CHAPTER 1

Speeding
The important thing is not to stop questioning. Curiosity has its own
reason for existing. One cannot help but be in awe when he [or she]
contemplates the mysteries of eternity, of life, of the marvelous structure
of reality. It is enough if one tries merely to comprehend a little of this
mystery every day. Never lose a holy curiosity.
—Albert Einstein

1 Special Relativity
Key idea: Concepts useful in exploring the very fast help us to examine spacetime
near very massive objects.
We use relativity to explore the boundaries of Nature. Special relativity Special relativity: fast objects
describes the very fast. General relativity—the Theory of Gravitation— General relativity: spacetime
near massive objects
describes matter and motion near massive objects: stars, galaxies, black
holes. General relativity also describes the Universe as a whole. This chap-
ter discusses a few key concepts of special relativity useful in exploring
general relativity. The treatment here is not designed to be an introduction
to special relativity; for introductory treatments see Section 11, Readings
in Special Relativity, and detailed references to our own introductory
treatment at the end of each section.

2 Wristwatch Time
Everyone agrees on the wristwatch time between two events.
What is the root of relativity? Is there a single, simple idea that launches us Begin relativity with
along the road to understanding? Alice's adventures in wonderland begin wristwatch time between
two ticks.
when a rabbit rushes past her carrying a pocket watch. Our adventure in
relativity begins when a small stone flies past us wearing a wristwatch.
The wristwatch ticks once at #1 and once at #2 (Figure 1). Wristwatch ticks
may be one second apart—or one microsecond. Measure the distance s
and time t between these ticks in a particular free-float or inertial refer-
ence frame. (The free-float frame is described in Section 8. Briefly, it is one
in which Newton’s first law holds: a free particle at rest remains at rest and
one in motion continues that motion at constant speed in a straight line.)
Special relativity warns us that a different observer passing us in uniform
relative motion typically records a different value of spatial separation s
and a different value of time lapse t between these two ticks. That is the
bad news. The good news is a central finding of special relativity:

Section 1 Special Relativity 1-1

 ick #2

Tick #1

Figure 1 Straight-line uniform-speed trajectory of a stone through

space. The stone wears a wristwatch that ticks and emits a flash at
#1 and then ticks again and emits a second flash at #2. These two
ticks are a distance s apart and have a time separation t as
measured in the frame of reference for which this diagram is
drawn.

All inertial observers, whatever their state of relative motion and what-
ever values they measure for s and t, agree on the value of the time τ
between ticks as recorded on the wristwatch carried by the stone. The for-
mula is simple:
2 2 2
τ = t –s [1]
Define wristwatch time. We use the Greek letter τ (tau) for the wristwatch time between these two
watch ticks. The wristwatch time is often called the proper time or, more
formally, the timelike spacetime interval (“timelike” because the time
separation t is greater than the space separation s). All observers agree on
the value of the wristwatch time between two events. In contrast, the
value of t and the value of s between these events will typically differ from
frame to frame. Call t the frame time and s the frame distance between
this pair of events. Wristwatch time τ can be used to describe the separa-
tion between any pair of events for which t is greater than s. It tells the
observer in any frame what the time lapse will be on a wristwatch that
moves uniformly from one event to the other.
Measure space and time For simplicity, the units of space and time are the same, such as light-years
in the same units. and years, or meters of distance and meters of light-travel time. In both
cases the speed of light c is the conversion factor between measures of
space and time. For example, the relation between seconds and meters of
light-travel time is
t ( in meters ) = ct second [2]
The metric: Key to all relativity Equation [1], which connects the wristwatch time between two adjacent
ticks to their space and time separations in a given frame, is called the
metric. The metric (with a minus sign between squared quantities) tells us
the separation between events in spacetime, just as the Pythagorean Theo-
rem (with a plus sign between squared quantities) tells us the distance
between points in a space described by Euclidean geometry. The metric is

1-2 CHAPTER 1 Speeding

central in both special and general relativity. In describing physical sys-
tems for which it can be derived, the metric provides the answer to every
possible question about (nonquantum) features of spacetime. And with a
simple extension it also predicts the trajectories of particles and light.
The fact that all free-float observers agree on the wristwatch time τ earns it Wristwatch time is an
the label invariant. Invariant means that all observers calculate the same INVARIANT.

value, independent of reference frame. In relativity every invariant quan-

tity is a diamond, to be treasured.
How fast does the stone travel between ticks? The stone’s speed depends Velocity v is a fraction of the
on the reference frame. For the frame of Figure 1, the speed (assumed to be speed of light.

constant) is v = s/t. Measure distance s and time lapse t in the same unit.
For example, a spaceship travels half a light-year of distance during one
year of time; its speed is then 0.5 year/year and the units cancel. As
another example, if an elementary particle moves 0.7 meter in one meter
of light-travel time its speed is 0.7. Hence the speed v has no units. In this
book the symbol v represents the speed of an object as a fraction of the
speed of light.

Fuller Explanations: Spacetime Physics, Chapter 1, Spacetime: Overview;

Chapter 3, Same Laws for All; Chapter 6, Regions of Spacetime.

SAMPLE PROBLEM 1 Wristwatch Times

PROBLEM 1A. An unpowered spaceship moving at constant PROBLEM 1B. An elementary particle is created in the target
speed travels 3 light-years in 5 years, this time and distance of a particle accelerator and arrives at a detector 4 meters
measured in the rest frame of our Sun. What is the time lapse away and 5 meters of light-travel time later, as measured in
for this trip as recorded on a clock carried with the spaceship? the laboratory. The wristwatch of the elementary particle
records what time between creation and detection?
SOLUTION 1A. The two events that start and end the space-
ship's journey are separated in the Sun frame by s = 3 light- SOLUTION 1B. The events of creation and detection are sep-
years and t = 5 years. Equation [1] gives the resulting wrist- arated in the laboratory frame by s = 4 meters and t = 5
watch time: meters of light-travel time. Equation [1] tells us that
2 2 2 2 2 2 2 2 2 2 2 2
τ = t – s = 5 – 3 = 25 – 9 = 16 years [3] τ = t – s = 5 – 4 = 25 – 16 = 9 meters [4]
τ = 4 years τ = 3 meters

which is less than the time lapse as measured in the Sun Again, the wristwatch time for the particle is less than the
frame. time recorded in the laboratory frame.

3 Proper Distance
Everyone agrees on the proper distance between two events.
Two firecrackers explode 1 meter apart and at the same time, as measured in
a particular free-float frame. In this frame these explosions are
simultaneous. No stone can travel fast enough to be present at both of
these explosions without moving at an infinite velocity, which is impossi-
ble. Therefore equation [1] is useless to define a wristwatch time τ between
these two events.

Section 3 Proper Distance 1-3

 SAMPLE PROBLEM 2 Speeding to Andromeda
At approximately what constant speed v must a spaceship Now, we assumed that v is very close to the speed of light. It
travel so that the occupants age only 1 year during a trip from follows that the time t for the trip in the Sun frame is very
Earth to the Andromeda galaxy? Andromeda lies 2 million close to the time that light takes to make the trip: 2 million
light-years distant from Earth. years. Substitute this value and also demand that the wrist-
watch time on the spaceship (the aging of the occupants
during their trip) be τ = 1 year. The result is
SOLUTION
2 2
τ 1 year
The word approximately in the statement of the problem tells 1 – v ≈ ------- = --------------------------------------------
2 12 2
us that we can make some assumptions. We assume that a 2t 2 × 4 × 10 year [7]
single free-float frame can stretch all the way from Sun to – 12
10 – 13
Andromeda, so special relativity applies. We also predict that = ------------- = 1.25 × 10
8
the speed v of the spaceship measured in the Sun frame is
very close to unity, the speed of light. That allows us to set
(1 + v) ≈ 2 in the last of the following steps: Equation [7] expresses the result in sensible scientific nota-
tion. However, your friends may be more impressed if you
report the speed as a fraction of the speed of light:
2 s 
2
2 2 2 2 2
τ = t – s = t  1 – ----- = t ( 1 – v ) v = 0.999999999999875. This result justifies the assumptions
 t 
2
[5] we made about the value of v and the time for the trip as
2 2 measured in the Sun frame. Additional question: What dis-
= t ( 1 + v ) ( 1 – v ) ≈ 2t ( 1 – v )
tance does the spaceship rider measure between Earth and
Andromeda?
Equate the first and last expressions to obtain
2
τ
1 – v ≈ ------- [6]
2
2t

Use simultaneous explosions Simultaneous explosions are thus useless for measuring time. But they are
to measure length of a rod. perfect for measuring length. Question: How do you measure the length of
a rod, whether it is moving or at rest in your frame? Answer: Set off two
firecrackers at the two ends and at the same time (t = 0) in your frame. Then
define the rod’s length in your frame as the distance s between this pair of
explosions.
Special relativity warns us that a different observer passing us in uniform
relative motion typically will not agree that the two firecrackers exploded
at the same time. That is the bad news (and the idea most difficult to
understand in all of special relativity). But there is good news: All inertial
observers, whatever their state of relative motion, can calculate the dis-
tance σ between explosions as recorded in the frame in which they do
occur simultaneously. The new metric is a variation of the old metric [1]:
2 2 2
σ = s –t [8]
Proper distance is an The Greek letter σ (sigma) labels what we call the proper distance
INVARIANT. between such events or, more formally, the spacelike spacetime interval
(“spacelike” because the space separation s is greater than the time separa-
tion t). All free-float observers agree on the value of the proper distance—
the proper distance is an invariant. In contrast, the value of t and the value
of s between these events typically differ, respectively, as measured in dif-
ferent frames. Proper distance σ can be used to describe the separation

1-4 CHAPTER 1 Speeding

between any pair of events for which s is greater than t. It tells the observer
in any frame what the distance σ is between the events as measured in a
frame in which they occur at the same time.

We attach special significance to the length of a rod measured in the frame

in which it is at rest. Let a firecracker explode at each end of a rod at the
same time in its rest frame. We call the distance between these explosions
the proper length of the rod. Any other inertial observer, whatever her
state of relative motion, can calculate the proper length of the rod from
equation [8] using the time t and distance s that she measures between
these particular explosions in her own reference frame.

As in equation [1], the units of space and time in equation [8] are the same,
such as light-years and years—or meters of distance and meters of light-
travel time.

The name spacetime interval is the collective name for the timelike space-
time interval (equation[1]) and the spacelike spacetime interval (equation
[8]).

Fuller Explanations: What happens to equations [1] and [8] when s and t
have the same magnitude? Find the answer in Spacetime Physics, Chapter 6,
Regions of Spacetime.

4 The Principle of Extremal Aging

The Twin Paradox leads to a definition of natural motion.

To get ready for curved spacetime (whatever that may mean), look further
at the motion of a free particle in flat spacetime, the arena of the free-float
frame (Section 8) in which special relativity correctly describes motion.
How does a free particle move in flat spacetime? We say: “What a ridicu-
lous question! Everyone knows that a free particle moves with constant
speed in a straight line—at least as observed in a free-float frame.” Ah yes,
but why does a free particle move straight with constant speed? What lies
behind this motion? Our answer for flat spacetime will be a trial run for
the description of motion in curved spacetime, the arena of general
relativity.
A deep description of motion arises from the famous Twin Paradox. Twin Paradox predicts the
Recall that one identical twin relaxes on Earth while her twin sister franti- motion of a free particle.

cally travels to a distant star and returns. When the two meet again, the
stay-at-home twin has aged more than her traveling sister. (This outcome
can be predicted by extending Sample Problems 1 and 2 to include return
of the traveler to the point of origin.) Upon being reunited, the “identical
twins” are no longer identical. Very strange! But (almost) no one who has
studied relativity doubts the difference in age, and experiments with fast-
moving particles verify it.
Which twin has the motion we can call natural? Isaac Newton has a defini- Being at rest is one natural
tion of natural motion. He would say, “A twin at rest tends to remain at motion.

Section 4 The Principle of Extremal Aging 1-5

 SAMPLE PROBLEM 3 How Slow Is “Speeding”?

A. Answer “yes” or “no” to questions (a) through (e): Equate the first and the last of the expressions in the last line
of [9] and multiply through by v/(2s) to obtain
Is the stay-at home twin older when they get together
again if the traveling twin sv
t – τ ≈ ----- [10]
2
(a) streaks to the Andromeda galaxy (2 million light-years
distant) and back? We need to express the velocity v as a fraction of the speed of
light. A speed of 1 meter per second is equal to
(b) soars to Alpha Centauri (4 light-years distant) and
back?
1 meter/second 1 meter/second
v = ------------------------------------ = ---------------------------------------------------
c 8
(c) flies to the planet Pluto and back? 3 × 10 meter/second [11]
–9
= 3.3 × 10
(d) hurries to Earth's Moon and back?

(e) strolls next door to the neighbor's house and back? Substitute this value of v into equation [10] to yield the time
difference for one leg of the round trip:
B. In case (e) of part A, what is the approximate difference in
aging between the twins if the traveling twin strolls at –9
100 meters × 3.3 × 10
1 meter per second and the next door neighbor's house is t – τ ≈ ----------------------------------------------------------
2
100 meters away? [12]
–7
3.3 × 10
= ------------------------- meters of light-travel time
SOLUTION 2

A. In principle, one should reply “yes”—the stay-at-home The round trip difference will be twice this value, or
twin will be older—for all cases in part A. Part B examines 3.3 × 10–7 meters of light-travel time. Divide the result by the
the actual value of the aging difference for small relative speed of light to obtain the time difference in seconds:
velocity.
–7
3.3 × 10 meter
 time difference ≈ ---------------------------------------------------
B. Solve equation [1] for s2 and apply it to the outward trip  for round trip  8
3 × 10 meter/second [13]
from the twins’ house to the neighbor's house. The word
– 15
approximately in the statement of the problem gives us = 1.1 × 10 second
permission to make assumptions.
(This result justifies our assumption that the two times t and τ
Usually we do not notice results of the Twin Paradox in our are very nearly equal.) So after her stroll next door and back,
everyday lives, so it seems reasonable to assume that the the traveling twin will be approximately 10–15 seconds
frame time t is very nearly the same as the wristwatch time τ younger than her stay-at-home sister. To measure this tiny
for the stroll next door. This allows us to set (t + τ) ≈ 2t in the time difference exceeds the sensitivity of even the most accu-
following steps. We also set t = s/v in one of the steps. rate atomic clock. That is why we do not notice relativistic
2 2 2 effects in our everyday lives! Nevertheless, Nature witnesses
s = t – τ = (t + τ)(t – τ) the difference by selecting the stay at home twin as the one
s [9]
2
s ≈ 2t ( t – τ ) = 2 --- ( t – τ ) whose motion (or whose lack of motion in this frame) is
v natural.

rest.” So it is the stay-at-home twin who moves in the natural way. In con-
trast, the out-and-back twin suffers the forces required to change her state
of motion—from outgoing motion to incoming motion—so that the two
sisters can meet again in person. The motion of the traveling twin is
forced, not natural.
Moving uniformly is another Viewed from a second relatively moving free-float frame, the stay-at-home
natural motion. twin moves with constant speed in a straight line. Hers is also natural
motion. Newton would say, “A twin in motion tends to continue this
motion at constant speed in a straight line.” So the motion of the stay-on-

1-6 CHAPTER 1 Speeding

Earth twin is also natural from the viewpoint of a second frame in uniform
relative motion—or from any frame moving uniformly with respect to the
original frame. In any such frame, the time lapse on the wristwatch of the
stay-at-home twin can be calculated from the metric (equation [1]).
The lesson of the Twin Paradox is that the natural motion of a free object Natural motion in general:
between two events in flat spacetime is the one for which the wristwatch Extremal wristwatch time

worn by the object has a maximum time reading between those two
events. Purists insist that we say not maximum reading but rather extremal
reading: either maximum or minimum. This book contains only examples
of maximum wristwatch time for natural motion. Still, let’s try to keep the
purists happy! Replace the two words maximum and minimum with the
single word extremal. The result is the Principle of Extremal Aging.

Principle of Extremal Aging: The path a free object takes between two events
in spacetime is the path for which the time lapse between these events, recorded
on the object’s wristwatch, is an extremum.
It turns out that the Principle of Extremal Aging describes motion even Principle of Extremal Aging:
when spacetime is not flat. The Principle of Extremal Aging accompanies works for general relativity
too.
us into curved spacetime, into the realm of general relativity. But for now
we stay in flat spacetime and use the Principle of Extremal Aging to derive
relativistic expressions for energy and momentum.

5 Energy in Special Relativity

The Principle of Extremal Aging tells us the energy of a free particle.
Combining the metric (Section 2) with the Principle of Extremal Aging Derive energy from the metric
(Section 4) leads to the relativistic expression for energy in flat space- plus the Principle of Extremal
Aging.
time—the formula for energy used in special relativity. Here is the plan in
outline: A free stone following its natural path carries a wristwatch that
emits three flashes. We consider all three flashes to be fixed in space and
the emission times for the first and last flashes also to be fixed. We then
adjust the time of the middle flash so that the wristwatch time from the first
flash to the last flash is an extremum. The outcome is the expression for a
quantity that is the same along every segment of the path—this quantity is
conserved. We identify the conserved quantity as the energy. Now fill in
some details.
Think of a free stone flying along a straight line in space as observed in an Three flashes: When will the
inertial frame (Figures 2 and 3). The stone emits three flashes #1, #2, and middle flash occur?

#3 bracketing two adjacent segments of its trajectory, segments labeled A

and B in the figures. These two segments need not be the same length. Fix
the positions of all three flash emissions in space, fix also the times for flash
emissions #1 and #3, then ask: At what time t will the free stone pass loca-
tion #2 and emit the second flash? Find this intermediate time t by
demanding that the total wristwatch time from #1 to #3 be an extremum.
In other words, use the Principle of Extremal Aging to find the time for the
middle flash. The result leads to a conserved quantity, the energy of the
stone.

Section 5 Energy in Special Relativity 1-7

 Figure 2 Three alternative cases of a stone moving along a straight line in space
as it emits three flashes, #1, #2, and #3. The space locations of emissions are the
same in all three cases, as are the times of first and last emissions #1 and #3. But
emission time for the middle flash #2 is different for the three cases. We ask: At
what time will a free stone following a natural path pass the intermediate point
and emit flash #2? We answer this question by demanding that the total
wristwatch time τ from first to last flash emissions be an extremum. From this
requirement comes an expression for the energy of the stone as a constant of the
motion.

Now for the full step-by-step derivation.

1. Let t be the frame time between flash #1 and flash #2 and let s be the
frame distance between these two flashes. Then the metric [1] tells us
that the wristwatch time τA along segment A is
2 2 1⁄2
τA = ( t – s ) [14]

To prepare for the derivative that leads to extremal aging, differentiate

this expression with respect to the intermediate time t:
dτ A t t
--------- = ---------------------------- = ------ [15]
dt 2 2 1 ⁄ 2 τA
(t – s )
2. Next, let T be the fixed time between flashes #1 and #3 and S be the
fixed distance between them. Then the frame time between flash #2 and
flash #3 is ( T – t) and the frame distance between them is (S – s). There-
fore the wristwatch time τB along segment B is

1-8 CHAPTER 1 Speeding

Figure 3 Three alternative cases of a stone moving along a straight line in space as it emits
three flashes, #1, #2, and #3. These are the same three cases shown in Figure 2, but here we
plot the stone’s path in space and time. Such a spacetime plot is called a worldline. On each
of three alternative worldlines, flash emissions #1 and #3 are fixed in space and time. Flash
emission #2 is fixed in space (horizontal direction in figure) but its time is varied (up and down in
the figure) to find an extremum of the total wristwatch time τ = τA + τB from #1 to #3. The
result is an expression for a quantity that is a constant of the motion: the energy of the stone.

2 2 1⁄2
τB = [ ( T – t ) – ( S – s ) ] [16]

Again, to prepare for the derivative that leads to extremal aging, differ-
entiate this expression with respect to the intermediate time t:
dτ B – (T – t) T –t
--------- = --------------------------------------------------------- = – ----------- [17]
dt 2 2 1⁄2 τB
[(T – t) – (S – s) ]
3. The total wristwatch time τ from event #1 to event #3 is the sum of the
wristwatch time τA between events #1 and #2 plus the wristwatch time
τB between events #2 and #3:
τ = τA + τB [18]

4. Now ask: When—at what frame time t—will the a stone, following its Use Principle of Extremal
natural path, pass the intermediate point and emit the second flash #2? Aging to find the time for the
middle flash.
Answer with the Principle of Extremal Aging: Time t will be such that
the aging (τ in equation [18]) is an extremum. To find this extremum set
the derivative of τ with respect to t equal to zero. Take the derivative of
both sides of [18] and substitute from equations [15] and [17]:

dτ dτ A dτ B t T –t
----- = --------- + --------- = ------ – ----------- = 0 [19]
dt dt dt τA τB

Section 5 Energy in Special Relativity 1-9

 5. The last equality in equation [19] leads to the equation
t T –t
------ = ----------- [20]
τA τB
Quantity whose value is the 6. In expression [20] the frame time t is the time for the particle to traverse
same for adjacent segments segment A. Call this time tA. The time (T – t) is the frame time for the
particle to traverse segment B. Call this time tB. Then equation [20] can
be rewritten in the simple form
tA tB
------ = ------ [21]
τA τB
. 7. The locations of segments A and B were chosen arbitrarily along the
straight path in space of the particle moving in a region of flat space-
time. Equation [21] holds for all pairs of adjacent segments placed any-
where along the path. We did not specify where segment A was to begin.
Nothing stops us from beginning the analysis with the second segment
B and adding to it a third segment C with which to compare it (which
may have a different length than either of the first two segments). Then
equation [21] applies to the second and third segments. But if the value
of the expression is the same for the first and second segments and also
the same for the second and third segments, then it must be the same
for the first and third segments. Continuing in this way, envision a
whole series of adjacent segments, labeled A, B, C, D, . . . , for each of
which equation [21] applies, leading to the set of equations
tA tB tC tD
------ = ------ = ------ = ------ = . . . [22]
τA τB τC τD

In brief, here is a quantity that is a constant of the motion for the free parti-
cle—a quantity that has the same value along any segment of the natural
path of a free particle moving in flat spacetime. Then equation [22] tells us
that
t
-- = a constant of the motion [23]
τ
E/m = t/τ is a constant of the What is this quantity? It is related to the relativistic expression for the total
motion. energy of the particle. If we have already studied special relativity, we
know that
t t t 1 E
-- = ---------------------------- = ----------------------------------- = --------------------------- = ---- [24]
τ 1 ⁄ 2 2 1 ⁄ 2 1 ⁄ 2 m
t 1 –  -
2 2 s 2
[t – s ] (1 – v )
 t

where m is the mass of the particle. Equation [24] gives the energy per unit
mass of a particle that moves with constant speed.

OBJECTION: Baloney! Everyone knows that a free particle moves with constant speed
along a straight path in space as observed in a free-float frame. So as this motion pro-
ceeds, every possible expression that depends only on v = s/t is also a constant of the

1-10 CHAPTER 1 Speeding

 motion, for example the expression v12, which is certainly not the correct expression
for energy! Your derivation proves nothing!

RESPONSE: You are almost right. Any function of velocity v = s/t is indeed constant for
the special case of a free particle in flat spacetime. And if v is constant, so is
t/τ, as witnessed by Equation [24]. But notice the priorities used in the derivation: The
Principle of Extremal Aging has highest priority; the expression for energy comes out
of this principle. Of all the quantities that remain constant because v is constant, the
Principle of Extremal Aging picks out t/τ = E/m as primary. (The following section
shows that a similar analysis picks out the relativistic expression for momentum as a
constant of the motion.) Chapter 3 contains a new and more general expression for
energy in curved spacetime. In that case the velocity is not constant—yet that more
general expression for energy is correct and a constant of the motion nevertheless.
Our derivation of the expression for E/m in flat spacetime is thus a trial run for the
derivation of the energy of a particle in the curved spacetime around a center of grav-
itational attraction.

If the particle changes speed, then it changes energy. In that case it makes
sense to talk about instantaneous speed and to use calculus notation. Let the
pair of flash emissions in Figure 1 be separated by the incremental frame
coordinates dt, ds, and incremental wristwatch time dτ. The equation for
E/m then becomes
E dt Particle energy in special
---- = ----- [25] relativity
m dτ
Ordinarily we use the ratio E/m in equations, instead of E alone. Why? .
Because it emphasizes two important principles: (1) Only spacetime rela-
tions between events appear on one side of equations such as [24] and
[25], reminding us that it is spacetime geometry that leads to these expres-
sions, not some weird property of matter. (2) The ratio E/m has no units.
Therefore, whoever uses these equations has total freedom in choosing the
unit of E and m, as long as it is the same unit. The same unit in the numera-
tor and denominator of [25] may be kilograms or the mass of the proton or
million electron-volts or the mass of Sun. If you insist on using conven-
tional units, such as joules for energy E and kilograms for mass m, then a
conversion factor c2 intrudes into our simple equation:
E joules dt
--------------- = ----- [26]
2 dτ
m kg c

Now view the particle from a reference frame in which the particle is at Rest energy: famous formula
rest. In this rest frame there is zero distance s between sequential flash
emissions. Equation [1] says that for s = 0 the frame time t and wristwatch
time τ have exactly the same value. For a particle at rest, then, equation
[26] reduces to the most famous equation in all of physics:
2
E joules rest = m kg c [27]

Note that equation [27] describes the rest energy of a particle. For a particle
in motion, the energy is given by equation [26].

Section 5 Energy in Special Relativity 1-11

 In equation [27], c has the defined value 2.99792458 × 108 meters/second.
An equation of the same form is correct if E is measured in ergs, m in
grams, and c in centimeters/second.

Fuller Explanations: Energy in flat spacetime: Spacetime Physics,

Chapter 7, Momenergy.

6 Momentum in Special Relativity

The metric plus the Principle of Extremal Aging give us an expression for momentum.
The relativistic expression for momentum is derived by a procedure anal-
ogous to the one used to derive the relativistic expression for energy. The
figures look similar to Figures 2 and 3, but in this case the time t for the
intermediate flash emission is fixed, while the position s for this event is var-
ied right and left to yield an extremum for the total wristwatch time from
the first flash to the third flash. (You carry out the derivation of momen-
tum in the exercises at the end of this chapter.) The result is a second
constant of the motion for a free particle:
s s s⁄t v p
-- = ---------------------------- = ------------------------------------- = --------------------------- = ---- [28]
τ 2 2 1 ⁄ 2 2 1 ⁄ 2 2 1 ⁄ 2 m
[t – s ] [1 – (s ⁄ t) ] (1 – v )

Equation [28] gives the momentum per unit mass for a particle moving
with constant speed. If the particle changes speed, then once again we use
calculus notation:
Particle momentum in special p ds
relativity ---- = ----- [29]
m dτ

Equation [29] has the same form as in Newton’s nonrelativistic mechanics,

except here the incremental wristwatch time dτ replaces the Newtonian
lapse dt of “universal time.”

Fuller Explanations: Momentum in flat spacetime: Spacetime Physics,

Chapter 7, Momenergy.

7 Mass in Relativity
Everyone agrees on the value of the mass m of the stone.
Find mass from energy and An important relation among mass, energy, and momentum follows from
momentum. the metric and our new expressions for energy and momentum. Suppose a
moving stone emits two flashes very close together in space ds and in time
dt. Then equation [1] gives the increase of wristwatch time dτ:
2 2 2
( dτ ) = ( dt ) – ( ds ) [30]

Divide through by dτ2 and multiply through by m2 to obtain

2 dt 2 2 ds 2 dt 2 ds 2
m = m  ----- – m  ----- =  m ----- –  m -----
2
[31]
 dτ  dτ  dτ  dτ

1-12 CHAPTER 1 Speeding

or, substituting expressions [25] and [29] for energy and momentum,
2 2 2 Energy (also momentum)
m = E –p [32] may be different for
different observers . . .
In equation [32], mass, energy, and momentum are all expressed in the
same units, such as kilograms or electron-volts. In conventional units, the
equation has a more complicated form:
2 2 2 2 2
( mc ) = E conv – p conv c [33]

where the subscript “conv” means “conventional units.”

Equations [32] and [33] are central expressions in special relativity. The . . . but mass is an invariant,
particle energy E will typically have a different value when measured in the same for every observer.

different frames that are in uniform relative motion. Also the particle
momentum p will typically have a different value when measured in dif-
ferent frames that are in uniform relative motion. However, the values of
these two quantities in any given free-float frame can be used to determine
the value of the particle mass m, which is independent of the reference
frame. Particle mass m is an invariant, independent of reference frame, just
as the time dτ recorded on the wristwatch between ticks in equation [1] is
an invariant, independent of the reference frame.

The mass m of key, car, or coffee cup defined in equation [32] is the one we
use throughout our study of both special and general relativity. Such a test
particle responds to the structure of spacetime in its vicinity but has small
enough mass not to affect this spacetime structure. (In contrast, the large
mass M of a planet, star, or black hole does affect spacetime in its vicinity.)
Wherever we are, we can always climb onto a local free-float frame (Sec-
tion 8) and apply special-relativity expression [32] or some other standard
method to measure the mass m of our test particle.

Fuller Explanations: Mass and momentum-energy in flat spacetime:

Spacetime Physics, Chapter 7, Momenergy.

No Mass Change with Velocity!

The fact that no object moves faster than the speed of light is atoms. Our viewpoint in this book is that mass is an invariant,
sometimes “explained” by saying that “the mass of a particle the same for all free-float observers when they use equations
increases with speed.” This interpretation can be applied [32] or [33] to reckon the mass. In relativity, invariants are
consistently, but what could it mean in practice? Someone diamonds. Do not throw away diamonds! For more on this
riding along with a faster-moving stone detects no change in subject, see Spacetime Physics, Dialog: Use and Abuse of
the number of atoms in the stone, nor any change whatever the Concept of Mass, pages 246–251.
in the individual atoms, nor in the binding energy between

Section 7 Mass in Relativity 1-13

 8 The Free-Float Frame Is Local
In practice there are limits on the space and time extent of the free-float (inertial)
frame.
The free-float (inertial) frame is the arena in which special relativity
describes Nature. The power of special relativity applies strictly only in a
frame—or in each one of a collection of overlapping frames in uniform rel-
ative motion—in which a free particle released from rest stays at rest and a
particle launched with a given velocity maintains the magnitude and
direction of that velocity.
Limits of local free-float If it were possible to embrace the Universe with a single free-float (inertial)
frames imply the need for frame, then special relativity would describe that universe and general rel-
general relativity.
ativity would not be needed. But general relativity is needed precisely
because typically inertial frames are inertial in only a limited region of
space and time. Inertial frames are local. The free-float frame can be real-
ized, for example, inside various “containers,” such as (1) an unpowered
spaceship in orbit around Earth or Sun or (2) an elevator whose cables
have been cut or (3) an unpowered spaceship in interstellar space. Riding
in these free-float frames for a short time, we find no evidence of gravity.

Free-float frame cannot be Well, almost no evidence. The enclosure in which we ride cannot be too
too large. large or fall for too long a time without some unavoidable changes in rela-
tive motion being detected between particles in the enclosure. Why?
Because widely separated test particles within a large enclosed space are
differently affected by the nonuniform gravitational field of Earth—to use
the Newtonian way of speaking. For example, two particles released side
by side are both attracted toward the center of Earth, so they move closer
together as measured inside a falling long narrow horizontal railway
coach (Figure 4, left). Moving toward one another has nothing to do with
gravitational attraction between these test particles, which is entirely neg-
ligible.

As another example, think of two test particles released far apart vertically
but one directly above the another in a long narrow vertical falling railway
coach (Figure 4, right). For vertical separation, their gravitational accelera-
tions toward Earth are in the same direction, according to the Newtonian
analysis. However, the particle nearer Earth is more strongly attracted to
Earth and gradually leaves the other behind: the two particles move far-

Elevator Safety
Could the cables snap and send an elevator plummeting down the shaft?

This is every rider's worst fear, but experts say there's no need to worry. You're being sup-
ported by four to eight cables, each of which could support the weight of the car by itself. In
fact, the only time an elevator has been known to go into freefall—with all of its cables
cut—was during World War II, when an American bomber accidentally hit the Empire State
Building [in New York City]. The plane's crew died, but the lone elevator passenger survived.

—Good Housekeeping Magazine, February 1998, page 142.

1-14 CHAPTER 1 Speeding

 Figure 4 Einstein’s old-fashioned railway coach in free fall. Left: horizontal orientation. Right: vertical
orientation.

ther apart as observed inside the falling coach. Conclusion: The large
enclosure is not a free-float frame.

A rider in either railway car shown in Figure 4 sees the pair of test parti-
cles accelerate toward one another or away from one another. These rela-
tive motions earn the name tidal accelerations, because they arise from
the same kind of nonuniform gravitational field—this time the field of
Moon—that account for ocean tides on Earth.

Now, we want the laws of motion to look simple in our free-float frame.
Therefore we want to eliminate all relative accelerations produced by
external causes. “Eliminate” means to reduce them below the limit of
detection so that they do not affect measurements of, say, the velocity of a
particle in an experiment. We eliminate the problem by choosing a room
that is sufficiently small. Smaller room? Smaller relative motions of objects
at different points in the room!
Let someone have instruments for detection of relative motion with any Reduce space or time
given degree of sensitivity. No matter how fine that sensitivity, the room extension to preserve free-
float frame.
can always be made so small that these perturbing relative motions are too
small to be detectable in the time required for the experiment. Or, instead
of making the room smaller, shorten the time duration of the experiment
to make the perturbing motions undetectable. For example, very fast par-
ticles emitted by a high-energy accelerator on Earth traverse the few-meter
span of a typical experiment in so short a time that their deflection in

Section 8 The Free-Float Frame Is Local 1-15

 Earth's gravitational field is negligible. The result: The frame of the labora-
tory at rest on Earth's surface is effectively free-float for purposes of ana-
lyzing these experiments.
Both space and time enter into the specification of the limiting dimensions
of a free-float frame. Therefore—for a given sensitivity of the measuring
devices—a reference frame is free-float only within a limited region of
spacetime.
Test for free-float property An observer tests for a free-float frame by releasing particles from rest
within the frame itself. throughout the space and noting whether they remain effectively at rest
during the time set aside for our particular experiment. Wonder of won-
ders! Testing for free float can be carried out entirely within the frame
itself. The observer need not look out of the room or refer to any measure-
ments made external to the room. A free-float frame is “local” in the sense
that it is limited in space and time—and also “local” in the sense that its
free-float character can be determined from within, locally.
. One way to get rid of “gravitational force” is to jump from a high place
toward a trampoline below. That is to say, a locally free-float frame is
always available to us. But no contortion or gyration whatsoever will
eliminate the relative accelerations of test particles that indicate the limits
of the free-float frame. These relative accelerations are the central indica-
tors of the curvature of spacetime. They stand as warning signs that we are
reaching the limits of special relativity.
General relativity requires How can we analyze a pair of events widely separated near Earth, near
more than one free-float Sun, or near a neutron star, events too far apart to be enclosed in a single
frame.
free-float frame? For example, how do we describe the motion of an aster-
oid whose orbit completely encircles Sun, with an orbital period of many
years? The asteroid passes through many free-float frames but cannot be
tracked using a single free-float frame. Special relativity has reached its
limit! To describe accurately motion that oversteps a single free-float
frame, we must turn to general relativity—the Theory of Gravitation—as
we do in Chapter 2.

Fuller explanations: Spacetime Physics, Chapter 2, Floating Free, and

Chapter 9, Gravity: Curved Spacetime in Action.

9 The Observer
Ten thousand local witnesses
Detect each event locally, How, in principle, do we record events in space and time? Nature puts an
using a latticework of clocks. unbreakable speed limit on signals—the speed of light. This speed limit
causes problems with the recording of widely separated events, because
we do not see a remote event until long after it has occurred. To avoid the
light-velocity delay, adopt the strategy of detecting each event using
equipment located right next to that event. Spread event-detecting equip-
ment over space as follows. Think of assembling metersticks and clocks
into a cubical latticework similar to a playground jungle gym (Figure 5).
At every intersection of the latticework fix a clock. These clocks are identi-
cal and measure time in meters of light-travel time.

1-16 CHAPTER 1 Speeding

 Figure 5 Latticework of metersticks and clocks

These clocks should read the same time. That is, the clocks need to be syn- Synchronize clocks in the
chronized in this frame. There are many valid ways to synchronize clocks. lattice.

Here is one: Pick one clock as the standard, the reference clock. At mid-
night the reference clock sends out a synchronizing flash of light in all
directions. Prior to emission of the synchronizing flash, every other clock
in the lattice has been stopped and set to a time (in meters) later than mid-
night equal to the straight-line distance (in meters) of that clock from the
reference clock. Each clock is then started when it receives the reference
flash. The clocks in the latticework are then said to be synchronized.
Use the latticework of synchronized clocks to determine the location and Measuring the space and
time at which any given event occurs. The spatial position of the event is time location of an event

taken to be the location of the clock nearest the event and the time of the
event is the time recorded on that clock. The location of this nearest clock
is measured along three directions, northward, eastward, and upward
from the reference clock. The spacetime location of an event then consists
of four numbers, three numbers that specify the space position of the clock
nearest the event and one number that specifies the time the event occurs
as recorded by that clock.

Section 9 The Observer 1-17

 Specify the location of an event as the location of the clock nearest to it.
With a latticework made of metersticks, the location of the event will be
uncertain to some substantial fraction of a meter. For events that must be
located with greater accuracy, a lattice spacing of 1 centimeter or 1 milli-
meter would be more appropriate. To track an Earth satellite, lattice spac-
ing of 100 meters might be adequate.
The lattice clocks, when installed by a foresighted experimenter, will be
recording clocks. Each clock is able to detect the occurrence of an event
(collision, passage of light flash or particle). Each reads into its memory
the nature of the event, the time of the event, and the location of the clock.
The memory of all clocks can then be read out and analyzed later at some
command center.
The “observer” is all the In relativity we often speak about the observer. Where is this observer? At
recording clocks in one one place or all over the place? Answer: The word observer is a shorthand
frame.
way of speaking about the whole collection of recording clocks associated
with one free-float frame. This is the sophisticated sense in which we here-
after use the phrase “the observer measures such-and such.”
What happens to our latticework of clocks in the vicinity of Earth or Sun
or neutron star or black hole? Suppose one of these centers of attraction is
isolated in space and we stay far away from it. Then there is no problem in
setting up an extensive latticework that starts far from the center and
stretches even farther away in all directions. Such an extensive far-away
lattice can represent a single valid free-float frame. And in studying gen-
eral relativity we often speak of a far-away observer.
The far-away lattice is not But there are problems in extending the far-away latticework of clocks
free float when extended to down toward the surface of any of these structures. A free particle
near Earth or black hole.
released from rest near that center does not remain at rest with respect to
the far-away lattice. A single free-float frame no longer provides a simple
description of motion.
Many local frames are To describe motion near a center of gravitational attraction we must give
required near Earth or black up the idea of a single global free-float frame, one that covers all space and
hole.
time around Earth or black hole. Replace it with many local frames, each
of which provides only a small part of the global description. A world
atlas binds together many overlapping maps of Earth. Individual maps in
the atlas can depict portions of Earth’s surface small enough to be essen-
tially flat. Taken together, the collection of maps bound together in the
world atlas correctly describes the entire spherical surface of Earth, a task
impossible using a single large flat map for the entire Earth. For spacetime
near nonrotating Earth or black hole, the task of binding together individ-
ual localized free-float frames is carried out by the Schwarzschild metric,
introduced in Chapter 2. The Schwarzschild metric frees us from limita-
tion to a single free-float frame and introduces us to curved spacetime.

Fuller Explanations: Spacetime Physics, Chapter 2, Section 2.7, Observer.

1-18 CHAPTER 1 Speeding

10 Summary

The wristwatch time τ between two events, the time recorded on a watch
that moves uniformly from one event to the other, is related to the separa-
tion s between the events and the time difference t between them as mea-
sured in a given frame. For space and time measured in the same units,
this relation is given by the equation
2 2 2
τ = t –s [1]

The wristwatch time τ is an invariant, the same calculated by all observers,

even though t and s may have different values, respectively, as measured
in different reference frames. Equation [1] is an example of the metric.

Of all possible paths between an initial event and a final event, a free parti-
cle takes the path that makes the wristwatch time along the path an extre-
mum. This is called the Principle of Extremal Aging.

From the metric and the Principle of Extremal Aging one can derive two
quantities that are constants of the motion for a free particle. One constant
of the motion is the energy per unit mass E/m:
E dt
---- = ----- [25]
m dτ

The second constant of the motion is the momentum per unit mass p/m:
p ds
---- = ----- [29]
m dτ

The spacetime arena for special relativity is the free-float (inertial) frame, one
in which a free test particle at rest remains at rest and a free test particle in
motion continues that motion unchanged. We call a region of spacetime
flat if a free-float frame can be set up in it.

In principle one can set up a latticework of synchronized clocks in a free-

float frame. The position and time of any event is then taken to be the loca-
tion of the nearest lattice clock and the time of the event recorded on that
clock. The observer is the collection of all such recording clocks in a given
reference frame.

Most regions of spacetime are flat over only a limited range of space and
time. Evidence that a frame is not inertial (so that its region of spacetime is
not flat) is the relative acceleration (“tidal acceleration”) of a pair of free
test particles with respect to one another. If tidal accelerations affect an
experiment in a region of space and time, then we say that spacetime
region is curved, and special relativity cannot validly be used to describe
this experiment. In that case we must use general relativity, the theory of
gravitation, which correctly describes the relations among events spread
over regions of space and time too large for special relativity.

Section 10 Summary 1-19

 Note on terminology: In this book we use the convention recommended by
the International Astrophysical Union that names for objects in the solar
system be capitalized and used without the article. For example, we say
“orbits around Sun” or “the mass of Moon.” This provides a consistent
convention; one would not say “orbits around the Mars.” We also capital-
ize the words Nature and Universe out of respect for our cosmic home.

11 Readings in Special Relativity

Spacetime Physics, Introduction to Special Relativity, Second Edition, Edwin F.

Taylor and John Archibald Wheeler, W. H. Freeman and Co., New
York, 1992, ISBN 0-7167-2327-1. Our own book, to which reference is
made at the end of several sections in Chapter 1 and elsewhere in the
present book.

Special Relativity, A. P. French, W. W. Norton & Co., New York, 1968,

Library of Congress 68-12180. An introduction carefully based on
experiment and observation.

A Traveler’s Guide to Spacetime, An Introduction to the Special Theory of Rela-

tivity, Thomas A. Moore, McGraw-Hill, Inc., News York, 1995, ISBN 0-
07-043027-6. A concise treatment by a master teacher.

Flat and Curved Space-Times by George F. R. Ellis and Ruth M. Williams,

Clarendon Press, Oxford, 1988, ISBN 0-19-851169-8. A leisurely, infor-
mative, and highly visual trip through special relativity is followed by
treatment of curved spacetime. See more on this book in the section
Readings in General Relativity at the end of the present book.

Space and Time in Special Relativity, N. David Mermin, Waveland Press, Inc.,
Prospect Heights, IL, 1989, ISBN 0-8813-420-0. Rigorous and mildly
eccentric.

Understanding Relativity: A Simplified Approach to Einstein’s Theories, Leo

Sartori, University of California Press, Berkeley, 1996, ISBN 0-520-
20029-2. Thoughtful and complete.

Relativity, The Special and General Theory, Albert Einstein, Crown Publish-
ers, New York, 1961, ISBN 0-517-025302. A popular treatment by the
Old Master himself. Published originally in 1916. Enjoyable for the
depth of physics, the humane viewpoint, and the charm of old-fash-
ioned trains racing past embankments.

Relativity Visualized, Lewis Carroll Epstein, Insight Press, 1985, ISBN 0-

935218-05-X. Eccentric visual treatment of special and general
relativity.

1-20 CHAPTER 1 Speeding

Of historical interest

Relativity and Its Roots, Banesh Hoffmann, Scientific American Books, New
York, 1983, ISBN 0-7167-1510-4. History of the subject by one of Ein-
stein’s collaborators.

The Principle of Relativity, A. Einstein, H. A. Lorentz, H. Weyl, H.

Minkowski, Dover Publications, Inc., New York, 1952, Standard Book
Number 486-60081-5. Translations of many of the original papers. See
the following reference for a more recent translation of Einstein’s spe-
cial relativity paper.

Albert Einstein’s Special Theory of Relativity: Emergence (1905) and Early Inter-
pretation (1905–1911), Arthur I. Miller, Addison-Wesley Publishing Co.,
Inc., 1981, ISBN 0-201-04680-6. Careful historical analysis of Einstein’s
original special relativity paper “On the Electrodynamics of Moving
Bodies,” the setting in which it was produced, and early consequences
for the scientific community. Includes a modern, corrected translation
of the paper itself.

12 Reference

Initial quote: Personal memoir of William Miller, an editor of Life maga-

zine, quoted in the issue of May 2, 1955. See The Quotable Einstein, edited
by Alice Calaprice, Princeton University Press, 1996, page 199.

Section 12 Reference 1-21