosmology. There are lots of good popular books on cosmology. Since the
subject is changing rapidly, pick one that is up to date. At the time of this
writing, we recommend:
The Extravagant Universe: Exploding Stars, Dark Energy, and the Accelerating
Cosmos, R. P. Kirshner (Princeton University Press, Princeton, 2002).
A good online source of cosmological information is:
Ned Wright’s Cosmology Tutorial,
http://www.astro.ucla.edu/∼wright/cosmolog.htm

22
The following cosmology textbooks are arranged in increasing order of technical
difficulty:
Cosmology: The Science of the Universe, 2nd ed., E. Harrison (Cambridge Uni-
versity Press, Cambridge, 2000).
Cosmology: a First Course, M. Lachièze-Rey (Cambridge University Press,
Cambridge, 1995).
Principles of Physical Cosmology, P. J. E. Peebles (Princeton University Press,
Princeton, 1993).
The Early Universe, E. W. Kolb and M. S. Turner (Addison–Wesley, New York,
1990).
The Large-Scale Structure of Spacetime, S. W. Hawking and G. F. R. Ellis
(Cambridge University Press, Cambridge, 1975).

Acknowledgment
E.F.B. is supported by National Science Foundation grant 0233969.

23