Vdoc - Pub Simply Einstein Relativity Demystified
Vdoc - Pub Simply Einstein Relativity Demystified
Vdoc - Pub Simply Einstein Relativity Demystified
Relativity Demystified
• • •
RICHARD WOLFSON
Cover
Title Page
Dedication
Preface
5 Ether Dreams
6 Crisis in Physics
8 Stretching Time
13 Is Everything Relative?
14 A Problem of Gravity
16 Einstein’s Universe
Glossary
Further Readings
Index
Copyright
PREFACE
• • •
Have you ever heard it said of a difficult idea that “it would take an Einstein
to understand this”? What could be more incomprehensible to us non-
Einsteins than Albert Einstein’s own work, the theory of relativity?
But relativity is comprehensible, and not just to scientists. At the heart
of relativity is an extraordinarily simple idea—so simple that a single
English sentence suffices to state it all. Some consequences of that
statement are disturbing because they violate our deeply held,
commonsense notions about the world. Yet those consequences flow
inexorably from a single principle so simple and obvious that it will take me
just a few pages to convince you of its truth.
This book’s title, Simply Einstein, reflects the fact that the basic ideas of
Einstein’s relativity are accessible to nonscientists and make eminent sense.
Even relativity’s startling implications about the nature of space, time, and
matter follow so directly from those basic ideas that they, too, become not
only comprehensible but also logically inevitable.
Relativity is behind many of the hot topics at the frontiers of modern
physics, astrophysics, and cosmology—topics ranging from black holes to
the ultimate fate of the Universe to the prospects for time travel. I’ll touch
on these topics here, and you’ll see how they flow from the essential ideas
of relativity. But my main purpose is not to explore the latest frontiers of
physics. There are plenty of good books on those topics, and I’ve included
some in the Further Readings. Rather, this is a book that aims to give you,
its reader, a clear understanding of just what it was that Einstein said about
the ultimate nature of physical reality. To help you get there, we’ll be
exploring together the history of ideas that culminated in Einstein’s simple
but remarkable vision. Then you’ll see how that vision alters your
commonsense notions of space and time in ways that would let you travel a
thousand years into the future in just a few short hours. You’ll come to a
new understanding of “past” and “future” that might surprise historians, and
you’ll begin to feel at home in the four-dimensional universe of relativistic
spacetime. Along the way I’ll anticipate your frequent questions: Why can’t
anything go faster than light? Will I really age more slowly, or is this just
something that happens to physicists’ clocks? Can I go backward in time?
What does E = mc2 really mean? Finally, in the end, we’ll return to some of
those contemporary hot topics that show just how prescient was Einstein’s
visionary insight.
You don’t need to do math to grasp the essence of Einstein’s relativity,
and you don’t need math to understand this book. Occasional numbers can
help make some points more concrete, and I’ll use them sparingly. What’s
important here are the big ideas—and they’re all expressed in words. Grasp
those ideas, and you know what Einstein’s relativity is all about. Enjoy!
CHAPTER 1
Could the Universe have created itself? What an absurd idea! Did the
Universe even have a beginning? That question, too, has an absurd ring. If
there was a beginning, what came before? Wasn’t that part of the Universe
too? Or has the Universe always existed, begging the question of its own
origin?
Whatever the answers to these questions, modern astrophysics makes
one thing clear: our Universe hasn’t existed forever unchanged. Rather, it’s
evolved from an earlier state of extreme temperature and density. Some 14
billion years ago, all the “stuff” that makes up ourselves, our planet Earth,
and all the stars and galaxies was crammed into a volume far smaller than a
single hydrogen atom or even the tiny proton at its core. The expansion of
that extreme state is the Big Bang that describes the Universe’s subsequent
evolution and ultimately accounts for the origin of stars, galaxies, planets,
and intelligent life.
What came before the Big Bang? What created that early, extreme state?
We’re back to the primordial question: Did the Universe have a beginning,
or has it always existed—albeit an existence marked by evolutionary
change?
To some cosmologists—scientists who concern themselves with the
origin and evolution of the Universe—the start of the Big Bang marks the
start of time itself. For them, it makes no sense to ask what came before
because the concept of “before” is meaningless if there’s no such thing as
time. Others have envisioned an ever existing Universe that undergoes a
series of oscillations. Each begins with a Big Bang and subsequent
expansion—the phase we’re now in—then eventually contracts toward a
Big Crunch of extreme density and temperature that starts another cycle.
In 1998 Princeton physicist J. Richard Gott and his student Li-Xin Li
published a novel answer to the ultimate question of the Universe’s origin.
Their paper, “Can the Universe Create Itself?,” shows how the laws of
physics may allow a time loop, in which time goes round and round in a
circlelike structure rather than advancing inexorably into a never-before-
experienced future. Like Bill Pullman’s character in the film Groundhog
Day, an occupant of the time loop might go to bed at night and wake up on
the morning of the day before! The “new” day would unfold, night would
come, and again the morning would bring the already familiar day. In this
loop, time advances circularly into a future that is a recycling through past
events. There’s no earliest event, any more than any point on a circle can be
called the beginning of the circle or any point on Earth’s surface is the place
where our planet “starts.”
Gott’s time loop doesn’t sound like a description of our Universe, but
hold on—there’s more. Time, in Gott’s theory, can branch, providing
different paths to different futures. Gott envisions a universe whose earliest
epoch includes a time loop. Every point on the loop both precedes and
follows every other point. There is no beginning instant, because one can
always trace time further back round the loop. But there’s a branch out of
the time loop, a branch into a more normal realm of time that advances,
without repetition, to a future of never-before events. That’s the kind of
time we know, with an as-yet-unknown, yet-to-occur future. Figure 1.1
depicts Gott’s time-loop, multibranched universe.
It’s the branching that reconciles Gott’s time loop with the more
ordinary time we experience today. The time loop unambiguously precedes
our present, and in that sense it’s closer to the beginning. But trace time
backward, through the branch and onto the loop. You can keep tracing back
but you’ll never find a beginning. Instead, events repeat as your historical
exploration circles backward around the loop. There’s no one event that
marks the creation of this universe. Every event on the time loop precedes
every other event, and in that sense Gott’s universe creates itself. Absurd!
Fig. 1.1 Gott’s time-loop universe. Arrows represent multiple directions of time, including the
circular time loop at the beginning, in which time goes round and round in an ever repeating
sequence. Circular cross sections of the “trumpets” represent position in a single spatial dimension,
and each trumpet is its own universe. Is this absurd? Maybe not, says the theory of relativity.
(Adapted with permission from J. Richard Gott and Li-Xin Li, “Can the Universe Create Itself?”
Physical Review D 58 (1998), p. 3501.)
Contact!
In the film Contact, based on Carl Sagan’s novel with the same title, actress
Jodi Foster plays the first astrophysicist to detect interstellar signals from an
advanced civilization. The signals convey a message—instructions for
building some sort of machine. Machines are built, against a backdrop of
political and religious intrigue, and eventually Foster’s character boards one
for a ride into the unknown. The machine takes her through a wormhole in
spacetime and deposits her on a distant world where she learns that entire
galaxies are actually cosmic engineering projects. She rides the wormhole
machine back to Earth only to find that no one believes her story because,
for folks on Earth, no time has elapsed while she was ostensibly touring the
cosmos.
Contact looks like science fiction. But many of its key ideas, including
wormholes through space and time, are based in sound physics. In fact,
author Sagan—a scientist himself—consulted colleagues about the validity
of the sci-fi ideas at the basis of Contact. As a result, Sagan’s novel spurred
a flurry of interest in wormholes and in the possibility of time travel. By the
turn of the century, leading researchers had published scores of papers on
these subjects in the most respected physics journals. Some show how
wormholes might connect seemingly distant parts of the Universe that are
actually less than an inch apart in a hidden dimension. Others debate the
mathematical possibility and philosophical implications of time machines
that might let us travel into the past. That’s getting pretty speculative, but
another form of time travel is solidly established. Read on.
A Common Absurdity
Time loops? A time before time? A many-branched Multiverse?
Wormholes? Leapfrogging into the future? Galactic telescopes peering to
the edge of space and time? What do all these seeming absurdities have in
common? All require conceptions of space and time that boggle our
common sense. All require that space and time bend, warp, and distort in
ways that have no counterparts in our everyday experience. All require that
we give up our tenth-grade notions of geometry, with straight lines and
perfect triangles. Straight lines through space, from here to there? Not in a
Universe where cosmic lenses focus multiple images of the same object!
Straight lines through time, from now to then? Not in a Universe with
wormholes, time loops, and high-speed star trips!
The geometry of space and time is not, in fact, the geometry you learned
in tenth grade. It’s a much richer geometry that allows for the curving,
bending, or warping of space and time. That richness, in turn, enables a host
of new phenomena: time loops that bend time back on itself; black holes,
where spacetime curvature becomes infinite; gravitational lenses that create
multiple paths for light from distant sources to reach Earth; high-speed
travel that’s a shortcut to the future. The strange geometry of spacetime is
not some unfathomable mystery, though. It’s described precisely by
Einstein’s theory of relativity—a theory that is as much about the geometry
of space and time as it is about phenomena of physics. In fact, relativity
suggests that we think of geometry as a branch of physics rather than
mathematics, setting the spacetime “stage” on which we and the rest of
physical reality strut and fret our roles in the larger Universe.
So what is this relativity theory that mixes geometry into physical
reality and leads to mind-boggling possibilities like time loops, wormholes,
and cosmic telescopes? Ultimately, the idea behind relativity is a very
simple one—so simple that I can state it in a single sentence. But it’s a very
big idea, too—an idea whose consequences and philosophical implications
go far beyond the confines of physics. In the next chapter I will introduce
you to the simple Principle of Relativity, and I’ll convince you that you
already grasp and even embrace this idea at the heart of relativity. From
there we’ll explore how relativity came to be, what it really means, and why
its consequences fly in the face of common sense. At the end, I’ll return to
the cosmic implications of relativity to show how exotic phenomena like
wormholes, time loops, and gravitational lenses follow from your newfound
understanding of Einstein’s remarkable vision.
CHAPTER 2
Picture this: You’re on a cruise ship, sailing straight and steady through
calm waters. You’re playing tennis on the ship’s indoor court. How does the
moving ship affect your game? As the ball approaches your racket, do you
sense the ship’s motion and adjust your swing accordingly? And what about
your opponent, who has her back to the ship’s motion? Does she have to
adjust her play in a way different from you?
The answer to all these questions is an obvious no. Your game on the
cruise ship proceeds exactly as it would on land. Neither you nor your
opponent needs to consider the ship’s motion whatsoever. Neither of you
needs to adjust your playing style to compensate for that motion.
Here’s something simpler than the complex volleying of a tennis match.
Suppose you’re standing on land, with no wind blowing, and you throw a
tennis ball straight up. Up the ball goes; eventually it slows to a momentary
stop, then drops straight back down into your waiting hand. Now try the
same thing in that indoor court on the cruise ship. (I’ve put the court
indoors to eliminate the wind you would experience on an outdoor court
because of the ship’s motion through the air.) Again the ball goes straight
up and drops straight back down. That the ship is moving doesn’t matter in
the least.
Tired of tennis, you stroll over to a microwave oven to heat a cup of tea.
Now how do you compensate for the ship’s motion? Do you put the cup at
the sternward end of the oven because the microwaves get “left behind” by
the moving ship? Do you adjust the power level because the microwaves
behave differently from how they would back on shore? Of course not! The
oven, like the tennis ball, behaves in just the same way on the moving ship
as it would on solid ground.
Of course, a cruise ship doesn’t go very fast—maybe 30 miles per hour
or so. So imagine now that you’re in a space colony on Venus, which at that
point in its orbit happens to be moving about 30 miles per second with
respect to Earth. The atmosphere inside the colony is the same as on Earth,
and Venus’s gravity has essentially the same strength as Earth’s. You step
into the colony’s recreation center for a game of tennis. As you start your
serve, do you need to account for the fact that the whole court is moving at
30 miles per second? How could you possibly compensate for that speed?
The answer, of course, is that you don’t have to. The tennis game on Venus
proceeds just as it would on Earth. Venus’s motion is irrelevant.
Now try the simpler act of throwing the ball straight up. One second
later, is it 30 miles away from you? Of course not! It goes straight up and
comes straight down, just as it would on Earth. And why not? Why should
the existence of Earth, millions of miles away and moving at 30 miles per
second with respect to Venus, have any significant effect on what happens
on Venus? Why should the rules governing the tennis ball be any different
on Venus than they are on Earth?
Tired of Venusian tennis, you stroll over to a microwave oven to heat a
cup of tea. This oven, of course, is moving at 30 miles per second. Do you
worry about that? Of course not! Again, Venus’s motion is irrelevant.
Maybe even 30 miles per second isn’t much. So now imagine you’re a
humanoid being on an Earthlike planet in a distant galaxy moving away
from Earth at 80 percent of the speed of light—about 150,000 miles per
second. (The Hubble Space Telescope routinely observes very distant
galaxies with speeds this high; their motion is part of the overall expansion
of the Universe.) Again, it’s tennis time. As you start your game, do you
think about the fact that you’re moving at 150,000 miles per second away
from Earth, and do you make a hopeless attempt to compensate for that
colossal motion? Of course not! You don’t even know about Earth; the
Milky Way galaxy, in which Sun and Earth are minor players, is but a faint
smudge in your civilization’s most advanced telescopes. Why should Earth
have anything to do with you and how you play tennis?
Again, the game’s over and it’s tea time. The microwaves that heat your
tea move through the oven at the speed of light—and this time the oven
itself is moving at 80 percent of that speed. Do the microwaves get left
behind? Does the microwave oven need a different set of instructions—or a
whole different design—to work on this distant planet because it’s moving
at such high speed? The answer, again, is a firm no. The microwave oven
works just fine, thank you, on this distant planet as it does on Earth.
These three examples of the cruise ship, Venus, and the distant galaxy
all illustrate a point that seems to make perfect sense: The physical “stuff”
of the Universe (tennis balls, atoms, microwaves—indeed, all matter and
energy) works the same way everywhere. Put another way, the laws of
physics—the rules that are somehow written into the structure of the
Universe and that tell things how to behave—are the same everywhere.
There’s no special, preferred place where the laws of physics are correct
while they need modifying everywhere else.
This idea that no place is special is hardly new. In 1543, Copernicus’s
On the Revolutions of the Heavens first challenged the ancient view of
Earth as the center of the Universe. Copernicus’s Sun-centered theory was a
major shift not only for science but also for philosophy and religion (which
is why the established church fought Copernicus’s ideas). Since Copernicus,
discoveries in science have only reinforced the view that there’s nothing
special about Earth in the cosmic scheme of things. We, its inhabitants, may
love our planet and consider it a special place—but that doesn’t mean Earth
gets special privileges where the laws of physics are concerned.
Actually, “place” isn’t quite the right word here. While it’s true that the
laws of physics apply everywhere, more important in my tennis examples is
that they apply in different states of motion. A tennis match and a
microwave oven—both governed by the laws of physics—work the same
way on Earth and on a hypothetical planet moving away from Earth at 80
percent of the speed of light. The fact of the planet’s motion is simply
irrelevant. If it weren’t irrelevant, there would be something mighty special
about Earth, in that everyone else in the Universe would have to reference
Earth in describing their own physical situations.
This idea that motion doesn’t matter in some absolute sense goes back
to Galileo and Isaac Newton, who were first to formulate quantitative laws
that describe the way objects move. But the idea found its complete fruition
with Albert Einstein, whose theory of relativity is based on the simple
statement that motion doesn’t matter, that is, that the laws governing
physical reality are just the same on that distant planet hurtling away from
Earth as they are right here on Earth itself. I’ll rephrase that absurdly simple
fact: The theory of relativity is, in its barest essence, just the simple
statement that regardless of one’s state of motion the laws of physics are the
same.
If you followed the simple example of the tennis match and the
microwave oven, then you already know and believe the basis of Einstein’s
theory of relativity. At its heart, this really is the essence of relativity: that
you can play tennis and use a microwave oven—or undertake any other
activity involving physical reality—in any state of motion, and you will
always get exactly the same results. Put in more scientific language, the
results of any scientific experiment will be the same for anyone who cares
to undertake that experiment—again, regardless of state of motion. By
“experiment,” I simply mean some activity that probes the behavior of
physical reality. In that sense, playing tennis is an experiment. So is heating
water in a microwave oven. Watching the tennis ball bounce or the water
boil tells you about how the physical world works. A simpler and more
controlled experiment is the third one I introduced: throwing a ball straight
up and observing its subsequent behavior. What relativity says about these
and any other scientific experiments is that otherwise identical experiments
performed by experimenters in different states of motion will give the same
results. By “identical” I mean all relevant physical circumstances are the
same. That’s why I put the tennis court on Venus and on an Earthlike planet
in a distant galaxy; that way, gravity was the same and so, therefore, was
the force with which you had to swing the racket to put the ball in your
opponent’s court. On Jupiter or the Moon your tennis game would go a bit
differently because of the stronger or weaker gravity, but ultimately you
could infer the same underlying physical laws. Stated more directly, anyone
who cares to experiment with physical reality will discover the same
fundamental physical laws, regardless of the experimenter’s motion.
So you already know and believe the essence of relativity. In that sense,
you can close this book and be done with it. Stripped to its essentials, the
theory of relativity is summarized in a single, simple English sentence: The
laws of physics are the same for all. This statement—called the Principle of
Relativity—is so obvious that you need read no further, so obvious that I’ve
probably belabored it too much already.
But wait! Although the essential idea of relativity is simple,
straightforward, and so eminently believable as to be obvious, the
consequences of that idea are anything but transparent. Nor, at first glance,
are those consequences particularly believable. I gave you some examples
in the preceding chapter; they included such seeming absurdities as time
loops, wormholes, and shortcuts to the future.
Consider the last of these, in which you and I, originally the same age,
find ourselves 20 years apart when you return from a trip to a nearby star. Is
this really possible? On what absurd new principles of physics does it rest?
It is possible and was verified, albeit modestly, by the round-the-world
atomic-clock experiment I described. For high-speed subatomic particles,
the observed effect is even more dramatic than for your hypothetical star
trip. But there’s no bizarre new principle involved—only the principle that
the laws of physics should be the same regardless of one’s state of motion.
I’ve just convinced you that the laws governing physical reality should be
the same on Earth, on Venus, and on a planet in a distant galaxy moving
away from Earth at nearly the speed of light. My task was easy because you
already knew that Earth is not the center of the Universe and you accept
that astrophysicists routinely observe distant galaxies moving rapidly away
from us. If the laws of physics apply throughout the Universe, surely they
should work as well in those distant galaxies as they do here—especially
given that neither we nor denizens of distant galaxies have any claim to
centrality in the grand scheme of things. So you came away from the
preceding chapter realizing that you already understand Einstein’s Principle
of Relativity—the simple statement that the laws of physics don’t depend
on your state of motion.
After showing that you already believe the basic premise of relativity, I
reviewed the strange tale of the star trip that could leapfrog you into the
future. That was disturbing because it violated what your common sense
tells you about the nature of time, and it was but a glimpse of other
disturbing things to come. So, although the idea behind relativity is simple,
its consequences are not. I could plunge further into those consequences
right now, but you’ll find it more convincing if we first explore how the
concept of relativity arose and why it required a genius of Einstein’s stature
to assert what is, in its essence, such a simple idea.
In the next few chapters I’ll back up and give a brief historical overview
of the evolution of our understanding of physical reality. I’ll show how the
idea of relativity became inevitable, but not until the early years of the
twentieth century. As you contemplate this history, remember that you have
the advantage of knowing a lot more than even the most advanced scientists
of earlier times. Pre-Copernican scientists lacked your broader perspective
on a Universe in which Earth’s place is not special. It was only around
1920, when Einstein’s relativity was already becoming widely accepted,
that astronomers recognized the existence of other galaxies. And not until
the last quarter of the twentieth century did instruments like the Hubble
Space Telescope make possible the discovery of distant galaxies moving
relative to us at speeds approaching that of light. The notion of a tennis
match in such a galaxy—a notion that helped convince you of the
obviousness of the relativity principle—would have had little credibility
much before your time.
A Matter of Motion
The history of physical science is intimately connected with our
understanding of the nature of motion. If you ever took a high-school
physics course, you studied motion—a subject you may not have found
particularly exciting. But the study of motion is profound, for several
reasons. First, motion is the source of all change. Imagine a world without
motion: Earth stops rotating, so it’s perpetual daytime. Earth stops
revolving around the Sun, so it’s always the same season. Your body can’t
move, so you’re stuck forever in one spot. Atoms cease moving, so there’s
no chemistry—no release of energy, no change in the substances of the
world. Nothing evolves, transforms, mutates, develops, or otherwise
changes. And there are no molecules jumping across the synapses of your
brain, so there’s no thought. Absent motion, everything stops. Period.
Conversely, if motion does exist (as it obviously does), then
understanding motion will help you understand night and day, the seasons,
the chemical reactions that result from the motion of atoms and the
electrons within them, and even the functioning of your brain as it’s based
in the motion of molecules. Understanding motion at the most fundamental
level goes a long way toward explaining the behavior of matter.
Furthermore, you’ve already seen hints that relativity is going to do
strange things to your commonsense concepts of time and space. Those
concepts are themselves closely associated with the idea of motion. What
does it mean to move? It means getting from one place to another, and
doing so in some time. Whatever else motion means, it involves passing
through time and through space. So motion holds the key to understanding
time and space.
Heresy!
Among the most profound advances in all of science is Polish astronomer
Nicolaus Copernicus’s suggestion that the Sun, rather than Earth, is the
center of the Universe, the point about which all other celestial bodies
revolve (Figure 3.1b). Copernicus (1473–1543) first hit on this idea early in
the sixteenth century and expounded it fully with the 1543 publication of
his De revolutionibus orbium coelestium libri vi (“Six Books Concerning
the Revolutions of the Heavenly Orbs”). This new idea—heretical because
established church dogma was rigorously Aristotelian—was the first in a
series of intellectual steps that stripped Earth of its special place in the
grand scheme of things. Einstein’s relativity, as I outlined it in the preceding
chapter, is the ultimate step in that series.
Copernicus’s heliocentric proposal was truly revolutionary, but in other
ways his new model for the Universe clung conservatively to older ideas.
Although he removed Earth from its central place, Copernicus nevertheless
maintained the distinction between terrestrial and celestial realms. For him,
the heavens remained the realm of perfection. Sun, planets, and stars were
perfect bodies, and they moved, appropriately, in perfect circles. Having
Earth and the other planets in circular paths around the Sun helped explain
many astronomical observations that had puzzled the ancients. In particular,
the retrograde motion of the planets and their apparent brightening and
dimming over time both followed naturally from the rather complicated
path a Sun-orbiting planet describes when viewed from a Sun-orbiting
Earth. Copernicus’s placement of Earth among the other planets raised the
question of how the terrestrial realm, with its imperfections and pestilences,
could be part of the heavens. Nevertheless, Copernicus maintained the
terrestrial/celestial distinction and did not change the Aristotelian answer to
the question, Is there a natural state of motion? On Earth, the Copernican
answer remained “at rest” and in the heavens it remained “motion in perfect
circles.”
Copernicus’s ideas inspired the Danish astronomer Tycho Brahe (1546–
1601) to undertake extensive, regular, and highly accurate observations of
stars and planets. Tycho ultimately rejected both the Ptolemaic and
Copernican views in favor of his own more complicated Earth-centered
model. When Tycho died, he left his prolific set of astronomical data to his
assistant, the German astronomer Johannes Kepler (1571–1630). Working
especially with Tycho’s data for Mars, Kepler conceived a radically simple
way to avoid the complex combinations of circles needed in the models of
Ptolemy, Copernicus, and Brahe. Compounding the Copernican heresy,
Kepler dispensed with the notion that celestial bodies must move in perfect
circles. Instead, he showed, the simplest explanation for planetary
observations was that the planets move in elliptical orbits, with the Sun not
at the center but at a special point called the focus of the ellipse (see Figure
3.2). And instead of incorporeal spirits pushing planets in their circular
paths about a passive Sun, Kepler proposed that the Sun itself exerts the
push that gives the planets their elliptical motion. Furthermore, Kepler
succeeded in formulating quantitative laws that describe how the speed of
the planets varies as they move about the Sun. Although Kepler described
planetary motions accurately, his theory couldn’t explain them.
Philosophically, though, he had taken a big step—freeing the celestial realm
from the constraint of perfectly circular motion.
Fig. 3.2 Kepler showed that each planet moves in an elliptical orbit, with the Sun at a special point
called the focus. The ellipse shown here is highly exaggerated; most planets’ elliptical orbits are
nearly but not quite circular.
A Theory of Everything?
Newton’s three laws of motion, coupled with his law of universal
gravitation, seemed capable of explaining essentially all physical
phenomena. With Newtonian mechanics as a foundation, physicists could
understand not only motion in the cosmos but also such seemingly
unrelated phenomena as the behavior of gases under conditions of changing
temperature and pressure. Although at first glance gas behavior seems to
have little to do with Newton’s laws, in fact it follows directly by applying
those laws to the motions of the individual molecules constituting the gas.
So great was the triumph of Newtonian physics that science held the hope
of explaining all physical phenomena ultimately in terms of the motions of
particles obeying Newton’s laws.
Galilean Relativity
At the heart of Galileo’s and Newton’s new understanding of motion is the
idea of uniform motion as a natural state, needing no further explanation. It
was this idea that helped Galileo accept Copernicus’s moving Earth.
Because everything on Earth shares a common motion and continues
naturally in that state, Earth’s motion is not obvious to us. Indeed, both
Galileo and Newton knew that the laws of motion didn’t favor a particular
place or state of motion. Galileo reasoned, for example, that a rock dropped
from a ship’s mast would behave just the same way if the ship were moving
steadily through calm water as it would if the ship were at rest. In both
cases the rock would fall straight down and land at the base of the mast. We
express this idea more generally with the principle of Galilean relativity,
which says that the laws of motion work equally well for anything in
uniform motion.
Your immediate surroundings—specifically, those things that participate
in the same motion as you—constitute what physicists call your frame of
reference. Earth is the frame of reference for someone at rest on the planet,
and Galileo’s ship constitutes the reference frame for someone doing
experiments on the ship. Put in terms of reference frames, Galilean
relativity becomes the statement that the laws of motion are valid in any
uniformly moving frame of reference.
Galilean relativity should sound a lot like the idea I introduced in the
preceding chapter: the laws of physics are the same, regardless of one’s
state of motion. That’s essentially what Galilean relativity says, especially
because, in Galileo’s and Newton’s time, many scientists believed that the
laws of motion could explain all of physical reality. So a statement about
the laws of motion becomes, to a Newtonian believer, a statement about all
of physics. Here’s the important point: Galileo and Newton, like you
yourself, already understood and accepted the principle of relativity. For
them, as for you, there was no conceivable experiment that could answer
the question Am I moving? Galileo’s rock dropped from a ship’s mast
provides an example: the rock always lands right at the base of the mast, so
the outcome of the experiment can’t be used to determine whether the ship
is moving. Other questions, such as Is the ship moving relative to the
shore?, Am I moving relative to Earth?, or Is Earth moving relative to the
Sun? are answerable—but these are questions about relative motion. That’s
why the statement that the laws of physics are the same, regardless of one’s
state of motion, is a relativity principle. It implies that motion itself is
undetectable and therefore meaningless; all that matters is relative motion.
Ultimately, the relativity principle traces back to Galileo’s recognition that
motion—at least uniform motion—is a natural state that requires no
explanation.
An everyday example should remind you how obvious and simple an
idea is Galilean relativity. Imagine you’re on an airplane, cruising through
calm air at a steady 600 miles per hour. You’re served a snack of airline
peanuts. Do you need to think about the airplane’s 600-mph motion in order
to get the peanuts successfully to your mouth? Do you need to modify your
understanding of the laws of motion to describe the peanuts’ behavior in the
“moving” airplane? Of course not! The airplane’s “motion” is irrelevant.
Newton’s laws work as well on the plane as they do on the ground. That’s
why I put the words “moving” and “motion” in quotes. In the context of
Newtonian physics, it makes no sense to consider that the airplane is
“moving” and Earth isn’t. The laws of physics work the same in both
places, so neither has a claim on some special state of being at rest. If you
want to claim that Earth is really at rest and the plane really moving, then
I’ll challenge you to come up with some physical test that will prove you
correct. Pull down the window shades so you can’t see Earth slipping past,
wait for the plane to be cruising through calm air at constant speed, and
then think up something you can do—that is, a physical experiment—
whose results will be different in the plane than they are if you performed
an identical experiment on Earth. You won’t find it. That’s what it means to
say there’s no experiment that can answer the question, Am I moving? By
looking out the window you can justifiably assert that the plane is moving
relative to Earth, but that’s as far as you’ll get. You’re just as correct in
asserting that Earth is moving relative to the plane. It simply doesn’t matter;
with Galilean relativity, there’s no such thing as absolute motion or absolute
rest. This is all just a rehash of the tennis-match argument from Chapter 2,
where the universal scope of the argument made it even more obvious how
absurd it would be to assert that Earth alone among all the cosmos is truly at
rest.
Throughout this book I’ll be making the point again and again that
there’s no experiment you can do to answer the question, Am I moving? or,
equivalently, that identical experiments performed in different reference
frames give exactly the same results. Let me make very clear what this does
and does not mean. What it does mean is that if you and I, each in uniform
motion but each moving relative to the other, set up and perform identical
experiments, then we’ll get exactly the same results. But those experiments
need to be truly identical in all respects except for our relative motion.
That’s why I set the tennis match in Chapter 2 on an indoor court, to
eliminate any wind that would result in an outdoor court from the ship’s
motion relative to the air. That’s why Galileo’s rock-drop experiment on the
ship should really be in an enclosed space, or maybe behind a sail, to block
the apparent wind caused by the ship’s motion. And to be really careful, I
should do the airplane experiment when the plane is at low altitude, to
minimize changes in gravity with distance from Earth. In practice, it may be
hard to set up circumstances that are identical except for relative motion,
but in principle it’s possible. And if we do, then identical experiments will
give identical outcomes.
What my statement does not mean is that observers in relative motion
must see the results of the very same individual experiment in the same
way. If I watch Galileo’s rock-drop experiment from onboard the ship, I’ll
see the rock fall straight down. If you watch the very same experiment from
the shore, you’ll see the rock following a curved path because it shares the
ship’s motion relative to you. But if we both perform different yet
identically staged rock-drop experiments, you on shore and I on the ship,
then we will get identical results—namely the rock falling straight down,
taking exactly the same time as measured by our different but identical
clocks, and so forth down to the last detail. We simply won’t be able to use
any of our results to decide which of us is moving and which isn’t.
We can express the principle of Galilean relativity in many ways. “The
laws of mechanics are the same for all observers in uniform motion” or
“The laws of mechanics are the same in all uniformly moving frames of
reference” are formal ways to state the principle. To say that “I am moving”
and “I am at rest” are meaningless statements is another way of putting
Galilean relativity. “You can eat dinner on an airplane” or “You can play
tennis on a cruise ship” are statements of Galilean relativity applied to
specific circumstances. “Uniform motion doesn’t matter,” “Uniform motion
is undetectable,” and “Absolute motion is a meaningless concept” are still
other ways to state the principle of Galilean relativity.
There are two subtle distinctions between the Galilean relativity
principle of this chapter and my more general introduction of the relativity
principle in Chapter 2. First, Galilean relativity applies only to the laws of
motion, whereas my earlier relativity principle was a statement about all of
physics. To one who believes that motion can explain all of physics, there’s
no difference. But if we should discover new areas of physics that aren’t
based in the laws of motion, then we’ll need to ask anew whether the
relativity principle holds there as well.
The second distinction is that I left out the qualifying phrase “in
uniform motion” in Chapter 2. But we definitely need that qualification in
expressing Galilean relativity. The laws of motion are decidedly not valid if
you’re not in uniform motion. Try eating those peanuts when the plane
encounters unexpected turbulence or while it’s accelerating down the
runway. Then the peanuts act in a decidedly non-Newtonian fashion. They
don’t stay put on your tray table. Try tossing one into your mouth and it
goes violently astray, landing instead on the floor. Or try playing tennis with
the cruise ship sailing through a hurricane. Your intuitive feel for how the
ball is supposed to behave will fail you; more formally, Newton’s laws of
motion just don’t hold in the nonuniformly moving reference frame of the
storm-lashed ship. So at least as far as Galileo and Newton are concerned,
statements about nonuniform motion, that is, changing motion—are
meaningful. It does make sense to say “this airplane’s motion is changing”
even though it makes no sense, in the context of Galilean relativity, to say
“this airplane is moving.” Absolute motion is meaningless in the
Galilean/Newtonian view, but change in motion is supremely meaningful
and, indeed, is what Newton’s laws of motion are all about.
As I move toward Einstein’s special relativity in the next few chapters,
I’ll continue to maintain the distinction between uniform and nonuniform
motion. Later, in Chapter 14, you’ll see how Einstein’s general relativity
blurs that distinction. But the distinction between the laws of motion (with
their associated Galilean relativity principle) and all the laws of physics will
continue to play a major part as we develop first the special and then the
general theory of relativity.
CHAPTER 4
Waves
We’ll begin with sound, a classic example of a wave. What’s a wave? Think
of an ocean wave or the wave of standing people that sweeps around a
sports stadium. In each case there’s a disturbance of some medium. For
ocean waves, that medium is the water. For stadium waves, the medium
comprises the people in the stadium. The disturbance moves through the
medium, temporarily upsetting the status quo and then moving on. But the
medium itself doesn’t go anywhere, although it may move briefly back and
forth or up and down as the wave passes by. That is, water from the distant
ocean doesn’t actually move toward shore with the waves. Watch a boat in
rough water: the boat bobs up and down as a wave passes, but it isn’t
carried shoreward with the wave (things get a bit more complicated in the
shallow water where the wave breaks or as a surfer intentionally rides a
wave, moving with the wave by sliding down its sloping front). Similarly,
those sports fans rise from their seats, then return; they don’t go anywhere,
even as the wave circles the stadium.
So what does move in a wave? Certainly the disturbance itself—the
displacement of water from its normal flat surface or the displacement of
the sports fans’ bodies from their sitting positions. In the process the wave
carries energy. In our examples that energy is associated with the temporary
lifting and movement of the water as the wave passes or with the standing
people in the stadium. So a good definition of a wave is that it’s a traveling
disturbance that carries energy but not matter.
Two simple ideas seem inherent in our definition of a wave as a
traveling disturbance—ideas that will play a key role in the development of
relativity. First, since a wave is a disturbance, it seems logical that it must
be a disturbance of something—and that something is what we’ve called the
medium for the wave. Second, since a wave is a traveling disturbance, it
must travel with some speed. What determines that speed? Simple: the
properties of the medium. For water waves, speed depends on things like
the density of water and the force of gravity. In fact, we can deduce that
speed by applying Newton’s laws to the motions of the water, taking
account of forces like gravity and water pressure. So we can add water
waves to our growing list of phenomena explained by Newton’s laws.
Now, on to sound. This, too, is a wave, and one that we can also
understand by using Newton’s laws. A sound wave is a disturbance of the
air, consisting of regions of high and low air pressure (Figure 4.1). The
disturbance moves through the air, but the air itself only moves back and
forth, and very slightly at that. The medium for sound waves is clearly the
air. We can directly measure the speed of sound or we can calculate it by
applying Newton’s laws and accounting for the force associated with the
sound wave’s air pressure variations. Either way, the answer for air under
normal conditions is about 700 miles per hour. That, incidentally, is also
about 1,000 feet per second, or one-fifth of a mile per second, which is why
you count the time between lightning and thunder and divide by five to get
the distance to the lightning strike. Each 5 seconds represents 5,000 feet or
about a mile of travel for the sound waves. Implicit in this rule of thumb is
an assumption that will soon become crucially important to us: the speed of
light from the lightning is so fast that the travel time for the light is
negligible.
Fig. 4.1 A sound wave consists of alternating regions of compressed (dark) and rarefied (light) air.
The entire structure moves with the speed of sound (large arrow). The air itself, however, only moves
back and forth as the wave passes (small arrows).
Fig. 4.2 Wave interference. (a) Constructive interference occurs when crests from two waves meet.
(b) Destructive interference occurs when crests meet troughs.
In 1801 the English physician and physicist Thomas Young performed
an experiment that led ultimately to the downfall of Newton’s particle
theory of light. Young let sunlight pass into a darkened room through two
pinholes, illuminating a screen opposite the holes. If light consisted of
particles, one would expect to find two bright spots on the screen opposite
the holes (Figure 4.3a). Instead, Young saw a series of alternating bright and
dark spots on the screen. It’s difficult to imagine how beams of particles
could produce such a pattern, but the wave theory provides an obvious
explanation (Figure 4.3b). Each hole acts like a source of waves, whose
circular crests spread outward from the holes much like ripples from a rock
dropped into a pond. The waves from the two holes meet and interfere. In
some regions, crests meet crests and troughs meet troughs; here the
interference is constructive and the wave strengthens. In optical terms, that
means brighter light. In other regions, crests and troughs meet. Here the
interference is destructive and the wave is diminished. Figure 4.3b shows
that these regions of constructive and destructive interference lie along lines
radiating from the vicinity of the two holes. Where the lines of constructive
interference meet the screen, bright spots appear. At the lines of destructive
interference, the screen is dark.
Fig. 4.3 Shining light through a pair of small holes in an opaque barrier would produce very different
effects, depending on whether light consists of particles or waves. (a) Particles would produce two
bright spots on the screen opposite the barrier. (b) Waves pass through the holes, producing circular
waves, which subsequently interfere. Thick lines mark regions where crests meet crests, and troughs
meet troughs. Here the interference is constructive, making the light brighter. Bright spots appear
where these regions hit the screen. Between, destructive interference results in dark spots. To the
right is an actual photo of the resulting pattern that appears on the screen.
What determines where the bright and dark spots appear? That depends
on the spacing of the two holes and on the wavelength of the light—the
distance between wave crests. The central bright spot, for example, forms
from waves that have traveled equal distances from the two holes and thus
meet “in step,” with crests joining crests and troughs joining troughs. The
dark zones on either side are from waves that have traveled further from
one of the holes by just the right amount that crests meet troughs, making
for destructive interference. Move the holes closer together and the bright
and dark patches move farther apart, making the interference pattern more
obvious. Move the holes farther apart, and the bright and dark patches get
so close that they soon blur together. The reason the wave nature of light
isn’t obvious to us is that the wavelength of light is very small—around 20
millionths of an inch—so it takes very closely spaced holes to make
Young’s interference pattern noticeable.
So now we have an answer to our question about the nature of light:
Waves or particles? That answer, clear from Young’s interference
experiment, is “waves.” But waves of what? To answer that question takes
us on a new and unexpected turn.
These four statements look rather different, but they actually comprise two
pairs that reflect a profound symmetry between electric and magnetic
phenomena. We need to develop that symmetry more fully in order to
answer our question about the nature of light.
Fields
Suppose Earth suddenly vanished. How and when would the Moon, a
quarter-million miles away, “know” that it should abandon its circular orbit
—the result of Earth’s gravity—and begin the straight-line motion that
Newton’s laws tell us it should take in the absence of a force? In Newton’s
view, the Moon would know about Earth’s disappearance instantaneously,
because in Newton’s description of gravity Earth “reaches out” across
empty space and immediately “pulls” on the Moon. Remove Earth, and that
pull—the gravitational force—disappears instantly. This Newtonian idea is
called action at a distance for obvious reasons.
There’s another way to view the Earth–Moon interaction, a way that at
first may seem needlessly complicating and abstract. But it will be crucial
in developing your understanding of light and will become absolutely
essential in the context of relativity. Even at this early stage, you may find
this new view more philosophically satisfying.
Here’s the issue: The action-at-a-distance description of gravity requires
the Moon to know what’s happening, right now, at the location of the
distant Earth. But how can this be? It would be much more believable if the
Moon only needed to know about its immediate vicinity and responded only
to local conditions. Enter the concept of field. We imagine that Earth creates
a kind of influence, called a gravitational field, everywhere in the space
around it. Put an object in Earth’s vicinity and it experiences not a
mysterious action-at-a-distance pull from Earth but rather the gravitational
field right at the point where the object is. It responds to that field by
experiencing a force toward Earth’s center, a force whose strength depends
on the object’s mass and on the strength of the gravitational field. Since
gravity decreases with distance from Earth, so must the strength of the field.
We can represent the gravitational field by drawing arrows that show its
strength and direction at selected points (Figure 4.4).
Conceptually, in introducing the field concept we’ve replaced the direct
but philosophically disturbing action-at-a-distance force of gravity with a
force that arises locally, from the gravitational field at any point in space.
Rather than Earth exerting forces directly, we have the more complex
situation in which Earth creates a gravitational field in its vicinity and
objects respond to that field. What we gain from this complexity is a new
simplicity: now an object doesn’t have to know about the situation at some
distant place but only about what’s happening in its immediate vicinity.
Of course, the ultimate outcome remains unchanged. The Moon, a
spacecraft, and a falling apple still behave as Newton predicted. Only our
description of how that occurs has changed; now these objects respond to
Earth’s gravitational field right where they are, rather than to Earth itself.
But for now the action-at-a-distance and field perspectives predict exactly
the same physical results.
Now back to the question What if Earth suddenly vanished? We can’t
answer that question for sure at this point, but at least the field concept
gives us some wiggle room. What happens at Earth’s location isn’t
important to the Moon; it’s the local gravitational field that determines the
Moon’s motion. So does the gravitational field everywhere vanish
instantaneously when Earth does? Or does it have some sort of independent
existence so it takes awhile for the field way out at the Moon’s location to
learn about Earth’s demise? In fact, the answer is the latter, as we’ll explore
in later chapters. For now, though, suffice it to say that the field concept
gives us a new and important physical entity—in this case the gravitational
field—that can have, at least temporarily, an independent existence.
Fig. 4.4 A representation of Earth’s gravitational field. The length of each arrow gives the strength of
the gravitational field at its location, and the arrows point in the direction of the field, namely, toward
Earth’s center. The field exists everywhere, but I can only show arrows to represent it at a few
selected points.
Electromagnetic Fields
I could have phrased the entire preceding section in terms of the electric
force between charges rather than the gravitational force between Earth and
Moon. I chose the latter only because you’re more familiar with gravity.
Having established the concept of the gravitational field, though, the
electric field follows directly by analogy. We say that one electric charge
creates an electric field everywhere in space and that a second charge
responds to the field in its immediate vicinity. Again we get the same
physical happenings as predicted by the action-at-a-distance electric force,
but now the behavior of a charge is determined by local conditions—in this
case the local electric field. Again we can speculate about what would
happen if one charge suddenly vanished or just moved a bit; would a
second, distant charge respond immediately or would it take a while for the
local field to learn about this change and thus to adjust to it?
If electric charges create electric fields, surely magnets must create
magnetic fields. And they do. Given our understanding of magnetism,
though, we can put all this more carefully: Electric charges create electric
fields, while moving electric charges create, in addition, magnetic fields.
And we can rephrase Faraday’s discovery as a statement about fields: the
loose phrase “changing magnetism gives rise to electricity” becomes “A
changing magnetic field creates an electric field.” This last statement has an
interesting implication: no longer is electric charge the sole source of
electric phenomena, in particular of electric fields. Now something quite
different, namely a changing magnetic field, can also create an electric
field. So electric fields have two possible sources: electric charge and
changing magnetic fields. Notice that the second of these entails a direct
relation between the fields themselves, where one directly creates the other
without the intermediary of electric charge or any other matter.
Let’s look in more detail at the last two of our earlier statements about
electromagnetism, now recast in terms of electric and magnetic fields:
ETHER DREAMS
• • •
(By the way, something about the sound from a moving source does
change, namely, its pitch. When the fire truck is coming toward you, the
siren sounds like it has a higher pitch. When it moves away, the pitch you
hear is lower. The analogous shift for light is what lets us measure motions
in double-star systems.)
In disabusing you of the notion that the speed of light might be c
relative to its source alone, I reverted naturally to the language of ether—
the medium in which light waves are purportedly a disturbance. Despite the
problematic aspects of the ether, what else are we (or, more precisely,
nineteenth-century physicists) to do? Light, after all, consists of waves, and
all other known waves are disturbances of some medium. The speed of each
type of wave—sound waves, water waves, stadium waves, earthquake
waves, etc.—is its speed relative to its particular medium. Why not the
same for light?
If you’re still troubled by the all-pervasive, tenuous, light-wave-
supporting ether, then again I challenge you to tell me relative to what light
travels at speed c. That’s a harder challenge now you know that the answer
cannot be “relative to its source.” So for now we’ll continue to follow the
nineteenth-century physicists’ line of thought, picturing light waves as
disturbances of an ether that pervades the entire Universe. If you really
don’t like the ether, though, hold on—eventually you’ll be vindicated!
A Broader Question
We were led to the ether concept by questioning relative to what light
travels at speed c. I want to convince you now that this question is really a
special case of a more general one: In what frame of reference are the laws
of electromagnetism (i.e., Maxwell’s equations) valid? The two questions
are related because one prediction of the laws of electromagnetism is that
there should be electromagnetic waves and that they should go at the speed
of light, c. When we answer “ether” to the question “Relative to what does
light go at speed c?”, we’re saying that Maxwell’s prediction of
electromagnetic waves that go at c is really valid only in a frame of
reference at rest with respect to the ether. Observers in a reference frame
moving through the ether will measure some other speed for light relative to
themselves; thus for them, in the context of their reference frame, the
predictions of Maxwell’s equations won’t be valid. That’s why my two
questions, one about the speed of light and the other about the validity of
Maxwell’s equations, are essentially equivalent. Maxwell’s equations
predict electromagnetic waves going at c, so those equations can only be
valid in a frame of reference where one will, in fact, measure c for the speed
of light.
So what’s the answer to our new question, In what frame of reference
are Maxwell’s equations valid? It’s obvious: In the context of nineteenth-
century physics, there’s only one frame of reference in which Maxwell’s
prediction about electromagnetic waves is valid, and that’s a frame of
reference at rest with respect to the ether.
Dichotomy in Physics
In Chapter 3 we discovered the principle of Galilean relativity, which states
that Newton’s laws of motion are valid in all uniformly moving frames of
reference. So if we ask explicitly in what frame of reference Newton’s laws
of motion are valid, then the answer is “in any uniformly moving frame of
reference.”
This question we’ve just asked about Newton’s laws of motion is
exactly the same question we asked in the previous section about Maxwell’s
laws of electromagnetism. But there we found quite a different answer.
Maxwell’s electromagnetism, it seems, isn’t valid in just any frame of
reference. Rather, the laws of electromagnetism should be valid only in one
very special frame of reference—a frame of reference at rest with respect to
the ether. Put another way, the laws of motion obey a relativity principle but
the laws of electromagnetism seem not to. So although there’s no
experiment we can do with the laws of motion to answer the question, Am I
moving?, there should be electromagnetic experiments that can answer this
question. That is, the concept of absolute motion is meaningless for
mechanics, but apparently it has meaning for electromagnetism. Although
there’s no privileged state of motion for mechanics, there seems to be a
privileged state for electromagnetism—namely, being at rest relative to the
ether.
Why the dichotomy? Why should one main branch of physics
(mechanics) not care about states of motion, while the other
(electromagnetism) does? Wouldn’t it be simpler and more coherent if both
branches of physics obeyed the relativity principle, or both didn’t?
Think back to the second chapter, where I asked about a tennis match
played on a cruise ship, on Venus, and on a planet in a distant galaxy
moving away from Earth at nearly the speed of light. You wisely and
logically agreed that it made perfect sense to expect that tennis playing—a
manifestation of the laws of motion—would work the same in all those
contexts or, as we would now say, in all those different reference frames.
That is, you intuitively accepted the principle of relativity as applied to the
laws of motion. Then I asked about heating a cup of tea in a microwave
oven, and you agreed that the microwave oven, like the tennis ball, should
also behave the same way in the different reference frames. But it’s
electromagnetism, not mechanics, that governs the microwave oven.
Through our exploration of the question, Relative to what does light go at
speed c?, we’ve just found that the laws of electromagnetism seem not to
obey the relativity principle. That is, the laws of electromagnetism should
not be valid in all reference frames—and electromagnetic experiments
should therefore give different results in reference frames that are in
different states of motion. So the microwave oven shouldn’t work the same
way in that distant galaxy moving at nearly c as it does on Earth!
What’s wrong with your intuition from Chapter 2? Again, the difficulty
is with the ether. When you blithely agreed that the microwave oven should
work the same way everywhere, you weren’t taking the nineteenth-century
view, with its unique reference frame of the ether, in which frame alone the
laws governing the oven should be valid.
So why not get rid of the ether, now troublesome not only because of its
improbable characteristics but because it’s also the cause of an illogical and
unsatisfying dichotomy between two branches of physics? That dichotomy
runs counter to your good sense that the laws governing physical reality,
whether tennis balls or microwave ovens, should work the same everywhere
and regardless of one’s state of motion. If you abandon the ether, then you
could eliminate that dichotomy.
But if you abandon the ether, then I challenge you once again to answer
the question, Relative to what does light go at speed c? or, equivalently, In
what frame of reference are the laws of electromagnetism valid? If you give
the same answer that I’ll happily accept for the laws of motion—“in all
uniformly moving reference frames”—then you’ll vindicate your intuitive
sense from Chapter 2 that the microwave oven should work the same in all
states of motion. But if you give that answer, you’ll find yourself on the
edge of a philosophical abyss. That’s because you’ll be insisting on a
seeming contradiction: that two different observers must each find valid the
Maxwellian prediction that light waves move at speed c—even if those
observers are moving relative to each other! Better not go there, at least not
yet; instead, we’ll stick for now with the nineteenth-century ether concept
and explore further its implications.
Insulting Copernicus
Consider first the possibility that Earth isn’t moving relative to the ether. I
can think of only two ways for this to be the case. First, the ether might be a
fixed substance that extends throughout the Universe. Then Earth alone
among all the cosmos would be at rest relative to the ether. I say “alone”
because all other celestial objects—the Moon, Mars, Venus, the other
planets, the Sun, other stars in our galaxy, and the other galaxies in the
Universe—all are moving relative to Earth. So if Earth is at rest relative to
the ether, then it alone is at rest. That makes us pretty special. If we’re the
only beings at rest relative to the ether, then Maxwell’s equations are valid
only for us, and only we measure c for the speed of light. Observers on
other celestial bodies measure different speeds for light in different
directions, and for observers moving very fast relative to Earth—like those
in distant galaxies—that effect must be dramatically obvious.
Copernicus would turn in his grave! It’s hard to imagine a worse insult
to the Copernican revolution than to make our planet so special that one of
the two main branches of physics is valid only on Earth. I spent most of
Chapter 3 presenting a history of science that led steadily away from the
notion of Earth being a special, privileged place in the Universe. Do you
really want to return to parochial, pre-Copernican ideas? Do you really
think you and your planet are so special that, in all the rich vastness of the
Universe, you alone can claim to be “at rest”?
On purely philosophical grounds, we should reject the notion that Earth
alone could be at rest relative to the ether. Now, philosophy isn’t science,
and I hasten to add that there’s plenty of good scientific evidence to support
this view. For example, we observe light-emitting processes in distant stars
and galaxies that seem to work the same there as they do here on Earth.
That suggests we don’t have any special status vis-à-vis the laws of
electromagnetism. So we can confidently reject the idea that Earth alone is
at rest relative to the ether.
Ether Drag
It might still be possible for Earth to be at rest relative to the ether if our
planet somehow “dragged” the surrounding ether with it. Presumably other
planets and celestial bodies would do the same, so each would be at rest
relative to its local blob of ether. Then observers everywhere and in
different states of motion would find the laws of electromagnetism to be
valid, and no one would have any claim to be special. Copernicus would be
a lot happier with that!
So does Earth drag the ether with it? Astronomical observations dating
to 1725 provide a clear answer. A simple analogy will help you understand
these observations. In Figure 5.2a you’re standing, holding an umbrella in
the rain. Obviously the best approach to keeping dry is to hold the umbrella
directly overhead. But what if you run through the rain? Now it’s better to
tilt the umbrella, holding it at an angle (Figure 5.2b). Figure 5.2c shows
why: viewed from your frame of reference, the rain is coming down at an
angle, and you want to hold the umbrella so the rain still hits the umbrella
top straight on. Since the rain is falling at an angle, you should hold the
umbrella at the same angle. If you run in the opposite direction, you should
still hold the umbrella tilted at the same angle in front of you, but now this
will be a different absolute direction.
However, suppose that somehow you drag a big blob of air with you as
you run—so big a blob that rain falling into it has time for the force of the
moving air to accelerate it to your running speed before it hits you or your
umbrella. Figure 5.2d shows the situation from your point of view. As you
run through the rain, the rain outside your blob of dragged air falls at an
angle as seen from your reference frame. But inside the blob, the moving air
accelerates the rain until it shares the blob’s motion. So now, relative to
you, it’s falling straight. The upshot is that you don’t have to tilt your
umbrella. Rather, you’ll stay driest if you hold it right overhead.
Fig. 5.2 An analogy for the aberration of starlight. (a) Standing still in vertically falling rain, you hold
your umbrella straight overhead to keep driest. (b) Running, you tilt your umbrella. (c) The situation
in (b), shown from the runner’s frame of reference. In this frame, the rain falls at an angle. (d) If the
runner drags a large blob of air, then rain entering the blob will take on the blob’s motion, and thus
will fall vertically relative to the runner.
CRISIS IN PHYSICS
• • •
Can we get any idea of how much difference to expect? Not entirely,
because we don’t know Earth’s speed relative to the ether. That, after all, is
what we’re trying to find out. But we know that Earth orbits the Sun at
some 20 miles per second, and because the orbit is circular, the direction of
that motion changes throughout the year. So even if the Sun, by some
bizarre coincidence, happened to be at rest in the ether, Earth would be
moving at some 20 miles per second. And whatever Earth’s speed,
physicists knew it had to change by some tens of miles per second
throughout the year, as Earth first heads in one direction at 20 miles per
second relative to the Sun and then, 6 months later, at 20 miles per second
in the opposite direction. If they could build a device to detect a speed of
that magnitude, they could answer the question of Earth’s motion through
the ether.
Now, 20 miles per second sounds fast but it’s slow compared with the
speed of light, some 186,000 miles per second. Worse, it turns out that a
successful measurement of Earth’s motion requires detecting the square of
Earth’s speed in relation to the square of the 186,000-mile-per-second speed
of light. That’s like measuring the difference between the numbers 1 and
1.00000001. So great was the challenge that many nineteenth-century
experimenters thought it impossible.
If the beam paths in Figure 6.2 were exactly equal, and if there were no
ether wind, then the beams would return exactly in step and the light waves
would interfere constructively. An observer looking into the viewer would
see bright light. But suppose the ether wind delayed one beam by exactly
enough to make its troughs line up with the other beam’s crests. Then we
would have destructive interference and darkness in the viewer. That, in
principle, is how Michelson’s interferometer could detect Earth’s motion.
Actually, it’s both a bit more complex but also simpler than that. First of
all, it’s impossible to get the two ordinary mirrors at exactly right angles,
and in any event the light rays in the beams aren’t exactly parallel but
diverge slightly and hit the half-silvered mirror at slightly different points.
The result is that light takes a whole lot of different paths, differing slightly
in length, on each of the two legs of Michelson’s apparatus. Some of it
returns in step, interfering constructively, and some of it returns out of step.
The result is not simply bright or dark in the viewer but a pattern of
alternating light and dark bands. (Figure 6.2 includes a photo of interference
bands from a modern-day Michelson interferometer). Furthermore, it’s
impossible to get the distances to the two mirrors exactly the same. But all
that does is to alter just which light interferes constructively and which
destructively. Even if the path lengths aren’t the same, we’ll still get a
pattern of light and dark bands. It’s just that the positions of the bands will
be a bit different.
None of these subtleties matters, though; in fact, they make the
experiment easier, since we don’t have to worry about getting the path
lengths equal or the mirrors exactly perpendicular. The reason they don’t
matter is that we aren’t really interested in the interference pattern itself,
which includes the effects not only of the ether wind but also of the various
imperfections in the instrument like path lengths and mirror alignment. But
now suppose we rotate the whole apparatus through 90 degrees. The path
that was initially along the ether wind is now at right angles to it, and the
one that was originally at right angles is now along the wind. Whatever the
relative timing for the light beams on the two paths was originally, the
important point is that the timing should now change. That change is due
entirely to the ether wind because nothing about the apparatus itself has
changed except its orientation in the wind. And how do we detect that
change in the relative travel times on the two paths? Simple: we watch the
interference pattern. A change in travel times should result in a shift in the
positions of the bright and dark interference bands. That shift, viewed as the
apparatus is rotated, gives a direct indication of Earth’s motion through the
ether. The magnitude of the shift is a measure Earth’s speed.
Fig. 6.3 Michelson and Morley should have seen a shift of nearly half a fringe, putting a dark band
where a light band had been. Photos, from an actual Michelson–Morley setup, are offset to show the
fringe shift.
Contradiction!
The starkly negative outcome of the Michelson–Morley experiment stands
as one of the most important experimental results in all of science. To see
why, remember where we are, logically, in 1887. The realization that light is
a wave, specifically an electromagnetic wave propagating at speed c, raises
the question of the medium in which light propagates. Nineteenth-century
physics answers that question by proposing the ether as the medium for
light and other electromagnetic waves. It then makes sense to ask about
Earth’s motion relative to the ether. Aberration of starlight shows that Earth
can’t be at rest relative to the ether, so Earth must be moving. Earlier
experiments fail to detect that motion, but they suffer either from
conceptual flaws or insufficient sensitivity. Now, in 1887, come Michelson
and Morley with an experiment much more sensitive than what’s needed to
detect something that must exist, namely, Earth’s motion through the ether.
And yet the experiment fails. Earth must be moving through the ether, yet
the Michelson–Morley experiment shows that it isn’t. That’s a pretty stark
contradiction, and it shook the foundations of physics in the late nineteenth
century.
I repeat: it’s the foundations of physics—the basis of our whole
understanding of physical reality—that are shaking, not some minor
inconsequential details. Why is this contradiction so profound, so dire?
Because it concerns a fundamental and sweeping prediction of one of the
two basic branches of physics, specifically Maxwell’s electromagnetism
with its prediction of electromagnetic waves propagating at speed c. That
prediction immediately gives voice to the question, Speed c relative to
what? It’s in attempting to answer that question that the contradiction
inherent in Michelson–Morley arises. If we can’t resolve that contradiction,
then there’s something drastically wrong with our supposed understanding
of physical reality.
No wonder nineteenth-century physicists sought at all costs to explain
away the negative Michelson–Morley result. Not to do so would be to admit
a logical fault so deeply ingrained as to threaten the entire edifice of
physics. The experiment itself seemed beyond reproach, so physicists
sought some explanation, some excuse, for its failure to detect Earth’s
motion. Michelson himself concluded disappointedly that Earth must be at
rest relative to the ether after all, despite the apparently opposite implication
of the starlight aberration observations. Others made more radical
suggestions. In particular, the Dutch physicist H. A. Lorentz and the Irish
physicist George Fitzgerald independently proposed that the ether squeezes
objects moving through it, contracting them in the direction parallel to their
motion. This contraction would shorten the ether-wind-aligned path in the
Michelson–Morley experiment and thus reduce the travel time for light
along that path. If the contraction were just right, the effect would eliminate
the time differences on the two paths and would therefore explain the
negative result of the experiment.
The Lorentz–Fitzgerald contraction would be very small, amounting to
a decrease of only about 3 inches in Earth’s 8,000-mile diameter. But that’s
all it would take to explain the Michelson–Morley result and to restore the
logical consistency of physics in the nineteenth century. But given ether’s
tenuous nature, why should the contraction occur at all? And why should it
be the same for all substances, regardless of what they’re made of? There
was no satisfying answer to these questions, and the Lorentz–Fitzgerald
contraction seemed a very ad hoc way out of the Michelson–Morley
contradiction.
Ingenious though the nineteenth-century physicists were, they weren’t
quite ingenious enough to break free of their nineteenth-century mindset
and resolve the contradiction in a simple, fresh, and radical way. That
resolution had to wait until 5 years into the next century, and when it came
it was truly revolutionary yet profoundly simple.
CHAPTER 7
Albert Einstein was 8 years old when Michelson and Morley performed
their 1887 experiment. Slow to speak, socially withdrawn, and stormy of
temper, the young Einstein seemed not especially promising. But a
magnetic compass his father showed him had evoked in 5-year-old Albert
the first glimmers of his fascination with the deepest nature of physical
reality. The compass needle’s mysterious response to something unseen and
unfelt made a lasting impression and stirred Einstein’s lifelong search for
the hidden principles governing the physical Universe.
That’s it. This one sentence implies all of Einstein’s special theory of
relativity. It’s conservative because it asserts for electromagnetism—indeed,
for all of physics—what had been known for centuries about mechanics,
namely, that there’s no favored state of motion, no preferred frame of
reference. If you look back in Chapter 3, where I introduced the principle of
Galilean relativity, you’ll find that Einstein’s statement is identical except
that the phrase “laws of motion” becomes generalized to “laws of physics.”
So in this sense Einstein’s relativity is nothing new. It’s just a generalization
of Galilean relativity to all of physics—including electromagnetism.
Einstein’s relativity is clearly simple, stated in its entirety in just one brief
sentence, but it’s also radical. For Einstein to state it, and for you to
understand it fully, requires dramatic restructuring of deep-seated
commonsense notions about the nature of space and time. That is what the
rest of this book is about, and it’s what makes relativity seem a challenging,
even mind-boggling subject. But I want to stress again and again that, in
essence, relativity is simplicity itself. It’s all based on the fact that there’s no
preferred state of motion for describing physical reality; that you and I, as
long as we’re both moving uniformly, will discover exactly the same
underlying physical laws—even though we may be moving relative to each
other. That’s exactly the idea that I introduced in Chapter 2, when I
convinced you that a tennis match and a microwave oven should work the
same on Earth, on Venus, and on that planet in a distant galaxy moving
away from Earth at nearly the speed of light. Motion—as long as it’s
uniform motion—simply doesn’t matter. There’s no one who can claim,
“You’re moving and I’m not.” All states of motion—at least uniform
motion—are equally valid. There’s no such thing as being absolutely at rest
or in motion. Only statements about relative motion make sense.
By the way, you might be questioning that caveat about uniform motion.
Why shouldn’t all states of motion be equivalent? We’ll get there
eventually, as Einstein did by 1915. But first things first: Einstein’s 1905
relativity is the special theory of relativity. Special here doesn’t mean that
the theory is particularly great and wonderful (although it certainly is
both!), but special in the sense of being specific and limited. Special
relativity is limited to the special case of reference frames in uniform
motion. Einstein’s general relativity, which I’ll introduce in Chapter 14,
removes that limitation.
(Historically, Einstein actually presented relativity in the form of two
postulates. The first was his assertion that the Principle of Relativity, which
I stated above, applies to all of physics. The second, which Einstein noted
“is only apparently irreconcilable” with the Principle of Relativity, asserts
that the speed of light is the same in all uniformly moving reference frames.
A more modern approach takes the view that the second postulate follows
from the first; indeed, by 1910 physicists had shown rigorously that the
second postulate is superfluous.)
Here’s how the Principle of Relativity takes care of those pesky
nineteenth-century dilemmas and contradictions. The Principle asserts that
all the laws of physics are the same in all uniformly moving reference
frames. Among the laws of physics are Maxwell’s equations of
electromagnetism. Those equations lead to the prediction that there should
exist electromagnetic waves and that those waves should propagate at a
particular speed—the speed of light, c. Speed c relative to what?
Nineteenth-century physicists foundered around in the ether trying to
answer this question, without success. But the Principle of Relativity
provides a simple answer. Because the prediction of electromagnetic waves
propagating at speed c is a prediction of the laws of physics, and because
those laws are valid in all uniformly moving reference frames, it must be
the case that electromagnetic waves propagate at speed c as measured in
any uniformly moving reference frame.
STRETCHING TIME
• • •
Fig. 8.1 Four observers each measure the same value c for the speed of light relative to themselves,
even though they’re in motion relative to one another.
Figure 8.1 also shows a friend driving by in a car at 60 miles per hour,
toward the traffic signal. I’ve given your friend a perfect meter stick and a
perfect stopwatch, identical to yours. Your friend uses exactly the same
experimental technique on exactly the same light flash, to determine the
speed of the flash. And what does she find? Because she’s heading toward
the traffic signal, you might guess that she’ll measure a slightly higher
speed than your 299 792 458 meters per second—higher by her speed of 60
miles per hour, or about 27 meters per second. But that answer isn’t
consistent with the Principle of Relativity, which states that the laws of
physics are the same in all uniformly moving reference frames. One
consequence of those laws is the existence of electromagnetic waves—
including light—that propagate at speed c. Both you and your friend are in
uniform motion, so that particular consequence of the laws of physics must
hold for both of you. Both of you must measure the same value for the
speed of light c. With perfect meter sticks and stopwatches, that value will
be 299 792 458 meters per second.
Let me make it perfectly clear that I’m saying that each of you gets
exactly the same value for c. You might think your speed relative to your
friend is so small that you just don’t notice the difference. So to make
things more dramatic, Figure 8.1 shows two more observers, one in a jet
plane going at 600 miles per hour and another in a rocket moving toward
the traffic signal at half the speed of light. Surely an astronaut on the rocket
sees the light approaching at 1.5c. But no! That result would be inconsistent
with relativity for the same reason it would be for the driver of the car. You,
your friend, the airplane pilot, and the astronaut are all in uniform motion
and, therefore, the laws of physics are equally valid for all of you. None of
you—including yourself standing on the road—can claim in any absolute
sense to be at rest. None of you can say that the laws of physics are correct
only for you. You’re all in equally good situations—reference frames—for
exploring physical reality, and one consequence is that you’ll all measure
precisely the same value for the speed of light, c.
But how can this be? How can different observers, moving relative to
one another, still measure the same speed for the same light? Ultimately the
answer is simple: The laws of physics are the same for all observers in
uniform motion, and the invariance of c follows directly from that principle.
Whenever you find yourself asking, How can this be? about relativity, the
answer always lies in the Principle of Relativity. Remind yourself how
intuitive that principle seemed in Chapter 2, and how the experimental
evidence and scientific quandaries of the nineteenth century led Einstein to
affirm the relativity principle as the basis of his theory. If you still don’t like
what relativity has to say, then you answer the question, Speed c relative to
what? in a way that’s consistent with experiments like Michelson–Morley
and that doesn’t put Earth, alone among all the cosmos, in a favored
position.
Fig. 8.2 Take away the ground and the traffic signal, and it’s obvious that the roadside and rocket
observers are in perfectly equivalent situations. Both think of themselves as being at rest, and both
measure c for the speed of light. (a) From the roadside observer’s viewpoint, the rocket is moving to
the right at half the speed of light. (b) From the rocket’s viewpoint, the roadside observer is moving
to the left at half of c.
Time Dilation
Exactly how do measures of time differ in different reference frames? Here
I’m going to argue that difference qualitatively and present without proof
how it works quantitatively. (Those who like math will find the details in
the Appendix.) First, though, I need to clarify an idea that’s already familiar
to you—the idea of an event. Events play a major role in relativity, because
they involve both time and space. An event is an occurrence, something that
happens at a specific place and time. Your birth is an event; it occurred
somewhere and at some time. Your reading this paragraph is another event;
it, too, is happening somewhere and at some time. The two events are not
the same, because they’re separated in time and probably in space as well.
(Although they could occur at the same place, at least in Earth’s reference
frame, if you happen to be reading in the very spot where you were born.)
An event is completely specified by giving a time and a place. For example,
the most recent major earthquake to strike the San Francisco Bay area
resulted from a specific event—a slippage along the San Andreas fault. The
event’s location (place) was at Loma Prieta, in the Santa Cruz Mountains,
60 miles south of San Francisco and at a depth of 11 miles below Earth’s
surface. Its time was 5:04 PM PST on October 17, 1989. Another event, the
landing of the NEAR spacecraft, marked humankind’s first contact with an
asteroid. This event occurred at time 12:02 PM PST on Monday, February
12, 2001, and its place was the asteroid Eros, then 196 million miles from
Earth. Since every event has a time and a place, it makes sense to talk about
the time intervals and spatial separations between any pair of events. For
our two events, the time interval is approximately 11.28 years and the
spatial separation is 196 million miles—as measured in a frame of reference
at rest with respect to Earth.
I’m now going to convince you that the time interval between two
events cannot be the same for two observers in motion relative to each
other. The argument I’ll use is based solidly in the Principle of Relativity
and its consequence, the invariance of the speed of light. I’ll use a rather
artificial-seeming situation involving a bouncing light beam, but don’t come
away thinking this is only about light or about the particular situation I’ll
describe. Rather, it’s about the nature of time. The example I present serves
to illuminate what is, in fact, a universal aspect of time—namely, that
measures of the time between two events differ in reference frames that are
in relative motion.
So here’s the situation, shown first in Figure 8.3a. There’s a rectangular
box with a light source at one end and a mirror at the other. A brief flash of
light leaves the source. We’ll call this occurrence—the departure of the light
flash from the source—event A. The light travels up the box, hits the
mirror, reflects, and returns to the source. We’ll call event B the return of
the light to the source. That is, event B is completely specified by stating
that it occurs at the bottom of the box (B’s place) at the exact instant that
the light flash reaches that place (B’s time). The light’s round-trip takes
some time, which is the time between events A and B as measured in a
frame of reference at rest with respect to the box. We need that clarification
because the time between two events is not absolute but depends on one’s
reference frame. That’s just what I’m trying to demonstrate.
Fig. 8.3 A “light box,” used as a clock to measure the time interval between two events. Event A is a
light flash leaving the source. Event B is the light returning to the source, after reflecting off a mirror
at the top of the box. Dashed line is the light path. (a) The situation in a reference frame at rest with
respect to the box. The light travels a total distance twice the box length L. (b) The situation in a
reference frame in which the box moves to the right at speed v. The light travels farther, but because
it has the same speed c, the time between the two events must be longer.
You could calculate the time between the two events quite easily if you
knew the length of the box. Let’s call that length L. Then the light goes a
total distance of twice L, because it makes a round-trip. And how fast does
the light go? At speed c, as always. Knowing the distance and speed, it’s
simple to calculate the time—and I do that explicitly in the Appendix. But
for this qualitative argument, all you need to know is that the light makes a
round-trip of twice the box length and that it does so at speed c.
I warned you earlier to be suspicious of pictures in relativity books.
Every picture is drawn from a particular point of view, that is, from the
perspective of a particular reference frame. To understand what’s going on,
you need to know exactly what frame that is. In the case of Figure 8.3a, the
reference frame of the picture is at rest with respect to the box, so it’s in that
reference frame that we’ve been discussing the time between events A and
B.
Now let’s look at the situation from another reference frame, one
moving relative to the box. Equivalently, the box is moving relative to this
new reference frame. So suppose the box is moving to the right, at some
speed v, relative to the new frame (v here is for “velocity”). Figure 8.3b
shows the situation in this reference frame. Because the box is moving
relative to the reference frame, I’ve needed to show it in different locations
at several different times. In particular, it’s shown at the time of event A,
when the light flash leaves the source; at the time of event B, when the light
returns; and at an intermediate time when the light reflects off the mirror at
the end of the box.
We want to know the time between events A and B as measured in the
reference frame of Figure 8.3b. To find that time we need the distance the
light travels between its departure from and its return to the source, and we
need to know how fast it goes. Without doing any math, you can see that
the light’s path in Figure 8.3b is longer than it is in Figure 8.3a, which was
drawn from the perspective of a reference frame where the box was at rest.
That’s because the light, relative to the reference frame of Figure 8.3b, takes
a diagonal path heading to the mirror and another diagonal path coming
back. Those diagonals are each necessarily longer than the length L of the
box, since they incorporate both that length and some motion of the light
sideways, along the direction of the box’s relative motion. Again, there’s
nothing fishy about this. If I’m in a bus and I throw a ball straight up, it
goes straight up and down relative to me. But if you’re standing by the road
you see the ball take a curved path that’s longer than the path it takes in a
reference frame at rest with respect to the bus. Similarly, the light takes a
longer path in a reference frame relative to which the box is moving.
The only other thing we need to get the time between the light’s
departure and return—that is, the time between events A and B—is the
light’s speed. Here’s where the relativity principle comes in. Again, one
consequence of that principle is the invariance of the speed of light. So the
light goes at c relative to the reference frame of Figure 8.3b. But it goes
farther in this reference frame than it did in the frame of Figure 8.3a, at rest
with respect to the box. Since it has the same speed, c, in both frames, it
must take a longer time in the reference frame with the longer path—that is,
in the reference frame relative to which the box is moving. So we’re forced
to conclude that the time between events A and B in a reference frame in
which the box is at rest is shorter than the time between the same two
events in a reference frame with respect to which the box is moving.
Let me be clear that I’m not saying simply that the light took two
different trips, one of them longer, and therefore that the longer trip took a
longer time. That would be obvious without relativity. Rather, the same
light took one and the same trip, and we examined that trip from the
viewpoints of two different reference frames. The beginning and end of that
trip are the events A and B, and observers in both frames agree about what
those events are and that they indeed mark the endpoints of the light’s trip.
What they disagree about is the time between those same events.
That’s all well and good, you might say, in this highly artificial situation
where the two events in question happen to involve a light beam that must
go at c in any reference frame. Surely, you say, this has nothing to do with
everyday events of the sort you and I experience. But it has everything to do
with such events. In principle, I could have set up a “light box” at any event
you care to name—for example, at your birth—and arranged for the flash to
depart at the instant of that event. By moving the light box at the right speed
I could arrange for it to arrive at a later event of your choosing—for
example, your reading this chapter. And by choosing the length of the box
(it would have to be very long for your birth and reading events!), I could
arrange for the light to return to the source just as the box arrived at where
you are as you read this chapter. All the arguments I made from Figure 8.3
would apply here, and we’re forced to conclude that the time between the
events of your birth and your reading this chapter is different in different
reference frames. This is about time, not about light beams and mirrors. The
light box serves as a device for exploring the nature of time and for
concluding that the time between two events is relative to one’s frame of
reference. But the light box doesn’t cause that difference; the difference is
intrinsic in the nature of time. Measurement of the time interval between
two events with ordinary clocks would, in principle, reveal exactly the same
discrepancies between different reference frames, regardless of the presence
or absence of the light-box device. In fact, all good clocks in a given
reference frame will measure the same time between two events, and that
time will be different from that measured by identical clocks in another
reference frame in relative motion. By a “good clock,” I mean any device
that accurately measures the passage of time. That includes the light box in
Figure 8.3, an atomic clock, your bedside alarm clock, or your wristwatch.
It also means time-dependent processes like the vibration of a radio wave,
the beating of your heart, or even the biological clocks that govern the
aging of your body. All measure the same underlying phenomenon—time
itself.
Before exploring further this strange new understanding of time, let’s tie
up a few loose ends. First, be very clear where and how the Principle of
Relativity entered my argument. It entered in the assumption that the speed
of light was c in both reference frames. Again, that assumption follows
directly from the relativity principle as applied to electromagnetism. Had
we not made that assumption, you could have argued that the light in Figure
8.3b was going faster than in Figure 8.3a because it shared the box’s
horizontal motion. Then, even though it was going farther, it would have
taken the same time in both frames. But you can’t make that argument for
the same reason that you can’t argue for different speeds of light for the
observers in Figure 8.1. It’s the Principle of Relativity that rules here, not
some commonsense but incorrect notions you’ve developed about how light
ought to behave. And the relativity principle implies that the speed of light
is the same in all uniformly moving reference frames.
There’s a more subtle point you might want to challenge. In arguing that
the light’s path in Figure 8.3b is longer than in Figure 8.3a, I’ve implicitly
assumed that the box length L is the same in both reference frames. But I’ve
mentioned repeatedly that relativity affects both time and space. So how do
I know the box length isn’t different in a way that shortens the path in
Figure 8.3b and thus voids the time difference? It is true that measures of
space, too, depend on reference frame (we’ll soon see just how) but that
only happens for measures taken along the direction of relative motion
between two frames. There’s no difference for measures, like that of the box
length L in Figure 8.3, that are perpendicular to the direction of relative
motion. You don’t have to take this on faith, because it follows directly
from the relativity principle. However, I’d like to hold that argument until
we’ve explored further our new discovery about the relativity of time.
Fig. 8.4 Two clocks, C1 and C2, are at rest with respect to each other, and the figure is drawn in the
reference frame of these two clocks. A third clock, C, moves relative to C1 and C2 in a direction that
takes it first past C1, then on past C2. Clocks C and C1 read the same time when the two coincide;
this is event A. When C reaches C2 (event B), C reads less elapsed time than C2. Thus, the time
between events A and B is shorter in C’s frame of reference. (Here C’s speed is such that its time is
half that measured by C1 and C2.)
Getting Quantitative
Just how significant is the time difference between events as measured in
different reference frames? That depends on the speed of the frames’
relative motion. For relative speeds small compared with the speed of light,
time dilation is not at all obvious and is very difficult to measure. For
relative speeds approaching c, though, the effect becomes dramatic.
Although I’m taking a nonmathematical approach to relativity in this
book, I nevertheless want to show you quantitatively the effect of time
dilation. One simple formula sums up that effect. If you like math, you can
follow its derivation in the Appendix. Even if you aren’t into math, you’ll
recall that sometime in high school or earlier you learned how to use the
Pythagorean theorem to find the diagonal of a right triangle. That involved
taking the sum of the squares of the two shorter sides, then taking the
square root. What’s this got to do with time dilation? Simply this: the light
path in Figure 8.3b forms the diagonals of two right triangles. As the
Appendix shows rigorously, finding the light travel time in the reference
frame of Figure 8.3b thus involves the Pythagorean theorem. So it’s no
surprise that the formula for time dilation involves squares and square roots:
I’ve labeled this time-dilation formula to make it clear exactly what
each term means. The formula gives the time, t' (read “t prime”), between
two events A and B, as measured in the reference frame in which two
events occur at the same place (e.g., the frame of clock C in Figure 8.4). To
the right of the equal sign, t is the time as measured in a reference frame in
which the two events occur at different places (e.g., the frame of clocks C1
and C2 in Figure 8.4). The symbol v under the square-root stands for the
relative speed of the two reference frames, given as a fraction of the speed
of light. That is, v = 0.5 means half the speed of light, v = 0.9 means 90
percent of the speed of light, and so forth. So what the formula says is that
the time t' in the frame where the events occur at the same place is given by
the time in the other frame, multiplied by the square root of 1 −v2. Let’s
take a closer look at this quantity. If v = 0, then the two frames aren’t in
relative motion and we’re left with the square root of 1, or just 1. So our
formula gives t' = t. Of course: In this case the two frames are really one
and the same reference frame, and the formula is just telling us that
observers in the same reference frame all agree on the time between events.
But if the two frames are in relative motion, then the two times are no
longer equal. Consider first a fairly low relative speed, say 10 percent of the
speed of light, so v = 0.1. Then v2= 0.01, and the quantity √ 1− v2 becomes
the square root of 0.99, which is very nearly 1. So even in this case—a
relative speed of nearly 19,000 miles per second—the two times are still
very close. It’s only when the relative speed becomes a significant fraction
of c that time dilation becomes substantial. With v = 0.8 for the relative
speed between two frames, for example, you can convince yourself that √
1− v2 = 0.6, meaning that the time between two events measured in a frame
where the events occur at the same place is only a little over half what it is
in the other frame. At v = 0.99, the square root is about 0.14, and the one
time is only about a seventh of the other. Finally, you might wonder about
the case v = 1, corresponding to relative motion at the speed of light. Here
the formula gives t' = 0. This suggests that time does not pass at all in the
frame where the two events occur at the same place! But as we’ll see in
Chapter 12, a relative speed of c is not possible for frames of reference
associated with physical objects like human observers or clocks of any
kind.
Wrapping It Up
The crucial point of this chapter is that measures of the time between events
need not be the same in two reference frames in relative motion. In
particular, observers in two different reference frames measure different
time intervals between the same two events—with the shortest time
measured by an observer for whom the two events occur at the same place.
Although I spent a long time elaborating on this point, it follows directly
from the Principle of Relativity and its consequence, the invariance of the
speed of light. If you accept the Principle of Relativity, then you can’t
logically escape this conclusion about the relativity of time.
But is this strange new behavior of time, as embodied quantitatively in
the time-dilation formula, at all relevant to anything? Can we find or
imagine situations in which it occurs and is important? Or is all this just an
academic exercise?
As the numerical examples in the preceding section show, we can
expect obvious time-dilation effects only when relative speeds are very high
—close to the speed of light. We might detect time dilation at lower relative
speeds, but then only with very sensitive experiments. At everyday relative
speeds, including those of jet aircraft and even today’s spaceflight, time
dilation is simply too small for us to notice directly.
Even so, time dilation does occur and is measurable. The effect shows
up dramatically in experiments involving subatomic particles moving,
relative to Earth, at speeds approaching c. As I outlined in Chapter 1, it’s
even been measured in clocks flown around Earth in ordinary aircraft—
although here the effect is minuscule. We can imagine a future with high-
speed space travel, where observers on Earth and passengers in spacecraft
would measure very different times between the same events. We’ll explore
these realities and possibilities in the next chapter.
*Isaac Newton, Principia (Berkeley, CA: University of California Press, 1934); originally published
in London in 1687 and trans. by Andrew Motte in 1729. This phrase appears near the beginning of
the work, in the “Scholium,” a discourse on time, space, and absolute versus relative motion that
follows Newton’s definitions of basic physical quantities.
CHAPTER 9
Does this make sense? Sure: If the ship were going at the speed of light (1
light-year per year) it would take exactly 20 years to make the 20 light-year
journey. It’s going a little slower, so the trip should take a little longer. If all
the clocks read 0 years when the ship starts out, then our answer for t shows
that the Earth and star clocks will read 25 years when the ship reaches the
star.
What about the ship clock? Like clock C in Figure 8.4, this is a clock
for which the two events of interest—the ship’s departure from Earth and its
arrival at the star—occur at the same place. So it’s going to take less time as
measured on the ship clock than it does on the Earth and star clocks—a
conclusion forced on us by the argument of the previous chapter, an
argument based on nothing more than the Principle of Relativity. The time-
dilation formula we developed in the previous chapter applies to just such a
clock as our ship clock, so we can work out the time t' as measured on the
ship:
(You could do the math on a calculator, but I showed all the details because
with v = 0.8 the square root works out so nicely to give 0.6 for the time-
dilation factor.) Figure 9.1 summarizes your star trip and the different times
involved.
The difference here is substantial! In the Earth–star frame, the time
between the ship’s departure and arrival is 25 years. In the ship’s frame, it’s
only 15 years—for a difference of 10 full years. Like everything that’s
come before, this conclusion follows inescapably from the Principle of
Relativity. But what does it mean, besides that the ship’s clock seems to
“run slow” compared with the Earth–star clocks? In particular, what
happens to you as you travel on the ship? Does the trip seem to take just 15
years, or is it “really” a 25-year trip in a spaceship whose clock somehow
isn’t reading right? If you’re thinking the latter, then you’re being
relativistically incorrect! The ship’s frame is just as good as the Earth–star
frame for exploring physical reality. What happens in the ship’s frame is
every bit as real and valid as what happens on Earth. It isn’t just that the
ship’s clock reads 15 years—it’s that, in the ship’s frame of reference, only
15 years of time have elapsed between departure from Earth and arrival at
the star. All manifestations of time reflect this interval. In particular, you
arrive at the star 15 years older than you were when you left Earth—not the
25 years older that your earthbound friends might expect.
Fig. 9.1 At 0.8c, the trip from Earth to star takes 25 years in the Earth–star reference frame but only
15 years in the spaceship’s frame. Times are shown on three digital clocks, one on the ship (shown
twice) and one each on Earth and star. Figure, including clock readings, is a composite of two
different times: first when the ship passes Earthand then when it passes the star.
Would you be aware, while traveling, that you’re aging more slowly
than usual? Would you feel your heartbeat slow down and notice your hair
growing more slowly? Absolutely not! To you, in your spaceship,
everything seems perfectly normal. It must! Why? Because you’re in a
perfectly good frame of reference for exploring physical reality. The laws of
physics work just as well for you as they do on Earth. If you felt something
strange because of the ship’s motion, something you didn’t feel on Earth,
then you could genuinely say that there was something unusual about the
spaceship’s reference frame, something that wasn’t true of the Earth’s frame
—namely, that the ship was moving in some absolute sense while Earth
wasn’t. But, of course, it’s just such specialness of one reference frame over
another that relativity denies. So everything—including the passage of time
in all its manifestations from atoms to clocks to human bodies—seems
perfectly normal onboard the spaceship.
At this point you’re probably full of questions. What if you radio back
to Earth periodically, maybe at each birthday? Will your friends back home
recognize that you’re aging more slowly than they? Suppose some friends
your own age had traveled earlier to the star, via a much slower spaceship.
When you stop at the star and join them, will you really be younger? Just
what is it that’s holding back time on the spaceship? What’s getting into
every clock, into the aging mechanism of each cell in your body, and into
every electron whirling about every atom, to slow them all down?
I’ll answer all these questions eventually, but for now I want to address
just the last one. The answer is that nothing is holding things back. Don’t go
looking for a mechanism that slows everything down. To do so is to say that
there’s really a universal time but that something unusual happens in the
moving spaceship to make all time-consuming processes take longer. That’s
no good because, again, it singles out Earth’s frame as the one in which this
time-slowing mechanism isn’t at work, and it identifies the ship’s frame as
one in which something unusual happens because the ship is moving. To
say those things about Earth and spaceship is to violate the relativity
principle.
There’s one subtle point I need to make. We’ve applied the idea of time
dilation—that the time between the same two events is different when
measured in reference frames in relative motion—to the duration of your
star trip. We can do that if both reference frames, namely the Earth–star
frame and the ship frame, are indeed in uniform motion. That’s because the
Principle of Relativity says that the laws of physics are equally valid in all
reference frames in uniform motion. If the ship starts from rest relative to
Earth, then travels to the star and slams on its brakes, it’s certainly not
moving uniformly while it starts and stops. The reason I’ve assumed the
ship can accelerate instantaneously to 0.8c relative to Earth, and then stop
abruptly at the star, is so that it spends essentially the entire journey in
uniform motion. To be completely correct, we should identify the departure
event as occurring just after the ship has accelerated. Since the acceleration
takes place instantaneously, the ship is still right at Earth but it’s now
moving relative to Earth. Its clock still reads 0 because the acceleration took
essentially no time. The ship then moves, uniformly, relative to Earth and
star, until the arrival event, when it’s at the star but hasn’t yet applied its
brakes. Between the departure and arrival events so defined, the ship indeed
moves uniformly, and we can apply the Principle of Relativity and its
logical consequence, the time-dilation formula, to reach our conclusion that
the time between departure and arrival is 10 years less for those riding the
ship than for those remaining on Earth or at the star.
A cleaner way to present the star trip would have been to imagine an
alien being in a spacecraft that happens to come whizzing by Earth at a
steady 0.8c. Later the alien passes a star that, as measured in the Earth–star
frame, is 20 light-years from Earth. We can draw exactly the same
conclusion in this case as we did before, namely, that the alien measures 15
years between passing Earth and passing the star, while observers with
synchronized clocks at Earth and star report a 25-year interval between the
time those on Earth see the craft go by and when observers at the star see it
pass. What’s simpler about this situation is that we don’t have to think about
the alien spacecraft starting and stopping; it’s truly in uniform motion
relative to Earth the whole time, so it’s completely obvious that the
relativity principle applies. But I emphasize again that it also applies to your
star trip, at least during the interval when you’re moving uniformly relative
to Earth. Later in this chapter we’ll reconsider the effects of starting and
stopping.
Squeezing Space
How can the spaceship get from Earth to star in only 15 years? After all, the
distance is 20 light-years, so even light should take 20 years. As you’re
probably aware, and as I’ll elaborate on in Chapter 12, no material object
can go faster than light. So how can the spaceship get to the star in only 15
years?
The answer lies in the fact that measures of space, as well as of time, are
different in different reference frames. Let’s look at things from your
viewpoint as you ride the spaceship. You know you’re moving at 0.8c
relative to Earth and star, and you know that your journey takes 15 years.
So how far have you gone? You’re in a perfectly good frame of reference
for doing physics, so the elementary-school formula distance = speed
× time works just fine for you. So to you, the distance you’ve traveled from
Earth to star must be
d' = (0.8 light-years/year) × (15 years) = 12 light-years.
(I’ve called this distance d' [“d prime”] for consistency with my earlier
notation, where I used t' [“t prime”] to represent time measured in the ship
frame.)
There’s no inconsistency here and no faster-than-light travel. To you in
the spaceship, the Earth–star distance is 12 light-years, and at 0.8 light-
years per year, it makes perfect sense that your trip takes a little over 12
years—15 years, to be precise. What’s troubling is that something you
thought was objectively real and absolute, namely the distance between two
objects, simply isn’t. Like measures of time, measures of space also depend
on one’s frame of reference.
Note, by the way, that the 12-light-year distance in the ship frame is
precisely 60 percent of the 20-light-year value that the Earth–star distance
has in the Earth–star frame. That 60 percent, or 0.6, is just the relativistic
factor that we calculated for time dilation with v = 0.8c. In fact, that’s
generally true. If the distance between two objects is d in a frame of
reference where the two are at rest, then in a reference frame moving at
speed v relative to the objects, the distance will be contracted by this same
factor, giving
d' = d × .
For this length contraction to occur, the relative motion must be along a line
between the two objects. If the motion is perpendicular to that line, then
there’s no contraction; if it’s at an angle, then the contraction factor is
somewhere between.
I referred to the shrinking of the distance between Earth and star, or any
to other objects at rest with respect to each other, as length contraction.
That’s because the two objects could be the opposite ends of a single
physical thing, like a ruler. To someone moving relative to the ruler, it’s
contracted to a shorter length. The ruler may still say “12 inches,” but as
measured by an observer moving relative to the ruler, it won’t measure 12
inches. But isn’t the ruler really 12 inches long? To ask that is to suppose
there’s one special frame in which measures of length or distance are
correct and that therefore they’re wrong in other frames. Obviously, that
view violates the relativity principle. On the other hand, one’s relationship
with an object like a ruler is certainly simplest in a frame of reference at rest
with respect to the object. For that reason the length measured when one is
at rest with respect to an object is called the object’s proper length. Here
“proper” doesn’t mean “correct” as much as it does “proprietary”—in the
sense of “belonging” to the object, or intrinsic to the object in its own frame
of reference. But that doesn’t mean the object’s proper length has any
transcending reality. Observers in different reference frames will measure
different lengths for an object, and they’ll all be correct. Measures of space
and time just aren’t absolute. An object is longest in a reference frame
where it’s at rest and shorter in any other frame. Just how short depends on
one’s speed relative to the object, as given by the formula above.
Once again I need to remind you of the difference between observing
and seeing. When you observe an object moving relative to you, you
measure its length by, perhaps, noting where its front and back ends are in
relation to a ruler or meter stick at rest in your reference frame. Since the
object is moving, this can be a bit tricky. You have to be careful to note the
positions of the front and back ends at the same time. But you can do that,
and the result will be the contracted length of the object as measured in
your reference frame. However, that doesn’t mean you’ll see the object
contracted. That’s because light from different parts of the object reaches
you at different times, and those time differences are significant for an
object moving fast enough relative to you that its length is noticeably
contracted. Remarkably, the object appears not contracted but rotated!
However, I’m not going into the details of how this comes about because
it’s a bit of a distraction from the bigger theme of the nature of space and
time.
Finally, a historical note. The length-contraction formula d' = d ×
is precisely the remedy Lorentz and Fitzgerald proposed in the late 1800s to
resolve the quandary of the Michelson–Morley experiment. If the
Michelson–Morley apparatus contracted by this amount in the direction of
its motion through the ether, Lorentz and Fitzgerald correctly argued, then
the travel times for light along the two legs of the apparatus would remain
the same, and the experiment wouldn’t be able to detect Earth’s motion
through the ether. So Lorentz and Fitzgerald got it partly right, in that they
correctly predicted a motion-induced contraction of material objects. But
they remained philosophically mired in a relativistically incorrect way of
thinking, because for them the contraction occurred against a background of
absolute space and time. Theirs was a contraction of material objects in an
uncontracted space. The relativistically correct interpretation of length
contraction is that measures of space itself differ in different reference
frames and that differing measures for the length of material objects reflect
this underlying relativity of space.
It Really Happens!
Does all this really happen? Does time really pass more slowly on your
spaceship than it does on Earth and star? Is the Earth–star distance really
less for you as you ride the spaceship? The answer to these questions is yes
—although if you’re in a deeply relativistic mode of thinking then these
questions themselves may be bothersome. Why, after all, can’t you on the
spaceship consider Earth to be moving, and therefore its clock to “run
slow”? You can, and I’ll return to this point later. For now, though, let’s
consider how we might show that it really happens. Unfortunately, we don’t
yet have the technology to carry out the star trip, so we can’t directly
experience time dilation and length contraction. We do have access to
objects that move relative to us at very high speeds but those objects are too
small for us to ride on, because they’re subatomic particles. Surprisingly,
though, they carry “clocks” that allow us to confirm quite dramatically the
effects of time dilation.
Here I’m going to describe an experiment that was done in the 1960s
with the express purpose of showing directly that time dilation and length
contraction do occur.* Special relativity had been thoroughly verified long
before the 1960s, so this experiment was designed and performed especially
to be convincing to folks other than scientists. I’m going to introduce the
real experiment first by imagining a fictitious but analogous experiment
closely related to our Earth–star trip.
Imagine I have some unusual clocks with the remarkable property that
they self-destruct in 20 years, exploding into a jumble of hands and gears.
These aren’t very accurate clocks, so some explode after only 18 actual
years, a very few as early as 12 or 15 years, while a few last 25 or 30 years.
If I take 100 new clocks and wait 20 years, roughly half of them will be
gone. If I wait 25 years, only a very few will be left. But if I wait only 15
years, nearly all of them will still be around.
Let’s load your spaceship with 100 of these clocks, and send you off on
that star trip at 0.8c. Your friends remain on Earth, although a few of them
have already traveled—again, on a much slower spaceship—and are
waiting at the star. When the ship reaches the star, what will you or your
star-based friends find? Will the spaceship be full of clocks or will most of
them have self-destructed before the journey is over? Your Earth–star
friends, who agree that your trip takes 25 years, might expect to find very
few clocks. But for you the trip takes only 15 years, so nearly all the clocks
should still be around. So which is it? This is something we can test
objectively by opening up the ship when it reaches the star and looking for
the clocks. If we don’t find many, then we can conclude that the trip time
was really 25 years, even for the traveling clocks. If we do find a lot of
clocks remaining, then we’re forced to conclude that the trip time as judged
by these clocks was considerably less than the 20 years it takes before half
the clocks will explode.
If we did this experiment, which of course is no more possible than the
original star trip with a single accurate clock onboard the ship, then
relativity says that we should indeed find lots of clocks, because in the ship
frame the trip time is only 15 years.
In fact, an almost identical experiment is possible, and it’s the one that
was done in the 1960s. Instead of clocks, the experiment uses subatomic
particles that are radioactive—meaning that they self-destruct by exploding
into several other particles. The particular particles are called muons, and
they’re formed when cosmic rays slam into atoms high in Earth’s
atmosphere. A steady rain of these muons comes downward through the
atmosphere at speeds approaching c. The experiment consists of two parts.
First, experimenters high on a mountaintop (Mount Washington, New
Hampshire) use a muon detector to determine the intensity of the “muon
rain” (Figure 9.2a). The detector is designed to catch only muons moving
with a narrow range of speeds very close to 0.995c. The experimenters
detect just over 500 such muons each hour at the mountaintop. Now the
time it takes muons to self-destruct is well known, although the 1960s
experimenters actually measure it again as part of their experiment. That
self-destruct time (analogous to the 20 years for our self-destructing clocks)
is short enough that even muons moving at 0.995c would seem very
unlikely to make it from the altitude of Mount Washington down to sea
level without self-destructing on the way.
Fig. 9.2 Muons are radioactive particles that act like clocks. They arrive in great numbers at the top
of Mount Washington. (a) They self-destruct at such a rate that very few would survive to reach sea
level, if time in the muons’ reference frame were the same as on Earth. (b) The muons experience
time dilation, and thus most of them reach sea level. Detection of these many muons confirms time
dilation
Now, that 500 muons per hour figure isn’t unique to Mount Washington;
we would measure very nearly the same quantity each hour anywhere
within a few hundred miles if we were at the same 6,300-foot altitude as
Mount Washington. So here’s what the experimenters do: They go down to
Cambridge, Massachusetts, at sea level, and set up the same muon detector,
again looking for muons moving at 0.995c. They know that overhead, some
6,300 feet up, there are 500 such muons every hour coming into a region
the size of their detector. If they didn’t believe in relativity, then the
experimenters should expect to find very few muons at sea level, as shown
in Figure 9.2a.
Now let’s look at things from the muons’ point of view. They’re like the
clocks in our spaceship, which are present both as the ship passes Earth and
again when it reaches the star. The muons are present as they pass the
6,300-foot altitude of Mount Washington, and—if they haven’t self-
destructed—they’re present again at sea level. So they’re clocks that are
present at two events, and for them the time between those events should be
shortened according to our time-dilation formula. For a speed of 0.995c,
that formula gives a relativistic factor equal to one-ninth. So the time
to get from Mount Washington to sea level, as judged by the muons, should
be only one-ninth what an observer at rest on Earth would expect. That
muon time is so much shorter that very few muons should decay on the way
from 6,300 feet to sea level.
We have a clear distinction between two possible outcomes. If relativity
is correct, and time dilation occurs, then very few muons will decay and the
experimenters should catch many muons in their detector at sea level
(Figure 9.2b). If relativity isn’t correct, then most of the muons should
decay on the way down, and the sea-level experiment should find very few
(Figure 9.2a). So what happens? The result is clear: Nearly as many muons
reach sea level each hour as arrive at the top of Mount Washington. The
muons are clearly experiencing time dilation, which at their speed relative
to Earth is a dramatic effect. A look at the actual numbers in relation to the
muons’ lifetime not only confirms that time dilation occurs but also that it
obeys precisely our time-dilation formula. Time for the muons elapses at
one-ninth the rate it does as measured by two clocks at rest on Earth, one on
the mountaintop and one at sea level.
We can also ask the same question here that we did for the spaceship.
How can the muons possibly get from Mount Washington’s altitude to sea
level in so short a time that they don’t self-destruct on the way? Doesn’t
that require them to be moving faster than light? No! Again, length
contraction comes to the rescue. Just as measures of time are different in the
muons’ reference frame, so are measures of space. For the muons, the
height of Mount Washington is contracted by the same factor (one-ninth),
so the 6,300-foot mountain is, to the muons, only 700 feet high.
You may still be dissatisfied, because subatomic muons are so outside
the range of your everyday experience. But the muon experiment really is
completely analogous to our star trip, and we have every reason to believe
that what holds for the muons would hold for ordinary clocks, people, and
spaceships. If that weren’t the case, then we would have the Principle of
Relativity applying in the subatomic realm of physics but not to the physics
of ordinary objects. Think back to Chapter 2, and you should find that
possibility very unsettling.
Still, it would be nice if we could measure time dilation in ordinary
clocks big enough for us to see and hold. Well, we can—but the effect is
less dramatic because we can’t get regular-sized clocks to speeds that,
relative to us, are anywhere near the speed of light. Furthermore, because
the effect of time dilation is so small at the relative speeds we can achieve,
we need clocks that are good enough to mark extremely small time
differences. We have such clocks—they’re atomic clocks like those used to
set the world time standards. These clocks are accurate to a few billionths
of a second. In an experiment aimed at confirming relativistic effects on
time, atomic clocks that initially read the same time ended up disagreeing
by several hundred billionths of a second as a result of relative motion at
the speeds of commercial aircraft. We’ll revisit and elaborate on this
experiment at the end of the next section.
At this point you may still be a bit uncomfortable with relativity and its
strange consequences like different-aged twins and time travel to the future.
But I hope you see the logic that takes us directly from the Principle of
Relativity to these unusual implications. The relativity principle itself
should be firmly grounded in your mind; after all, it made complete sense in
Chapter 2, and by Chapter 7 it had emerged from the quandaries of late-
nineteenth-century physics as the one clear way to resolve the muddle of
motion and ether.
A Test of Faith
Now I’m going to put your faith in relativity to the test. Consider once
again a one-way trip to that distant star. To make things really simple, forget
about starting at Earth and stopping at the star. Rather, the spaceship just
comes zooming past Earth at a steady 0.8c, then later zooms past the star. In
other words, the ship is always in uniform motion, so it’s always a good
reference frame in which to do physics.
The preceding chapter should have convinced you that, from the
viewpoint of observers on Earth, time on the spaceship “runs slow.”
Specifically, the time between the event of the ship passing Earth and the
later event of the ship passing the star is shorter in the ship’s frame of
reference than it is in the Earth–star frame.
Now here’s my question: What do you, a passenger on the spaceship,
have to say about clocks in the Earth–star frame? To a friend on Earth, your
clock runs slow. Do you want to say, then, that from your shipbound
perspective, Earth and star clocks must run fast? Sure sounds like a logical
answer. After all, if your clock is running slow compared with clocks on
Earth and star, then surely Earth and star clocks must be running fast
compared with yours.
But wait! I’ve set things up in this example so that you in the spaceship
and your friend on Earth really are in equivalent situations, namely, the
state of uniform motion. Your friend says, “I’m at rest on Earth, and the
spaceship is moving relative to me, so the ship’s clock is running slow
compared with my clocks.” But your situation is perfectly equivalent, so
you must be able to say the analogous thing: “I’m at rest in the spaceship,
and Earth is moving relative to me, so Earth’s clocks are running slow
compared with mine.” Each observer—you in the spaceship and your friend
on Earth—must claim that the other’s clocks are running slow. How can
that possibly be? Yet if you really accept relativity—that the laws governing
physical reality are the same in all uniformly moving reference frames—
then you’re forced to this seemingly absurd conclusion.
You might still object that the spaceship is “really moving” while Earth
isn’t, so the situations aren’t really equivalent. But by now you’re surely
past such relativistically incorrect thinking! Unlike the round-trip we
considered in the preceding chapter, or even a one-way trip with starts and
stops, the scenario here really has both ship and Earth in completely
equivalent situations. Neither changes its motion in any way, so both are in
the state of unchanging, uniform motion that makes them both suitable
reference frames for applying the laws of physics. Or maybe you want to
say that the ship is moving in one direction relative to Earth, and Earth is
moving in the opposite direction relative to the ship, so that makes their
situations different. But nothing in our arguments leading to time dilation
depended on the direction of the relative motion. I could have redrawn
Figure 8.3b in a reference frame in which the light box was moving to the
left, and the argument from that figure would not have changed one bit. So
the direction of relative motion can’t matter.
So which is it? Do Earth’s clocks run slow or fast from the viewpoint of
the ship’s clock? The answer, if we accept relativity, has to be that they run
slow. To understand this answer, we have to find our way out of the
contradiction it seems to imply—the contradiction that, relative to Earth,
the ship clocks run slow while at the same time, relative to the ship, the
Earth and star clocks run slow. Escaping this contradiction will give you
further insights into the nature of time and especially the status of
simultaneous events, that is, pairs of events that occur at the same time.
Simultaneity Is Relative!
What does it mean to say that two events occur at the same time? If the two
events also occur at the same place, then there’s no question. If we see the
events simultaneously, then they must have occurred simultaneously. But
suppose the events occur at different places. Then an observer watching the
two events has to compensate for the time it takes light to get from each
event to determine if they were, in fact, simultaneous. You’ve already seen
enough of relativity to know that the constancy of the speed of light might
make that compensation different for observers in relative motion. So if one
observer determines that two events are simultaneous, it becomes an open
question whether another observer, moving relative to the first, will deem
them simultaneous.
Another approach to determining simultaneity is to set up clocks at the
locations of the two events. At each clock, station a reporter who will report
to you the time of the event occurring at the location of that clock. If both
times are the same, then you can claim that the events were simultaneous.
For this to work, you have to be sure those clocks are synchronized. What
does that mean? It means both clocks read the same time, say noon, at
exactly the same instant. In other words, the event of one clock’s hands
pointing to noon is simultaneous with the other clock’s hands pointing to
noon. You could synchronize the clocks by standing midway between them
and sending a light flash to your reporters at the clocks. Each reporter sets
the clock at the instant the light flash arrives. But again we’re back to
judging simultaneous events by methods that involve sending light signals
through space and, again, relativity leaves open the question of whether
observers moving relative to you will agree that the events are
simultaneous. In this case, that means such observers might not agree with
you that your clocks are synchronized.
In fact, events that are simultaneous in one frame of reference may not
be simultaneous in another reference frame. I’m now going to demonstrate
this rigorously in a way that follows from the Principle of Relativity. To do
so, I’ll invoke relativistic length contraction—a result which, as I showed in
the preceding chapter, follows logically from the relativity principle. Recall
that an object is longest in a reference frame in which it’s at rest and shorter
as measured in any other reference frame. In our earlier examples, one
“object” in question was the Earth–star pair, whose separation contracted
from 20 light-years in the Earth–star frame to 12 light-years in the ship
frame. Another object we considered was Mount Washington, 6,300 feet
high in the Earth frame but contracted to only 700 feet in the reference
frame of muons for which Mount Washington was moving at 0.995c.
I’m now going to consider two distinct objects and look at how length
contraction applies to each of them in different reference frames. The
objects are two identical airplanes, and they’re flying toward each other at a
substantial fraction of the speed of light (these are no ordinary airplanes!).
The airplanes pass, one just above the other. I want to consider two events
associated with this passing. The first event will be the nose of the upper
airplane passing the tail of the lower one. We’ll call this event A. The
second event will be the tail of the upper airplane passing the nose of the
lower one. This is event B. We want to know whether events A and B are
simultaneous.
Figure 10.1 shows the situation as observed in a reference frame in
which the two planes approach with the same speed. Obviously, neither
plane is at rest in this reference frame, so each is shorter, because of length
contraction, than it would be in a reference frame where it was at rest.
Because both are moving at the same speed relative to the reference frame
of Figure 10.1, each is contracted by the same amount. So in this reference
frame the two planes have the same length. When they’re alongside each
other, as in Figure 10.1b, the nose of the upper plane coincides with the tail
of the lower one at the same time that the tail of the upper plane coincides
with the nose of the lower one. That is, events A and B are simultaneous in
the reference frame of Figure 10.1.
Fig. 10.1 (a) Two identical airplanes approach. The figure is shown in a reference frame in which the
planes approach with the same speed. Event A is the nose of the upper plane passing the tail of the
lower plane; event B is the tail of the upper plane passing the nose of the lower one. (b) In this
reference frame, events A and B are clearly simultaneous.
Finally, Figure 10.3 shows a reference frame in which the lower plane is
at rest. Now the lower plane is longer, the upper one shorter, and it’s
obvious that in this reference frame event B occurs before event A.
The example of the airplanes shows clearly that events that are
simultaneous in one reference frame may not be simultaneous in another
reference frame moving relative to the first. Comparison of Figures 10.2 and
10.3 shows something even harder to swallow: the order of two events can
be different in different reference frames. That might seem particularly
disturbing if you think about events that are causally related—like your
birth and your now reading this book. Are there really observers for whom
the reading comes before the birth? Fortunately, not. Causality is a big
enough issue that I’ll devote the entire next chapter to it; for now, let me
just assure you that not all pairs of events can have their time order
reversed. All we’ve learned from the airplane example is that those events
that are simultaneous in some reference frames may not be simultaneous in
other frames, and it’s precisely such events whose time order itself depends
on the observer’s frame of reference.
Fig. 10.3 Like Figure 10.2, but now in the reference frame in which the lower plane is at rest. Here
event B comes first. Incidentally, Figures 10.1, 10.2, and 10.3 are scaled so that the speed of the two
planes relative to the reference frame of Figure 10.1 is 0.6c.
By the way, I’ve been saying that events simultaneous in one frame
“may not be simultaneous” in another frame. Why “may not” as opposed to
“are not”? Are there cases where events can be simultaneous in different
frames? Yes, if the relative motion of the two frames is perpendicular to the
line joining the two events. For example, observers moving from the bottom
toward the top of the page in Figure 10.1 will judge events A and B to be
simultaneous, no matter how fast they’re moving relative to the frame of
Figure 10.1. For them, contraction of the airplanes squeezes them in the
vertical direction, not the horizontal, so it affects both planes the same way.
Fig. 10.4 Star trip of Figure 9.1, shown at two different times: (a) the situation at the instant the ship
passes Earth; (b) its arrival at the star. Figure is drawn from the Earth–star frame of reference.
Fig. 10.5 Star trip, now drawn from the ship’s reference frame. Here the ship is at rest, while the
Earth and star move to the left at 0.8c. Note that the ship is longer in its own frame, while Earth, star,
and the distance between them are contracted. (a) As the ship passes Earth, both ship clock and Earth
clock read 0. In the ship frame the Earth and star clocks aren’t synchronized, and the star clock is, in
fact, 16 years ahead. (b) The ship clock advances 15 years between the times Earth and star pass the
ship, just as in Figure 10.4, but Earth and star clocks are “running slow,” and they advance only 9
years. Because it was 16 years ahead, the star clock still reads 25 years as the star passes the ship, just
as in Figure 10.4.
Let’s sort all this out. Turn back to Figure 10.4, where (a) shows the
situation in the Earth–star frame when the ship passes Earth and (b) shows
the situation in this frame when the ship passes the star. In this frame all the
clocks read 0 as the ship passes Earth (Figure 10.4a). Later, as the ship
passes the star, its clock reads 15 years while Earth and star clocks—which
are synchronized in the Earth–star frame—both read 25 years (Figure
10.4b). The interpretation of the clock readings in the Earth–star frame is
that 25 years have elapsed in that frame but only 15 years have elapsed in
the ship frame because the ship clock runs slow.
Now look again at Figure 10.5, where (a) shows the situation in the ship
frame when Earth passes the ship and (b) shows the situation in this frame
when the star passes the ship. As in the Earth–star frame, both ship clock
and Earth clock read 0 as Earth and ship pass. But in the ship frame, the
clocks in the Earth–star frame aren’t synchronized; in fact, the star clock
already reads 16 years when Earth and ship pass. Later, as the star and ship
pass, the ship clock has advanced from 0 to 15 years. The star clock,
running slow from the viewpoint of the ship frame, has advanced only 9
years. But since it was ahead to begin with, it now reads 25 years. The
interpretation of the clock readings in the ship frame is that 15 years have
elapsed in that frame while only 9 years have elapsed in the Earth–star
frame.
Yet observers in both frames agree about what the clocks actually read
whenever two clocks are right next to each other so observers can
unambiguously compare times. In particular, observers in both frames agree
that 15 years elapse on the ship’s one clock. They also agree that the
reading of the Earth clock at the Earth/ship passing differs by 25 years from
the reading of the star clock at the star/ship passing. What they disagree
about is the interpretation of this 25-year interval. To observers in the
Earth–star frame the clocks at Earth and star are synchronized, so 25 years
is a legitimately measured time between two events. To observers on the
ship, Earth–star clocks are running slow, and advance only 9 years between
the two events. The 25-year difference in clock readings occurs because the
star clock is ahead of the Earth clock by 16 years.
Which interpretation is right? By now you should have enough faith in
relativity to know that’s a meaningless question. Observers in both frames
are correct in asserting that the other’s clocks run slow, and they’re saved
from outright contradiction by the fact that events simultaneous in one
reference frame aren’t simultaneous in another frame in motion relative to
the first.
It’s easy to conjure up situations in relativity that seem contradictory,
and books on relativity are often full of paradoxes that seem to embody
such contradictions. But there’s always a way out, and it usually involves
recognizing that simultaneity is relative. Careful consideration of the timing
of events in two different reference frames almost always resolves any
apparent contradiction.
CHAPTER 11
That simultaneity is relative has just got us out of the seeming contradiction
of observers in relative motion each finding that the other’s clocks “run
slow.” But in demonstrating the relativity of simultaneity, I introduced what
may seem an even more disturbing thought—that the time order of events
may depend on one’s frame of reference. Because relativity gives every
uniformly moving reference frame equal status, this reversal of time order
isn’t just some illusion. It’s really true that I can observe event A to occur
before B, that you can observe B before A, and that we’re both right. But
how can that be? Doesn’t it wreak havoc with causality?
Excitement on Mars
When NASA’s Mars Rover was exploring the Martian surface in 1997,
Mars was 10 light-minutes from Earth. That means it took light, as well as
the radio waves used to communicate with Rover, 10 minutes to get from
Earth to Mars or from Mars to Earth. You’ve probably heard that the speed
of light is the maximum possible speed in the Universe—a consequence of
relativity that I’ll explore in detail in the next chapter. For now, let’s accept
this cosmic speed limit which, more accurately, states that information
cannot be communicated at speeds faster than c. So that 10-minute travel
time for light between Earth and Mars is the shortest time that any
information could take to travel between the two planets.
Now, let’s put ourselves in NASA’s Mars Rover control room, during
the time of the Rover mission. To make things concrete, let’s say we’re just
beginning a coffee break. Suppose further that, 5 minutes before our break
began, Rover’s TV camera spotted a real, live Martian strolling across its
field of view. Is that event in our past? It happened 5 minutes ago, so it
occurred before our present event. But could it influence us right now? Can
we possibly know, right now, of Rover’s discovery? No, and we won’t be
able to know for another 5 minutes, when we receive Rover’s TV picture.
So Rover’s discovery of extraterrestrial life cannot possibly influence us,
right now at the start of our coffee break here in NASA’s control room. We,
the Rover scientists, aren’t excited by what’s happened, because we simply
can’t know about it. By our “influence” definition, then, Rover’s discovery
is not in our past. It cannot be a cause of the event, namely our coffee break,
that’s occurring right now in the earthbound Rover control room.
We have in Rover’s discovery on Mars and our earthly coffee break 5
minutes later a pair of events that just cannot be causally related, because
no information can travel between them. They’re too far apart in space and
too close in time for even light to be present at both events. It’s precisely
such pairs of causally unrelated events that different observers, moving in
this case relative to our Solar System, may see with different time orders.
For example, there could be an observer for whom the two events are
simultaneous and others for whom the coffee break occurs first. Still other
observers would agree with us on Earth that the Martian discovery occurred
first, but they wouldn’t agree that the time interval was 5 minutes. I won’t
go into the mathematics of all this. But imagine in Figure 10.5 replacing
Earth with Mars and the star with Earth. Then you can see that an observer
moving in the direction from Mars toward Earth will observe Earth time
advanced relative to Mars time. That advancement will be 5 minutes—
meaning the two events will be simultaneous—if the observer happens to be
moving at half the speed of light. Observers moving faster in the same
direction will judge the coffee break on Earth to occur before Rover’s
discovery of Martian life. Observers moving slower but in the same
direction will judge the Mars event to occur first, but by less than 5
minutes.
Of course all these observers are correct, because they’re all in
uniformly moving reference frames that are equally valid situations for
applying the laws of physics. As you well know by now, measures of time
simply aren’t absolute, but depend on one’s reference frame. What’s new
here is that the time order of events may also depend on reference frame.
But there’s no contradiction. That’s because the only events for which
different observers claim different time orders are those that can’t be
causally related, because they’re too far apart in space and too close in time
for any influence traveling at c or less to get between them. Too far apart by
whose measures? By anyone’s. Even though the different observers
disagree about the time between the events, and for that matter about the
distance between Earth and Mars, all agree that the distance is greater than
the distance light could travel in the time between the events. More on this
in Chapter 13.
Poor Mars Rover! It’s made a great discovery, but, alas, its navigation
system has failed and it’s heading straight toward a deep crater. We, in the
NASA control room, realize this just as we start our coffee break, and we
calculate that in 5 minutes Rover will topple into the crater and be
destroyed. Is Rover’s demise in our future? Can we send a radio signal to
stop Rover and prevent the impending disaster? No; Rover is 10 light-
minutes away, and we have only 5 minutes until Rover reaches the crater.
Nothing we do at our present moment can influence what happens on Mars
5 minutes from now. So, according to our influence definition of the future,
the event of Rover falling into the crater is not in our future. Here’s another
pair of events—the start of our coffee break and Rover’s unfortunate
accident—that cannot be causally related. And, sure enough, there could be
other observers, moving relative to Earth and Mars, who find these events
simultaneous, and still others for whom Rover’s demise occurs before our
coffee break starts.
The Elsewhere
If the events of Rover’s alien discovery and its crater accident aren’t in our
past or our future, where are they? Relativity opens a new realm of time—
or, more precisely, of spacetime. Events that aren’t in the past or the future
of a given event are in its elsewhere. They’re events that cannot
communicate with the given event, so the two cannot be causally related—
and different observers can disagree about their time order. On Mars, for
example, any events occurring between 10 minutes before and 10 minutes
after the start of our NASA coffee break are in the elsewhere of the coffee-
break event. Similarly, events in the 2-million-light-year distant Andromeda
galaxy are in our elsewhere if they occurred anytime more recently than 2
million years ago or will occur within the next 2 million years. Even events
that occur on the other side of the room you’re sitting in are in your
elsewhere, provided they occurred less than a few nanoseconds (billionths
of a second) ago or will occur less than a few nanoseconds from now.
Clearly, the farther away a place is, the broader the range of time that lies in
the elsewhere of your here and now.
You can picture the elsewhere with the help of Figure 11.1. Here I
assume for simplicity that we live in a space with just one dimension—
meaning we can move back and forth along a line but not in any other
direction (it gets too hard or even impossible to draw with more
dimensions!). I’ll represent place—that is, where an event occurs—by its
position along a horizontal line. I’ll represent time—when an event occurs
—by its position along a vertical line. To make things simple, I’ll measure
position in light-years and time in years. Every point on the diagram
represents a specific place and a specific time so the diagram is therefore a
spacetime diagram. But time and place are what specifies an event, so each
point on the spacetime diagram marks an event. The point at the center,
where the time and space axes cross, is the here (position 0) and now (time
0).
Fig. 11.1 Spacetime diagrams, showing only one spatial dimension. Each point in the diagram
represents an event, its location in space given by its horizontal position, time by its vertical position.
(a) In the Newtonian view, the present is the horizontal line separating future from past. All events
happening right now, no matter where they are, occur in the present. (b) In relativity, future and past
have different meanings for each event. Diagram shows the past, future, and elsewhere for event A.
Tick marks are at intervals of 1 light-year in space and 1 year in time. The past comprises all those
events that can influence A; the future those that A can influence. The text discusses the relations
among the six events A–F.
First Principles
First I’m going to give a quick and easy answer, one that follows directly
from the Principle of Relativity. It won’t be particularly satisfying, but it
will be correct. In the sense that all the implications of relativity follow
from the principle, it will be as complete as can be. Later, though, I’ll
provide answers you may find more satisfying.
The relativity principle states that the laws of physics are the same in
any uniformly moving frame of reference. Among those laws are Maxwell’s
equations of electromagnetism, and among the predictions of Maxwell’s
equations is the existence of light waves going at speed c. That prediction
must hold in all reference frames, meaning that all observers must measure
the same speed c for light. I’ve used that argument many times before, so it
should be tediously familiar.
Now suppose you and I are standing together as a light beam goes by.
You hop into a fast rocket and try to catch up with the light. If you succeed,
then you’ll be moving with speed c relative to me, and you’ll be at rest with
respect to the light. If you were at rest with respect to the light, then you
would be measuring a speed for light, namely zero, which is not equal to c.
But light is supposed to have speed c with respect to any uniformly moving
reference frame. If it didn’t, the Principle of Relativity would be violated.
So your situation—moving with speed c relative to me and therefore being
at rest with respect to the light—must be impossible.
This question of catching up with light puzzled Einstein from the age of
16 until he resolved it with his special theory of relativity. He imagined
running alongside a light wave, so the wave would be at rest with respect to
him. Einstein understood Maxwell’s electromagnetism, and he knew that
Maxwell’s equations dictated a particular form for an electromagnetic
wave, a form in which electric and magnetic fields are perpendicular to
each other and to the direction of the wave’s motion. What bothered
Einstein was that a stationary structure of that form was not a valid solution
of Maxwell’s equations. In fact, the only valid solution had the
electromagnetic wave moving at speed c. So there had to be something
wrong with the idea of running alongside a light wave so that the wave
would appear to stand still.
Before relativity, there was no real problem here. Maxwell’s equations
were thought to be valid in only one particular frame of reference, namely
the frame at rest with respect to the ether. You wouldn’t expect
electromagnetic waves as seen from other frames to satisfy Maxwell’s
equations, and there would be no problem with an observer moving through
the ether in such a way that an electromagnetic wave appeared at rest.
But take away the ether, as Einstein did, and require that the laws of
physics be valid for all uniformly moving observers. In other words, insist
on the Principle of Relativity. Then all observers must measure c for the
speed of electromagnetic waves, and no observer can be at rest relative to
such a wave. Thus the Principle of Relativity asserts that it’s impossible for
you to run alongside a light wave and see it at rest, so it’s impossible for
you to move relative to me at speed c.
This argument is logically airtight and follows directly from the
relativity principle. But it isn’t very satisfying, because it doesn’t explain
why you can’t get yourself up to the speed of light. What actually prevents
that? I’ll now explore several more satisfying answers to that question.
However, there’s no new physical principle involved; like everything else in
relativity, these answers, too, are ultimately grounded in the Principle of
Relativity.
Leapfrogging to c
Here’s an obvious way to beat the cosmic speed limit, c. Suppose we have a
big rocket capable of going, relative to Earth, at three-quarters of the speed
of light (0.75c). Inside the big rocket, we build a miniature version with the
same technology. We fire up the big rocket and zoom away from Earth at
0.75c. You climb into the small rocket, fire the engine, and soon you’re
moving at 0.75c relative to the big rocket (Figure 12.1). So now you must
be moving at 1.5c relative to Earth, right?
Fig. 12.1 The big rocket is moving past Earth at 0.75c. A smaller rocket moves at 0.75c relative to
the big one. So is the smaller rocket going at 1.5c relative to Earth? Common sense suggests it is, but
in fact the speed of the small rocket relative to Earth is only 0.96c.
Wrong! Why wrong? Because measures of space and time aren’t the
same in different frames of reference. The small rocket is indeed moving at
three-quarters of light speed relative to the big rocket, and the big rocket is
moving at three-quarters of light speed relative to Earth. But determining
the speed of the small rocket relative to the big one involves measurements
of distance and time in the reference frame of the big rocket. An observer
on Earth doesn’t agree with those measures and comes up with a different
result for the speed of the small rocket. I won’t go through all the math, but
that speed is, in fact, 0.96c.
What I’m saying here is that 0.75c and 0.75c don’t add to give 1.5c.
That’s an odd statement! Surely if I jog at 5 miles an hour down the aisle of
an airplane going 600 miles per hour relative to the ground, then I’m going
at 605 miles an hour relative to the ground. Why isn’t it the same for the
rockets? Actually, it is the same in both cases, but it’s the example of the
airplane that’s wrong. In fact, my speed relative to Earth is a tiny bit less
than 605 miles per hour—and for the same reason, namely that measures of
time and space in the airplane’s reference frame aren’t quite the same as on
Earth. Here the difference is negligible, because the relative speed of Earth
and airplane is tiny compared with the speed of light. The effect becomes
more dramatic as relative speed increases.
To get a little more abstract, suppose that the big rocket is moving at
some speed u relative to Earth, and the small rocket is moving at speed v
relative to the big one. Common sense suggests that the speed of the small
rocket relative to Earth—call it v’—should be just v’ = u + v. But relativity
modifies this, giving
For obvious reasons, this equation is called the relativistic velocity addition
formula. Here I’m assuming that all speeds are given as fractions of the
speed of light. The numerator in the formula is just what we’d expect from
common sense; it’s the sum of the two speeds u and v. But the denominator
has the effect of reducing that value. That effect is small if either of the
speeds u or v is much less than the speed of light, since then the product u ×
v is much less than 1 and the denominator remains essentially 1. At high
relative speeds, though, u × v is substantial and the speed of the small ship
relative to Earth is a lot less than common sense would suggest. For
example, let both u and v be 0.75 as in our example, and you’ll see that the
result is v’ = 0.96.
It won’t help to make the small rocket go even faster. No matter what its
speed relative to the big rocket, as long as that speed is less than c then its
speed relative to Earth will also be less than c. We can also ask what will
happen if you shine a light beam inside the big rocket. Obviously, the light
goes at c relative to the big rocket. What’s its speed relative to Earth?
Relativistic velocity addition tells us that, too. The light takes the place of
the small rocket, so we put 1 in for v in our relativistic velocity addition
formula (that’s 1 because we’re measuring all speeds as fractions of c, so
the speed of light itself is just 1):
In other words, the speed of the light is unchanged. It’s the same relative to
Earth as it is relative to the rocket, namely 100 percent of c. Of course, we
already knew this; invariance of the speed of light is at the basis of
relativity. Relativistic velocity addition, like all else, is based on the
Principle of Relativity, so of course it gives a result consistent with that
principle.
So we can’t leapfrog to a speed greater than c by adding two sub-c
motions. That rules out another idea you might have for achieving faster-
than-light travel. Think of taking one of those people movers found in
airports and put another people mover on top of it. People on the second
mover should be going twice as fast relative to the ground as those on the
first. Of course, that still isn’t very fast. So pile more and more people
movers, one on top of the other. Get enough of them and folks on the
uppermost one should be going faster than c relative to the ground. You
should be able to get to that superluminal speed by jumping from one
mover to the next, undergoing just a tiny increase in speed each time. But it
won’t work. Even on the second mover, you aren’t going quite twice as fast
as people on the first one. That slight discrepancy compounds, according to
relativistic velocity addition, so no matter how many people movers you
pile up, none will be going faster than c relative to the ground.
Incidentally, understanding relativistic velocity addition helps dispel the
common misconception that it’s only light that exhibits unusual behavior.
The unusual behavior of light is, of course, that it has the same speed for all
observers regardless of their states of motion. For an object moving at less
than c relative to one observer, another observer in motion relative to the
first will measure a different speed for the object. This makes it seem that
the unusual behavior—invariant speed—is unique to light. But as the
relative speeds of observers and object approach c, the speeds measured by
the two observers become nearly the same and their difference becomes
much less than the relative speed of the two observers. So this relativistic
velocity addition effect applies to everything, not just light. The only
distinction is that with light there’s no difference whatsoever in the speed as
measured by different observers. For objects with relative speeds below c,
there is a difference, but that difference becomes very small as the speed
approaches c. Again, the effect here is not about light but about something
more fundamental, namely, the nature of space and time.
E = mc2
E = mc2 is surely the most famous equation in all of physics, and for most
people it’s synonymous with Einstein and relativity. It’s also, in the popular
mind, the basis of nuclear weapons. So what’s E = mc2 doing here, a
seeming footnote buried deep in a chapter on why faster-than-light travel is
impossible?
In fact, E = mc2 is itself a footnote to special relativity. It doesn’t appear
in Einstein’s famous paper of June 1905. Later that year, Einstein published
a second paper in which he asserted, “The mass of a body is a measure of
its energy content,” or, stated mathematically, E = mc2. He elaborated more
fully on this relation between mass and energy in a 1907 paper.
E = mc2 is important but it’s not the essence of relativity, and it’s no
more the basis of nuclear weapons than it is of a burning candle or a cave
dweller’s fire. Here I’m going to explore the meaning of E = mc2, then
come round to show how it gives what is probably the most satisfying
reason that faster-than-light travel is impossible.
E = mc2 asserts a fundamental equivalence, or, more precisely, an
interchangeability, between matter and energy. Before Einstein, matter and
energy were the two substances that populated the physical Universe. I use
the word “substance” here in the sense of being “substantial”—that is,
having ongoing, indestructible, existence. Matter is obviously substantial;
it’s the stuff we and everything else seem to be made of. Matter is
quantified by its mass, a measure, roughly, of how much of it there is. We
can transform matter from one form to another, as in chemical reactions.
Burning coal is an example; carbon in the coal combines with oxygen in the
air to make carbon dioxide. The total amount of matter seems unchanged;
what’s happened is a rearrangement of atoms to make a new chemical
substance. But the atoms themselves seem unchanged. In other words,
matter seems to be conserved; it can be rearranged, but not created nor
destroyed.
Energy is less tangible than matter, but it too was thought in
prerelativity times to be conserved. Like matter, energy can change forms.
Sunlight falls on a dark surface, transforming what was the energy of
electromagnetic waves (light) into the random molecular motions we call,
loosely, heat. Falling water turns an electric generator, transforming the
energy of motion into electrical energy. Coal burns, transforming chemical
energy into heat. You charge a battery, turning electrical energy into stored
chemical energy. You put the battery in your laptop computer, and the
chemical energy changes back into electrical energy to run the computer.
You step on your car’s brakes, changing the energy of the car’s motion into
useless heat energy (unless you’ve got a hybrid or electric car, in which
case the energy goes back into the battery for later use). I could go on and
on, since transformations among different forms of energy are involved in
virtually everything that happens. In all these transformations, it seems that
the total amount of energy remains unchanged.
What E = mc2 says is that matter and energy are interchangeable. A
piece of matter with mass m could, in principle, be transformed into pure
energy, where the amount of energy, E, is the product of the mass m and the
square of the speed of light. Because c is so big, this means that a little bit
of matter could yield a large amount of energy. The equation works the
other way, too. It says that an amount of energy E could be transformed into
matter with mass m given by E/c2. So matter and energy are, individually,
no longer conserved. What is conserved is a new universal substance, which
for lack of a better name we might as well call mass-energy. The total
amount of mass-energy remains the same, but how much of it is in the form
of mass and how much of it is energy can change.
Is the conversion of matter to energy and vice versa really possible?
Yes, but total conversion occurs only in very special cases. It turns out that
every one of the elementary particles that makes up everyday matter has an
associated antiparticle, identical in mass but opposite in electric charge and
other properties. Corresponding to the negatively charged electron, there’s a
positively charged antielectron, also called a positron. Corresponding to the
positively charged proton is a negative antiproton. Collectively, these
antiparticles constitute antimatter. Our Universe, for reasons that are not yet
entirely clear, seems to consist almost entirely of matter. A few antimatter
particles are created in high-energy collisions and in some nuclear
reactions, either naturally or in laboratory experiments. When a matter
particle and its antimatter opposite meet, the two annihilate, disappearing
altogether in a burst of energy. If the particle and antiparticle each have
mass m, then the total mass that disappears is 2m, and the energy that
appears in its place is therefore 2mc2. The opposite process can occur, too.
Under the right conditions, energy, in the form of a pulse of electromagnetic
waves, can disappear and in its place a particle and its antiparticle come
into existence. Figure 12.2 shows such a pair-creation event, as observed
with a detector used in high-energy physics experiments.
Complete conversion of matter to energy is such a powerful process that
if you had a box of ordinary raisins and a box of antiraisins (much harder to
get hold of!), the energy released in the annihilation of one raisin–antiraisin
pair would be enough to supply all of New York City’s energy needs for a
day. If you had a power plant capable of harnessing that energy, you could
drop in one raisin and one antiraisin each day and that would be all the fuel
you’d need to keep New York going.
Fig. 12.2 Simulated image of a pair-creation event, as observed in the particle detector of a high-
energy physics experiment. The detector records the paths of electrically charged particles. An
uncharged, and therefore invisible, gamma ray–a high-energy bundle of electromagnetic energy–has
entered the region from below. At the point marked, it ceases to exist and its energy becomes that of
an electron and its antiparticle,a positron. A magnetic field points perpendicular to the page and
causes the two new particles to spiral in opposite directions.
Quantum Weirdness
It’s not so much faster-than-light speed that relativity prohibits, rather, it’s
the transmission of information at speeds greater than c. As we saw in the
previous chapter, that prohibition is what saves the logical order of cause
and effect from the otherwise devastating fact that different observers
ascribe different time orderings to the same pairs of events. So tachyons and
other superluminal occurrences aren’t strictly banned, but if they exist they
have to do so in such a way that precludes information transmission at
faster-than-light speed.
Recently there’s been considerable interest in a strange phenomenon in
quantum physics whereby two objects are somehow “entangled” in such a
way that they seem able to communicate instantaneously even when they’re
far apart. This unusual possibility was first suggested in 1935 by Einstein
and his colleagues Boris Podolsky and Nathan Rosen. Einstein, Podolsky,
and Rosen thought up the so-called EPR paradox as a way of questioning
the probabilistic interpretation of quantum physics—an interpretation that
Einstein rejected, most famously in his claim, loosely translated, that “God
does not play dice with the Universe.” Much later, in 1964, the British
physicist John Bell showed how experiments on the EPR phenomenon
could distinguish between probabilistic and deterministic interpretations of
quantum physics. Experiments done in the 1980s and subsequently have
confirmed the probabilistic interpretation, upholding the apparently
instantaneous communication between two widely separated particles.
In a typical EPR experiment a pair of particles is created and moves
away from each other in opposite directions. The particles have certain
properties that can take on particular values; for one property those values
are described by the words “up” and “down.” It’s known for certain that if
one particle is up, then the other must be down. Now here’s the weird thing:
according to the probabilistic interpretation of quantum physics, the
particles don’t have well-defined values “up” or “down” until someone
measures them. It’s not just that those values aren’t known before
measurement—it’s that they’re not even determined. So the particles fly
apart and, later, experimenters measure one of the particles to have the
value “up.” According to quantum physics, the act of measurement is what
causes the particle to have that value; before the measurement it was
undetermined. When the one particle is measured to be up, then the other
must, immediately, become down. Indeed, experiments confirm that this is
the case. Somehow the act of measuring the up or down state of one particle
in the pair has instantaneously fixed the state of the other—even though it’s
far away.
Is this superluminal communication? Are the two particles in violation
of special relativity? Physicists think not. The two particles have, in some
weird quantum-physics sense, a correlated existence. Even when separated,
they act somehow like a single entity. Perhaps this entangled pair of
particles does communicate instantaneously. Here’s the rub: there’s no way
you can use this strange entanglement to send information, because it’s
entirely random whether you measure up or down for the first particle. So
you can’t say, to paraphrase Paul Revere, “‘Up’ if by land and ‘down’ if by
sea . . . ,” because you have no way of influencing whether your
measurement of the first particle will yield up or down.
The EPR phenomenon is philosophically fascinating and reveals a
startling interconnectedness at the heart of physics. You can find out a lot
more about EPR from works on quantum physics. Here, in a book on
relativity, it’s enough to note the phenomenon and to recognize that it
appears not to violate relativity’s prohibition on transmitting information at
speeds greater than c.
Reports appear from time to time on other seemingly superluminal
phenomena, often in astrophysical situations. So far, all have proven to have
other explanations. So solidly grounded is special relativity that if you hear
of a new discovery of something moving faster than light, be skeptical! But,
like any good scientist, remain open-minded too. By the way, I need to
assert again that the cosmic speed limit is the speed of light in vacuum, c,
and that it applies to the transmission of information. It’s entirely possible
for an object to move through a material medium at a speed faster than the
speed of light in that medium. For example, the speed of light in water is
about two-thirds of c. When electrons or other particles move through water
at faster than this speed, they produce shock waves analogous to the sonic
booms from supersonic aircraft. These shock waves are essentially intense
emissions of electromagnetic radiation, produced when waves pile up on
each other because they can’t move fast enough to get away from the
rapidly moving particles. Some very high energy sources of x-rays and
visible light, used in research and technology, take advantage of this
mechanism. You may also hear of certain kinds of waves with speeds
purportedly greater than c. In the mathematical description of wave motion,
there are two different speeds associated with waves. One of them can,
indeed, be greater than c—but this speed is not associated with the
transmission of any information. The second speed, that of the information
carried in the wave, is always less than or (for electromagnetic waves in
vacuum) equal to c.
The cosmic speed limit c really does appear to be inviolable!
CHAPTER 13
IS EVERYTHING RELATIVE?
• • •
Spacetime
The eminent relativity physicists Edwin Taylor and John Archibald Wheeler
begin their delightful book Spacetime Physics with the “Parable of the
Surveyors” (check out the Further Readings list for details). Here they
describe a fictitious town that is mapped by two different surveyors. One
works in the daytime and uses a magnetic compass to establish directions.
The other works at night and takes directions from the North Star. As
anyone who has worked with map and compass knows, magnetic and true
north are not quite the same; hence, the surveyors’ maps are slightly
different, even though they map the same physical town.
Suppose you want to get from point A to point B in Taylor and
Wheeler’s fictitious town. You check the daytime surveyor’s map (Figure
13.1a) and see that you can do this by going 4 miles east and then 3 miles
north. But if you use the nighttimer’s map (Figure 13.1b) you’ll want to go
nearly 5 miles east and then just short of 1.5 miles north. Which map is
right? They’re both right, of course. Either way, you get from A to B. It’s
just that one surveyor’s definitions of north and east aren’t the same as the
other surveyor’s definitions. Each has produced an accurate map, and if you
follow either surveyor’s directions carefully, you can get to any point you
want.
The daytime surveyor says, “Point B is 4 miles east of A, and 3 miles
north.” The nighttimer says, “B is nearly 5 miles east of A, and just under
1.5 miles north.” They disagree about how to describe the position of B
relative to A, but there is something both agree on: If you walk in a straight
line from A to B, rather than following compass directions, you will walk a
distance of 5 miles. Figure 13.1c shows this obvious fact. On either map,
the straight-line distance is the hypotenuse of a right-angled triangle. Either
surveyor can calculate the length of that hypotenuse using the Pythagorean
theorem. Despite the fact that they’re working with different triangles
whose sides have different lengths, both get the same value—it happens to
be 5 miles—for the straight-line distance between points A and B.
Fig. 13.1 (a), (b) Two maps describing the same region but using different definitions of the compass
directions. Instructions for getting from A to B differ depending on which map one is using, but, as
(c) makes clear, the straight-line distance from A to B doesn’t depend on one’s choice of map.
This is all very obvious, but let me make clear why it is so. It’s because
the distance between points A and B is an objective fact about the world, a
fact that does not depend on any particular choice of compass directions. A
hundred different surveyors, with a hundred different choices for the
direction of north, would come up with a hundred different maps and a
hundred different sets of instructions for getting from A to B. But all would
agree that the straight-line distance is 5 miles. In other words, the distance
from A to B is an invariant, a quantity that doesn’t depend on one’s
particular point of view as embodied, in this case, in the choice of which
direction to call north. Invariant quantities can claim to be objectively real
in a way that quantities dependent on one’s point of view cannot.
Taylor and Wheeler’s surveyor parable is an analogy for relativity. In
relativity, the points A and B are not points in space; they are events
characterized by their locations in both space and time. Spatial distances in
relativity are like east–west distances in the fictitious town. They depend on
one’s point of view—in the case of relativity, on one’s state of uniform
motion rather than one’s choice of north. Time intervals are like north–
south distances. They, too, depend on one’s point of view, that is, on one’s
choice from among the infinitely many possible uniformly moving
reference frames. The spatial distance between two events and the time
between those events are not absolute but relative to one’s frame of
reference. This, of course, you know from earlier chapters.
But everything is not relative, and there are absolutes even in relativity
theory. In Taylor and Wheeler’s parable, the different surveyors could use
their different east–west and north–south distances to calculate the same,
invariant distance between points A and B. Similarly, in relativity, observers
in different reference frames can use their different measures of spatial and
temporal separation between two events to find an invariant “distance” not
in space alone but in spacetime. All uniformly moving observers, no matter
what their states of relative motion, will agree on the value of this
spacetime interval between any two events. Because its value is
independent of one’s point of view (read “frame of reference”), that interval
can claim a reality that relativity denies to measures of space alone or time
alone.
The different surveyors in Taylor and Wheeler’s town chose to break the
two-dimensional space of their town into north-south-east-west in different
ways, but the underlying reality of the town itself didn’t depend on those
arbitrary choices. Similarly, we break the underlying reality of spacetime
into space and time in different ways, depending on our states of relative
motion. It’s not space and time that have absolute, invariant reality, but
rather their combination in the single entity called spacetime. Where the
surveyors’ space was two-dimensional, spacetime must have four
dimensions because there are three independent spatial dimensions and one
of time.
The German mathematician Hermann Minkowski, whose classes
Einstein had attended as a student, later studied relativity and was first to
conceive the idea of four-dimensional spacetime and the invariance of the
spacetime interval. Minkowski summarized his insight in an oft-quoted
statement: “Henceforth space by itself, and time by itself, are doomed to
fade away into mere shadows, and only a union of the two will preserve an
independent reality.”*
Given our map analogy, you might expect that Minkowski’s spacetime
interval follows from applying the Pythagorean theorem to the space and
time intervals between two events. Actually, though, the mathematics is
different. Where the Pythagorean theorem gives the square of a right
triangle’s hypotenuse as the sum of the squares of the two sides, the
spacetime interval in relativity has its square given by the difference of the
squares of the time and space intervals. That simple change from a positive
to a negative sign tells us that spacetime does not obey the rules you learned
back in tenth-grade geometry. Furthermore, it raises the possibility that the
square of a spacetime interval can be positive, negative, or zero. That
mathematical division into three classes corresponds to Chapter 11’s
classification of event pairs as being either causally related or not. For those
events that are close enough in space and far enough apart in time that an
observer can be present at both—and thus the events can be causally related
— the square of the spacetime interval is positive. Such an interval is called
timelike because an observer with just the right motion can be present at
both events, so for that observer the events are separated in time only. The
numerical value of the interval is just the time read on that observer’s clock.
In other reference frames the separation is a mix of space and time, but the
time separation always dominates; hence the name “timelike.” On the other
hand, if the two events are so far apart in space and so close in time that
they can’t be causally related, then the square of the spacetime interval is
negative and no observer can be present at both events. For an observer in
just the right reference frame, the two events will be simultaneous. For this
observer their separation is purely spatial; hence, the interval is spacelike.
Take the negative sign off its square, take the square root, and the numerical
result is the distance between the events in the frame where they’re
simultaneous. Between the general cases of timelike and spacelike intervals
is the lightlike interval, whose numerical value is zero even though the two
events are located at different places and occur at different times. Events
separated by lightlike intervals correspond to the emission and arrival of a
light flash. We’ll soon see how the unusual geometry suggested by
relativity’s modified Pythagorean theorem becomes central to Einstein’s
general theory of relativity.
The Meaning of c
There’s one further complication, and in Taylor and Wheeler’s parable it
involves an unusual religious convention. In their fictitious society, north–
south distances are considered sacred and are measured in a special unit—
the mile. East–west distances are not sacred and are measured in ordinary
meters. Different surveyors cannot consistently describe the straight-line
distance between points A and B because that distance involves two
different units for measuring spatial separation. Their religious convention
keeps them from doing the obvious—converting all distances to the same
unit, be it miles or meters. Then a brilliant and creative new surveyor
arrives in town. This newcomer compares the different surveyors’ maps,
and—heresy!—converts all distances to the same unit. Henceforth all
surveyors find that they agree on the straight-line distance between any two
points. In other words, they discover an invariant aspect of reality that
doesn’t depend on their different points of view.
Einstein is the newcomer. What his relativity shows is that time and
space, before considered separate quantities, are actually aspects of the
same underlying physical reality. That reality is obscured by the convention
of measuring time in one unit (seconds, for example) and space in another
(meters). In reality, we should use the same units for both. To do so we need
to know the conversion between seconds and meters—just as the surveyors
needed to know the conversion between miles and meters to put their
measurements in the same units. What is that conversion in the case of
meters and seconds? Simply the speed of light, c. What’s a meter of time?
It’s the time it takes light to cross a meter of space. What’s a second of
space? It’s the distance light goes in 1 second of time. We can use either
unit for both space and time, or any other distance or time unit we choose.
In this book we haven’t gone quite all the way to treating space and time
with exactly the same units. Instead, we’ve been measuring time in years
and distance in light-years. But the effect is the same; the conversion
between time and space units is the speed of light, in this case 1 light-year
per year. Had we dropped the “light-” we would have been measuring both
distances and times in precisely the same unit, namely, the year.
Viewed in this sense, the speed of light isn’t so much a speed as it is a
conversion factor between units of space and time. It’s pure coincidence
that it has the value 299 792 458 meters per second; that number is just an
artifact resulting from our arbitrary decisions to define the meter and second
as we first did—definitions that were made long before anyone knew
anything about relativity. If we embrace relativity fully, we should dispense
with one of those units and choose a single unit for both space and time. In
that more philosophically sensible system, the speed of light is just the
number 1. Of course, measuring time and space in the same units isn’t very
practical for beings who creep about their planet at speeds much less than c.
A 60-mile-per-hour highway speed limit would have to read 1/10,000,000
in those units! But when we’re discussing hypothetical high-speed space
trips, or for physicists working with elementary particles moving through
their labs at relativistic speeds, it makes eminent sense to use units where c
= 1. In those units space and time are implicitly aspects of the same
underlying reality.
Is Everything Relative?
We now have a firm answer to the question that forms this chapter’s title.
Everything is not relative. If it were, there would be nothing absolute about
physics and that science could hardly claim to be a description of an
objective physical reality. Cherished absolutes of our common sense—in
particular, space and time—have lost their absolute status and have become
relative to our particular point of view (i.e., frame of reference). But they’re
not lost for good; they’ve simply merged into the greater, more
encompassing absolute of four-dimensional spacetime. Space and time
aren’t the only players on this four-dimensional stage; many other physical
quantities that once seemed distinct have merged into four-dimensional
entities with an absoluteness that transcends particular frames of reference.
Ultimately, relativity has enhanced, not diminished, our sense of an
objective and absolute reality.
* Hermann Minkowski, quoted in E. F. Taylor and J. A. Wheeler, Spacetime Physics, 2d ed. (New
York: W. H. Freeman, 1992), p. 15.
† See, for example, R. Wolfson and J. M. Pasachoff, Physics for Scientists and Engineers, 3d ed.
(Addison Wesley Longman, 1999), pp. 1035–7; the sections just before this also explore relativistic
invariants. A more thorough analysis is given in E. M. Purcell, Electricity and Magnetism, 2d ed.
(New York: McGraw-Hill, 1985), chap. 5.
CHAPTER 14
A PROBLEM OF GRAVITY
• • •
A Shaky Foundation
The foundation of special relativity is the principle that physics is the same
for all observers in uniform motion. How do you know whether you’re in
uniform motion? At first glance that seems easy. In an airplane, for
example, you can tell the difference between smooth air and turbulence. In
the former the plane’s motion is uniform; in the latter it’s anything but
uniform. You know the motion in smooth air is uniform because things
around you, like the airline peanuts you’ve spread over your seatback tray,
stay put. In other words, things in the plane obey Newton’s first law, the law
of inertia: if they’re at rest, they remain at rest. You could play tennis on
this uniformly moving plane and the tennis ball would behave as you expect
—that is, according to Newton’s laws. Turbulent air, in contrast, would send
your peanuts flying all over the place, and no way could you have a
satisfactory tennis match. So it looks like you can tell pretty easily whether
you’re in uniform motion.
We can make the criterion for deciding if you’re in uniform motion
more concise and scientific: You’re in a reference frame in uniform motion
if, in your reference frame, objects obey the law of inertia. That is, if an
object at rest in your reference frame remains at rest without the need for
any forces to hold it at rest, then your reference frame is in uniform motion.
Another, and actually better, name for such a reference frame is an inertial
reference frame, so called because of the link to the law of inertia. On the
other hand, if you need to apply forces to keep objects at rest—as you
would have to do to your airline peanuts when the plane encounters choppy
air—then you know you’re not in an inertial frame.
Or do you? How do you know for sure that there aren’t forces acting on
objects to keep them at rest? If you thought there weren’t, but there really
were, then you would mistakenly conclude that you were in an inertial
reference frame. After all, the most fundamental forces, like electricity,
magnetism, and gravity, are invisible. Having an electric force acting isn’t
as obvious as seeing your hand holding the peanuts to keep them at rest.
Maybe—just maybe—your airline peanuts are electrically charged and
there’s some clever arrangement of electric field holding them in place on
the tray despite what might, in fact, be nonuniform motion. Farfetched?
Yes, but you just might be deceived.
For electric and magnetic forces, fortunately, there’s a simple test. Not
all objects are electrically charged. So find one that is and one that isn’t.
Watch how they behave. If they behave the same way, then there’s no
electric force acting. If they behave differently, then the difference must be
caused by an electric force. The same idea works for magnetism: some
objects are magnetic and some aren’t. So simple experiments will tell you
whether electric or magnetic forces are present.
Gravity is different. Gravity affects all objects. Furthermore, as Galileo
purportedly showed by dropping different objects from the Tower of Pisa,
gravity affects all objects in the same way. That is, all objects, regardless of
how massive they are, undergo the same acceleration in response to gravity.
So gravity could be deceiving you. That is, you might think objects remain
at rest because you’re in a uniformly moving reference frame with no forces
acting, when really gravity is acting to keep them in place in a reference
frame whose motion is not uniform. Or, you might think that objects are
undergoing accelerated motion relative to you because your reference frame
is accelerating when in fact it isn’t but the objects are accelerating due to
the force of gravity. (If all this sounds a bit obscure, be patient; I’ll be
giving concrete examples in the next section.)
But surely you can tell the difference between a reference frame that is
“really” in uniform motion and one that “really” isn’t. Be careful! I
cautioned earlier about using the word “really” in cases where relativity
renders it meaningless, as in “You’re really moving and I’m really at rest.”
Statements like that are ultimately meaningless because no physical
experiment can distinguish absolutely between being at rest and moving.
Only relative motion is meaningful. That, of course, is what special
relativity is all about. Now we’re finding that there seems in principle to be
no way of telling with absolute certainty whether you’re in uniform motion.
So statements like “I am in uniform motion” are as meaningless as the
statement “I am moving.” This presents a big problem for special relativity,
which is a theory about the laws of physics being the same for observers in
uniform motion. If you can’t tell for sure that you’re in uniform motion,
how can you ever know if special relativity is valid for you?
So special relativity is on shaky ground and, again, the culprit is gravity.
The problem of instantaneous information transmission in Newtonian
gravity means that we need a new theory of gravity and a more general
theory of relativity that encompasses that new gravity. Now we have the
additional problem that gravity make us unable to tell for sure whether
we’re in uniform motion. For that reason, too, we need a relativity theory
that doesn’t just apply to uniform motion. Such a general theory must
necessarily consider gravity.
Fig. 14.1 The Principle of Equivalence. (a) A ball released in a reference frame at rest on Earth falls
to the floor with constant acceleration. (b) Exactly the same thing happens in an accelerating
reference frame in intergalactic space, far from any source of gravitation.Here the acceleration is
upward and provided by the rocket motor. The situations of (a) and (b) are equivalent as far as the
laws of physics are concerned.
Now, the instant you let go of the ball, there’s no longer any force acting
on it. So what does it do? It obeys the law of inertia, and it remains at rest
relative to a uniformly moving reference frame that had the same speed as
the room at the instant you released the ball. The room itself is not that
uniformly moving frame, since it’s accelerating. Relative to the uniformly
moving reference frame I’ve just identified, the ball remains at rest and the
floor of the room accelerates toward it at the rate of 10 meters per second in
each second. What’s this look like to you in the room? You share the room’s
motion, so it looks to you like the ball accelerates toward the floor, gaining
speed (relative to you) at the rate of 10 meters per second in each second
(Figure 14.1b). But this is just what happened when the room was at rest on
Earth! From your point of view, inside the room, the outcome of the ball-
drop experiment is the same in either situation. That experiment can’t be
used to decide whether you’re in an unaccelerated reference frame in the
presence of Earth’s gravity or, instead, in an accelerating reference frame in
the absence of gravity. Nor, in fact, can any other experiment involving the
laws of motion.
In the accelerating room, incidentally, you also feel just as you would
when standing on Earth. The floor of the room pushes up to accelerate you,
and by Newton’s third law you push down on the floor. The muscles in your
body tense and compress just as they would when standing at rest on Earth.
So your bodily sensations, too, give you no way to tell whether you’re in an
accelerating reference frame, absent gravity, or in an unaccelerated frame
with gravity present.
Now consider two other situations. First, put the little room in
intergalactic space again, but now turn off the rocket engine. What
happens? Gravity is completely negligible, so when you release the ball, it
just stays there. In this case the room is not accelerating, so it constitutes a
uniformly moving reference frame in which the law of inertia is valid. The
ball was at rest relative to this frame, and it remains at rest. It just floats
there, next to your hand (Figure 14.2a).
Finally, consider that the little room is an elevator, whose cable has
snapped. It’s plummeting toward Earth, accelerating downward at that same
10 meters per second each second. You and the ball experience exactly the
same acceleration, as shown by Galileo’s purported experiment (and by
many other more modern and very precise measurements). So when you
release the ball, relative to you in the falling elevator room, it again just
floats next to your hand (Figure 14.2b). In the two situations of Figure 14.2
—being in a truly gravity-free environment and falling freely in the
presence of gravity—the ball-drop experiment has exactly the same
outcome. Within the confines of the room, that experiment can’t help you
decide which situation you’re in. Nor could any other experiment involving
the laws of motion.
Fig. 14.2 Two other equivalent situations. (a) In intergalactic space, far from any source of gravity.
With the rocket engine off, the reference frame is unaccelerated. You float about the room, and a ball
released from rest doesn’t go anywhere. (b) Exactly the samething happens in a reference frame
falling freely toward Earth because you, the room, and the ball are all accelerating downward at the
same rate.
So the two situations of Figure 14.1 are indistinguishable, and so are the
two situations of Figure 14.2. Indistinguishable, that is, by experiments
involving the laws of motion. Why this indistinguishability? The reason,
ultimately, is that all objects experience the same acceleration in the
presence of gravity. In Figure 14.1b, the room’s acceleration mimics the
effect of gravity on any object, because all objects near Earth would fall
with the same acceleration. In Figure 14.2b, the ball floats beside your hand
because you, the ball, and the room all experience the same gravitational
acceleration.
Because no experiment with the laws of motion can distinguish them,
the two situations in Figure 14.1 are in a sense equivalent. So are the two in
Figure 14.2. What is the ultimate reason for that equivalence? It’s the fact
that gravity has exactly the same effect on all objects. If that weren’t true,
different objects in Figure 14.1a would fall with different accelerations, but
all objects in Figure 14.1b would still have the same acceleration relative to
the room—and the situations wouldn’t be equivalent. Absent other forces,
objects with dramatically different masses really do fall with exactly the
same acceleration. Apollo 15 astronaut David Scott dropped a hammer and
a feather on the Moon; with no air resistance, they hit the ground at exactly
the same time.
The simple fact of equal gravitational acceleration for all objects,
known since Galileo’s time, is nevertheless remarkably profound. Here’s
why: The mass of an object is a measure of its inertia, that is, of how hard it
is to change its motion. That’s what Newton’s second law says; it tells us
that it takes a bigger force to give a more massive object the same
acceleration as a less massive object. There’s nothing whatsoever about
gravity in that statement. On the other hand, the mass of an object is also a
measure of the force gravity exerts on that object. Drop a bowling ball and a
tennis ball. It takes a much greater force to give the bowling ball the same
acceleration as the tennis ball yet they fall with the same acceleration.
Why? Because the gravitational force on the bowling ball is greater by just
the amount needed to compensate for its greater resistance to being
accelerated. In other words, mass as a measure of resistance to changes in
motion goes hand-in-hand with mass as a measure of gravitational force.
But why?
Why should an object’s inertia—the property that determines its
resistance to changes in motion—be essentially the same as the property
that determines the force of gravity on the object? Until Einstein’s time, this
essential equivalence of two different meanings of the term “mass” was
considered a coincidence, and physicists spoke of “inertial mass” and
“gravitational mass” as distinct properties of an object that happened to
have the same value. What Einstein saw, though, was a hint of a deeper
relation between gravity and accelerated motion. Einstein elevated to the
status of a fundamental principle the equivalence inherent in Figures 14.1
and 14.2, and the underlying equivalence between the two meanings of
“mass.” With the same incisively simplifying genius he showed when he
applied the principle of special relativity to all of physics, Einstein declared
the two situations of Figure 14.1, and the two of Figure 14.2, to be
fundamentally indistinguishable by any physical experiments, not just those
involving the laws of motion. This Principle of Equivalence is at the heart
of general relativity.
Weightless!
You might object to my assertion of equivalence between the two situations
in Figure 14.2 on the grounds that the hapless elevator occupant in Figure
14.2b will smash to smithereens when the elevator hits the ground. That’s
true, but beside the point. What I’m declaring equivalent are an
unaccelerated reference frame in the true absence of gravity and a reference
frame in accelerated motion under the influence of gravity alone. When the
elevator hits the ground, strong nongravitational forces act on it and its
unfortunate occupant, and the equivalence is no more. But there’s a way
around this unhappy outcome. There is nothing sacred about falling down.
The equivalence would still hold if, in Figure 14.2b, I launched the elevator
sideways, so it fell in a curving path. Because everything in the elevator
would share its initial motion, all those objects would follow the same
curving path and would appear to float relative to the elevator. Of course
the elevator would still hit the ground, just not directly below its launch
point. But—as Newton first suggested, and as modern spaceflight confirms
daily—if I launched the elevator fast enough, and from a point above the
atmosphere to avoid the force of air resistance, it would “fall” around the
Earth in the circular path we call an orbit (recall Figure 3.4). Because all
objects experience exactly the same acceleration due to gravity, you, the
ball you attempt to drop, and everything else in this orbiting elevator would
float freely about relative to the elevator. That’s the origin of the apparent
weightlessness that occurs in spaceflight. It isn’t that there’s no gravity in
space or that objects in a spacecraft have somehow ceased to experience
gravity. It’s just that all objects experiencing only gravity have exactly the
same acceleration, so they all “fall” together and thus experience no relative
motion.
Being “weightless” has nothing, inherently, to do with being in space.
It’s just that in space one gets rid of the pesky, nongravitational force of air
resistance and the danger of hitting the ground. NASA’s “Vomit Comet”
training aircraft executes an arcing flight that essentially cancels the forces
of air resistance. Those onboard temporarily experience the same
“weightlessness” as astronauts on the Space Station. “Weightless” scenes in
the movie Apollo 13 were filmed on the Vomit Comet. Reviews praised the
film for its realism in “simulating the weightlessness of outer space.”
Nonsense! That was no simulation. Any reference frame under the influence
of gravity alone is a reference frame in which objects are “weightless.”
The state of motion under the influence of gravity alone is often called
free fall: free because nothing other than gravity is acting; fall because
gravity is acting to change the motion of an object in the direction toward a
gravitating body like Earth or Sun. Most of us have trouble shaking the
Aristotelian view that gravity should make things move downward, as
opposed to the Newtonian view that gravity changes motion toward the
downward direction without necessarily moving an object downward. For
that reason free fall conjures up images of objects actually plummeting
toward Earth. That state is, indeed, free fall—but so is the motion of the
International Space Station, in its never-ending circular “fall” around Earth.
For that reason the term free float is a more apt description of motion under
the influence of gravity alone. The Space Station is in free float, along with
everything inside it. So is the plummeting elevator, at least for a while. Free
float simply means that the only influence acting is gravity.
What Is Gravity?
What you think of as gravity—the heaviness you feel standing on Earth or
the force that accelerates a plummeting stone—is not real because it
disappears in some reference frames (i.e., in any free-float frame). That
doesn’t mean there’s no such thing as gravity; rather, the real gravity must
be something that can’t be transformed away with a change of reference
frame.
I’ve argued that gravity disappears completely in a free-float reference
frame, like an orbiting spacecraft or the little room of Figure 14.2b.
Actually, that conclusion isn’t quite true. Let’s make the room very much
bigger, as suggested in Figure 14.3. You, too, have become gigantic, with an
arm span of several thousand miles. You perform the ball-drop experiment
again, this time dropping not one ball but two, one from each hand. What
happens? Each ball accelerates downward, falling on a straight-line path
toward Earth’s center. Of course, you and the giant room are also
accelerating downward, so again the balls appear, at first, to float right next
to your hands. Then something strange happens. The balls begin to
approach each other, the distance between them shrinking as time goes by.
You can see why this occurs: The falling room is now so large that the
direction “down” is significantly different for the different balls. Each falls
on a straight-line path toward Earth’s center, but those paths converge, as
Figure 14.3 shows. As a result, the balls drift closer together in the free-
float reference frame of the huge, falling room.
Fig. 14.3 Tidal forces arise even in a freely falling reference frame. In the Newtonian view,these
forces result from differences in the direction or strength of gravity. Here the two balls fall toward the
center of the Earth and thus their horizontal separation decreases, an effect evident even to the freely
falling observer. As (b) shows,this effect also squeezes the observer’s body in the horizontal
direction, while the decrease in gravity’s strength with distance from Earth results in a stretching in
the vertical direction. I’ve assumed the falling room is sufficiently rigid that it isn’t significantly
stretched or squeezed.
Here’s an effect that doesn’t go away when you jump into a free-float
reference frame. It’s an effect you can use to distinguish between the
situations of Figure 14.2. But it’s a distinctly nonlocal effect—an effect you
can’t notice unless your experience spans a large region of space. (Actually,
I should say, a large region of spacetime, because you won’t notice the
effect even in the large room of Figure 14.3 unless you let the experiment
proceed for some time.)
Newton would say that the drifting together of the two balls in Figure
14.3 is a subtle effect resulting from the variation of Newtonian gravity
from place to place, in this case a difference in the direction of the
Newtonian gravitational force at the locations of the two balls. In a related
experiment, you could imagine dropping two balls, one at your head and
one near your feet. Because Newtonian gravity is stronger closer to Earth,
the lower ball falls with a slightly greater acceleration than the upper one;
as a result, the distance between the balls would increase as they fell. A
solid but deformable object, like a blob of Jello or even your body, would
experience these effects as a squeezing sideways and a stretching vertically
(Figure 14.3b). These subtle effects, known since Newton’s time, explain
the squeezing and stretching of Earth’s oceans that we call tides. For that
reason, they’re called tidal forces. In Newton’s view the tides aren’t caused
by gravity itself but by variations in gravity from place to place.
To Einstein, what Newton calls gravity doesn’t exist, because it
disappears in a free-float reference frame. Tidal effects, however, don’t
disappear, so they can claim to be real. For Einstein, those effects are the
true manifestation of gravity. Newton says, “Tidal effects result from
variations in gravity.” Einstein says, “Tidal effects result from gravity
itself.” To Einstein, gravity in the Newtonian sense doesn’t exist, so there’s
nothing to vary.
Curved Spacetime
So what is gravity? To Newton, it’s a mysterious force that somehow
reaches out across an immutable, unchanging, universal, three-dimensional
space to influence distant objects. To Einstein it’s much simpler. Gravity,
Einstein says, is synonymous with the geometry of spacetime.
Imagine yourself a tiny being confined to the surface of a large, smooth
sphere. Because you’re so tiny, your day-to-day existence occupies only a
small portion of the sphere. You study your world by doing the sorts of
things you did in tenth-grade geometry. You draw triangles, for example,
and discover that their angles add up to 180 degrees. You draw parallel
lines, follow them for some distance, and find that they don’t intersect
(Figure 14.4a). You can take a journey halfway around the sphere and
repeat these geometric experiments in another small region. You’ll come to
the same conclusion—that your world obeys the laws of tenth-grade
geometry. This geometry is also called Euclidean geometry, after the
mathematician Euclid, and so you declare that your space is Euclidean.
At least, it’s locally Euclidean. The region in Figure 14.4a is so small
that you don’t notice the curvature of your world. Your measuring
instruments just aren’t sensitive enough to tell the difference between a
truly flat world and a very large sphere. That’s why Euclidean geometry
works for you. (This situation is not so far-fetched. We all know that Earth
is round, but that’s because we’re told so. Direct evidence for a round Earth
is not at all obvious from everyday life, and if we’re not airplane pilots or
astronauts, most of us could get along just fine in our local communities by
treating Earth as flat.)
Fig. 14.4 (a) A small portion of a sphere is indistinguishable from a flat plane and obeys the laws of
Euclidean geometry. Parallel lines never meet, and the angles of a triangle add to 180 degrees. (b) On
a larger scale the sphere’s surface is decidedly not Euclidean.Parallel lines approach and would
eventually intersect. The large triangle shown has three right angles, which add to 270 degrees.
If you explore a larger portion of your sphere world, then you begin to
notice strange things. For example, you and a friend start out walking on
perfectly parallel paths. You both continue walking absolutely straight, but
after a while you notice that you’re getting closer together (Figure 14.4b).
Why? Because the straightest possible lines on the surface of a sphere, the
shortest paths between any two points, are circular arcs centered on the
sphere’s center (in geography they’re called great circles). Follow any two
of those arcs and eventually they intersect (at two points, no less, if you go
far enough). Or draw a really big triangle. I’ve shown one in Figure 14.4b
that is so big it extends from equator to pole, and I’ve made it an equilateral
triangle (all sides the same length). But its angles aren’t the 60-degree
angles of a Euclidean equilateral triangle; rather, they’re each 90 degrees,
and sum not to 180 degrees but to 270 degrees! Euclidean geometry just
doesn’t hold on this spherical world. The reason is obvious to those of us
who can see the whole sphere from afar: it’s because the sphere’s surface is
curved, not flat.
Confined to the sphere’s surface, you might try to explain these strange
results by proposing a mysterious “force” that tugs on you and your friend,
pulling you away from “true” straight lines and causing your paths to
converge. Similarly, the tape measures you use to stake out that big triangle
might be deflected by the same force, accounting for the unusual angles.
You might try to learn more about this force by varying your experiments.
For example, replace your friend by an elephant and repeat the parallel-path
experiment. Obviously, the same thing happens. You and the elephant
gradually approach each other, despite the most meticulous effort at
following straight paths. That gradual approach is just the same for the
elephant as it was for your friend; therefore, you conclude that your
proposed “force” has the same effect on all objects, independent of how
massive they are. But your force idea is complicated and cumbersome, and
invokes something that just can’t be detected in a small, local region of your
globe. How much simpler is the correct explanation! The geometry of this
spherical world is not Euclidean; rather, it has curvature and that curvature
accounts fully for strange effects like the gradual approach of objects on
parallel, straight-line paths.
I’ve just presented an analogy for Einstein’s conception of gravity. We
live, says Einstein, in a four-dimensional spacetime. The geometry of
spacetime exhibits curvature, and that spacetime curvature is gravity. Not
“is a manifestation of gravity” or “is caused by gravity” or “causes gravity.”
No: spacetime curvature is gravity. Gravity is not some force that affects
objects in spacetime. Gravity is no more and no less than the curved
geometry of spacetime.
In a small, local region of spacetime—that is, in a local free-float
reference frame—you don’t notice gravity because spacetime curvature is
negligible over small regions of space and time. Similarly, a small region on
the surface of our hypothetical globe is essentially flat and obeys the laws
of Euclidean geometry. In either case, one’s world is simple when viewed
locally. On a small patch of the sphere, you can use tenth-grade Euclidean
geometry. Objects move in straight lines, parallel lines don’t intersect, and
triangles have 180 degrees. In a small, localized free-float reference frame
in spacetime, physics is simple. The law of inertia holds, with free objects
moving in straight lines at constant speed. It’s only when you go beyond
your local neighborhood that you notice deviations from simple geometry
and simple physics.
A description of physical reality in a local free-float reference frame
makes no mention of gravity. Objects at rest remain at rest, and objects in
motion remain in straight-line motion at constant speed. Make that
reference frame a space station orbiting Earth and this simple description
remains true. There’s no mention of Earth and its “gravity.” Objects in the
space station obey very simple physical laws because, locally, spacetime is
flat. Here’s a big philosophical advantage of Einstein’s view: There’s no
such thing as “action at a distance.” Rather, matter takes its “marching
orders” from its immediate vicinity, that is, from the local geometry of
spacetime. When that geometry is flat, as it always is in a small enough
region of spacetime, those marching orders say to remain at rest if you’re at
rest, and to keep moving uniformly if that’s what you’re doing. What could
be simpler?
Things get more complicated only when we consider larger regions of
spacetime, large enough that spacetime curvature becomes noticeable.
Then, as in Figure 14.3 (and its analog, Figure 14.4b), we notice the effects
of that curvature on the paths of widely separated objects. We can call such
effects “gravitational,” but we could equally well call them “geometrical.”
Gravity is synonymous with the curved geometry of spacetime.
Locally, physics is always simple. The laws of physics work the same
way in every small free-float frame of reference. But because of gravity—
because of spacetime curvature—there is no one free-float frame that spans
all of spacetime. The local free-float frame in one small region of spacetime
is not the same as the local free-float frame in another region. For example,
a freely falling elevator in New York is not the same free-float reference
frame as a freely falling elevator in Bombay, on the other side of the Earth.
That physics works equally well (and simply) in both elevators is the
essence of special relativity. What general relativity does is to provide the
link between different free-float frames, giving a description of physics that
is consistent across large regions of spacetime. As we’ll soon see, general
relativity also tells how and why spacetime is curved.
Before moving on to summarize general relativity, I want to counter an
objection you might have to my sphere analogy. Aren’t there “really”
straight lines that cut through the interior of the sphere? Doesn’t the big
triangle “really” have 60-degree angles if I use those “truly straight” lines
as its sides? Yes—but I want you to consider only the surface of the sphere.
That’s a two-dimensional surface because on it you can move in only two
mutually perpendicular directions. Now, you can’t imagine that sphere
without seeing it curved in three-dimensional space. As the hypothetical
tiny creature on the sphere, though, you can do experiments that
demonstrate your world’s curvature without ever leaving its surface. Walk
in a straight line and eventually you come back to your starting point.
Measure a big triangle and find that its angles add to more than 180
degrees. You just can’t explain those happenings in a Euclidean world. You
don’t have to imagine three-dimensional space to recognize that the
geometry of your world is not Euclid’s geometry. In fact, you can give an
accurate and consistent description of the sphere’s curved two-dimensional
surface without any need for a third dimension. The spherical surface’s
curvature is an intrinsic property of the surface itself, complete without any
reference to a higher dimension. It’s just that your three-dimensional brain
finds it much easier to acknowledge that curvature if you picture the
spherical surface as being curved in a third dimension.
Similarly, you and I have difficulty wrapping our minds around the idea
that we live in a curved four-dimensional spacetime. Four dimensions are
hard enough, and now they’re curved as well? What are they curved in, a
fifth dimension? No: just like the denizens of that sphere world, we can do
experiments in our four-dimensional spacetime that tell us its geometry is
not flat. That curved geometry is an intrinsic property of spacetime, and it
doesn’t require a fifth dimension. That curved geometry is gravity.
Fig. 14.6 A two-dimensional analogy for curved spacetime. The large sphere distorts a rubber sheet.
Smaller objects moving on the sheet naturally follow curved paths. One object is moving fast enough
that it gets deflected, then continues on; the other is trapped in a closed orbit around the large sphere.
In both cases the motions result not from a force of attraction to the sphere but from the curved
geometry of the rubber sheet.
Now suppose we roll a little ball along the curved sheet. Far from the
large ball, the sheet remains nearly flat, and the little ball rolls in a straight
line. As it approaches the large ball, the small one encounters substantial
curvature and is deflected from its straight-line path. In fact, you can even
put a smaller ball in “orbit” around the larger one, as shown in Figure 14.6.
Let’s make the rubber sheet perfectly transparent and look straight down
on it. All you see are the large ball and the small one. You study the motion
of the small ball and conclude that it doesn’t follow a straight line but gets
deflected toward the larger ball. That deflection is minimal far from the
large ball but becomes substantial closer in. You can even give the small
ball just the right speed that it’s deflected at a constant rate and circles
endlessly around the large ball. You might well conclude that there is an
attractive force between the two balls and that the force gets weaker with
increasing distance between them. You would, of course, be reasoning like
Newton, and discovering what you would happily call the “force of
gravity.”
But those of us who can see and touch the rubber sheet know better.
There is no attractive force, no “force of gravity.” There’s only the rubber
sheet, whose geometry is curved in the presence of the large ball. The
amount of curvature depends on how close you are to the large ball. Small
objects move across this curved sheet on simple paths, namely, the
straightest lines possible in the curved geometry. This analogy, in a nutshell,
describes general relativity. There is no “force of gravity.” There’s only
spacetime, whose geometry is curved in the presence of matter. Objects
move through this curved spacetime on the simplest, straightest paths
possible in the curved geometry.
Note, by the way, that the actual curvature of the rubber sheet is not
synonymous with the curved path of the small ball. You can see that it isn’t
by looking at how the sheet is distorted. Furthermore, my rubber-sheet
example is incomplete because it shows only curvature in space, while
gravity is the curvature of spacetime. So what does an object’s path—say,
Earth’s orbit around the Sun—look like in curved spacetime? That’s
impossible to draw accurately, for the same reason you can’t make an
accurate map of the spherical Earth on a flat sheet of paper. So the picture I
come up with cannot be a realistic depiction of Earth’s motion in curved
spacetime. In particular, the orbital path I show doesn’t look at all straight
or “straightest” any more than the “straightest lines” on the globe in Figure
14.4 look straight. Those caveats understood, take a look at Figure 14.7a.
The orbital path—Earth’s worldline, in the language of Chapter 11—is a
spiral, because Earth goes around a circle in space while advancing straight
into the future in time. Now, the diameter of Earth’s orbit is about 200
million miles—that’s 16 light-minutes, or about 30 millionths of a light-
year. It takes Earth, of course, 1 year to complete 1 orbit. To show a few
years’ motion, then, I’ve had to use very different scales for the space and
time measures in Figure 14.7a. If I had used the natural relativistic units of
the spacetime diagrams in Chapter 11 (i.e., light-years for space and years
for time, with each unit occuping the same physical distance on the
diagram), then the time axis in Figure 14.7a would have to be about a
thousand miles long or the spatial extent shrunk to some tens of millionths
of an inch! I simply can’t make a correctly scaled diagram without either
going way off the page or shrinking the spatial scale so small that the
circular orbit becomes microscopic. (I’ve tried in Figure 14.7b, but even
there the scales are nowhere near correct.) Still, you get the point: Earth’s
path through spacetime is a very, very loose spiral; our planet advances
much more in time (1 year) in each orbit than it does in space (an orbital
circumference, or 200 millionths of a light-year). Drawn to scale, that spiral
is very nearly a straight line—an indication that spacetime curvature in
Earth’s vicinity, although enough to produce our planet’s circular orbit, is
just not very great.
Fig. 14.7 Spacetime diagrams of Earth’s orbital motion, showing two spatial dimensions and time.
Helical paths represent Earth’s worldline, its path in spacetime. Time and space scales are very
different. In both versions the time axis spans 2.5 years, while in (a) the spatial extent is only about
20 light-minutes. Dashed circle shows Earth’s orbit in space alone. In (b) the spatial scale spans about
250 light-minutes, and Earth’s worldline shows more correctly as the nearly straight line our planet
follows in the modestly curved spacetime of the Solar System.
Curved spacetime, geodesics, matter and energy—in general relativity
we have a radically different understanding of gravity, an understanding
that arose almost single-mindedly in the brain of Albert Einstein.
Philosophically, Einstein’s theory of gravity is light-years different from
Newton’s. What does it tell us physically about our Universe? We’ll go
there in the next chapter.
CHAPTER 15
Errant Orbits
By the nineteenth century, Newtonian gravitation reigned supreme
throughout the Solar System. Meticulous calculations of the gravitational
effect not only of the Sun but also of the other planets had accounted for
nearly every detail of the observed motions of the planets. But there was
one tiny discrepancy in the motion of the planet Mercury, a discrepancy that
remained even when the effects of the other known planets were included.
In an idealized Newtonian universe containing just the Sun and one
planet, the planet’s orbit would be a perfect ellipse that closes back on itself
and repeats exactly forever (Figure 3.2 showed such an orbit). The real
Solar System is more complicated, largely because of the gravitational
effects of the planets on each other. We can subtract out those effects and
ask if, in their absence, a planet’s orbit would be the ideal, closed ellipse.
For Mercury, the answer is “not quite.” Mercury’s ellipse doesn’t quite
close, meaning that the long axis of the ellipse rotates slowly with time, as
shown in Figure 15.1. “Slowly” is the operative word here, for in 100 years
the orbit swings through a mere 43 seconds of angle. You probably don’t
have a good feel for a “second of angle,” but you know a “degree of angle”
as a very small angle indeed; there are 360 degrees in a full circle, 90 in a
right angle. One second of angle is a minuscule 1/3,600 of a degree. So the
orientation of Mercury’s orbit changes by only about 43/3,600 of a degree
in a century, or just over a ten-thousandth of a degree per year! That’s a
pretty small angle, and its measurement is all the more impressive because
Mercury’s orbit is not the obvious ellipse of Figure 15.1, but is nearly
circular. Nevertheless, astronomical observations are accurate enough that
nineteenth-century astronomers were confident in their measure of
Mercury’s anomalous orbital motion. They called the phenomenon “the
precession of the perihelion of Mercury,” where precession is a slow
rotation like the change in the orientation of a spinning top and perihelion is
the point on the orbit where the planet is closest to the Sun. You can see
from Figure 15.1 that precession of the orbital axis carries with it that point
of closest approach.
Fig. 15.1 Orbital precession. In Newtonian gravitation, a planet’s elliptical orbit would repeat forever
(thick ellipse),but general relativity predicts that the orientation of the orbit should change with time
—an effect observed in Mercury’s orbit.The precession shown here is highly exaggerated, and
Mercury’s orbit is also much closer to circular.
An Astrophysical Interlude
Observed differences between Newtonian gravitation and general relativity
remained mostly small, marginal effects through the first two-thirds of the
twentieth century. We’ll soon look at some of the other classical tests that
affirmed Einstein’s theory using very sensitive measurements on Earth or
elsewhere in our Solar System. But in the 1960s things began to change.
Astrophysicists discovered amazing objects whose gravity was really strong
in the sense I’ll define shortly. The first such objects discovered were stars
that, having expended the nuclear fuel that kept them shining, had collapsed
under the influence of their own gravity to form one of two remarkable
stellar endpoints. Neutron stars, the somewhat less bizarre of these two
“dead star” possibilities, pack roughly the mass of the Sun into a region
only a few miles across. Under the resulting intense gravity, electrons and
protons collapse together to make neutrons, hence “neutron star.” The
neutron star is somewhat like a giant atomic nucleus, so dense that a
teaspoon of neutron-star matter would weigh more than 100 million tons.
An unusual kind of pressure arising from all those neutrons keeps the star
from collapsing further, but there’s an upper limit to the mass neutrons can
support. Above this limit, the burned-out star must collapse completely,
forming a black hole. Much more on black holes later—they’re in this
chapter’s title—but for now suffice it to say that massive stars collapse at
the ends of their lives to form either neutron stars or black holes. In both
cases, gravity in the immediate vicinity of the collapsed star is most
definitely strong. (By the way, our Sun is not sufficiently massive for such a
dramatic fate; it will eventually become a white dwarf star, with its mass
packed into roughly the size of the Earth.)
It was in the 1930s that physicists first used general relativity to predict
the theoretical possibility of neutron stars and black holes, but it wasn’t
until the 1960s that these remarkable objects were actually discovered. I
won’t go into the details of how we detect neutron stars and black holes,
except to note that they often occur in binary-star systems and are observed
through their effect on a companion star. (You may recall from Chapter 5,
where observations of binary stars showed that the speed of light does not
depend on the speed of its source, that about half the stars in our galaxy are
in binary systems.) In 1974 astrophysicists Joseph Taylor and Russell Hulse
made a remarkable discovery: a binary system containing not one but two
neutron stars in very close orbits. Each has about 50 percent more mass
than our Sun and their orbits are decidedly elliptical. At closest approach,
they’re only about half the Sun’s diameter apart. With such tight orbits,
each completes a full orbit of the other in a mere 8 days. The gravity each
neutron star experiences is much stronger than anything in our Solar
System, and the effects of general relativity are much more pronounced.
Thus this unusual binary system provides a laboratory for studying general
relativity.
How do we know so much about the Taylor–Hulse binary system?
Because neutron stars, in addition to being dense, often rotate at high
speeds. This rapid rotation occurs for the same reason that a figure skater or
ballet dancer goes into a rapid spin by pulling in his or her arms. As a large
star collapses to a neutron star, its initially slow spin gets amplified as the
matter of the star “pulls in” to occupy a much smaller space. (The physics
term for this is “conservation of angular momentum.”) Furthermore,
neutron stars have strong magnetic fields, and these fields channel outgoing
electromagnetic radiation—light, radio waves, x-rays—into narrow beams
that sweep through the cosmos as the star rotates. When a beam sweeps by
Earth, our telescopes or x-ray detectors record a pulse of radiation. As the
neutron star spins, we get one such pulse for each rotation; for that reason
we call these spinning neutron stars pulsars. By timing the pulses, we can
measure the spin rate with amazing accuracy; in the Taylor–Hulse system,
for example, the spin rate of one neutron star was measured to be
16.940539184253 rotations per second. (This measurement was made in
1986, and the rate has changed since then—more on this later!)
Now, as I noted in Chapter 5, the motion of stars in a binary system
toward or away from us changes the frequency or color (but not the speed!)
of the light we receive. Similarly, the pulse rate we measure depends on
whether a pulsar is moving toward or away from us. (This phenomenon,
called the Doppler effect, also occurs for sound and is probably familiar to
you. For example, when you stand near a highway, you hear a high-pitched
sound as a truck approaches, then a low-pitched sound as the truck passes
and heads away from you: “aaaaaaaaaeeeiiiooooooooo.” The high pitch
occurs because wave crests pass you more often as the truck approaches;
the low pitch because they pass less often as the truck recedes. The same
idea holds for light-wave crests or pulsar pulses.) So by timing a pulsar’s
pulses, we can keep track of its orbital motion. Even though the system is
way too distant for us to see the individual neutron stars, pulse timing
allows us to construct a picture of the orbital motion. Because the pulse rate
is known with such precision, that picture is very accurate.
What’s all this got to do with general relativity? With objects as massive
and as close as those in the binary pulsar, general relativity predicts a
precession of the orbital axis by more than 4 degrees per year, some 35,000
times Mercury’s paltry 43 seconds of angle per century. Given our accurate
picture of the pulsar’s orbit, derived from timing its pulses, the orbital
precession is obvious and easy to measure. The result, from more than 20
years of observations, is a precession that’s right in line with the general
relativistic prediction.
So here’s one way in which general relativity and Newtonian gravitation
differ: elliptical orbits don’t quite close, but precess at a rate that depends
on how strong gravity is at the location of the orbit. In our Solar System,
gravity is so weak that this precession is barely noticeable. Elsewhere in the
cosmos, though, nature provides remarkable systems—here the binary
pulsar—where general relativity’s deviation from Newtonian physics is
much more obvious. That will be the story of the rest of this chapter: we
identify a general relativistic effect and confirm it early on with a very
sensitive terrestrial or Solar System–based experiment. Later, toward the
end of the twentieth century, astrophysics provides much more dramatic
confirmation of the same effect. Today, in the twenty-first century, general
relativity is a solidly verified theory, a working tool of many
astrophysicists.
Light Bends!
In the previous chapter I introduced the Principle of Equivalence as the
heart of general relativity. This principle says that no experiment can
distinguish the state of free float—motion under the influence of gravity
alone—from truly uniform motion in the complete absence of gravity.
Figure 14.2 made that point very clear for a simple mechanical experiment
like dropping a ball. Einstein extended the equivalence principle to all of
physics, including electromagnetism and the behavior of light. He then
turned his reasoning around and used the principle to deduce a new physical
phenomenon, namely, the bending of light in the presence of gravity.
Consider once again the little room of Figure 14.2a, way off in
intergalactic space where there’s no gravity whatsoever. Suppose we mount
a flashlight on one wall and shine the light directly across the room. The
light goes straight across the room to make a bright spot on the opposite
wall, directly opposite the flashlight (Figure 15.2a). Now switch to the
situation of Figure 14.2b, where the room was shown falling freely toward
Earth. Einstein, with his Principle of Equivalence, says that this situation is
indistinguishable from that of Figures 14.2a and 15.2a. That means any
experiment we do in the falling room must have exactly the same outcome
that it did way off in intergalactic space. If it didn’t, we would have a way
to distinguish the two situations. So if we turn on the flashlight here, the
light must still hit the wall directly opposite. In either situation, an observer
in the room sees the light go through the room in a straight path to the
opposite wall. Now think about what this looks like to someone on Earth
watching the falling room. To the earthbound observer, light leaves the
flashlight at the left side of the room. By the time it gets to the opposite
wall, the room has fallen some distance, yet the light still hits the wall right
opposite its source. How can that be? Only if the light, too, has “fallen.”
That is, the light itself must describe a curved path, as suggested in Figure
15.2b.
Fig. 15.2 (a) Light shined across the little room in gravity-free space hits the opposite wall. (b) The
freely falling room is equivalent, so here the light should hit the opposite wall the same distance
below the ceiling as it did in (a). But the room is accelerating downward,so the light—s path must
bend. The room is shown twice, first when the light departs the left wall and again (lighter image)
when it hits the right wall.
Einstein’s conclusion that gravity must affect light shows the power of
the equivalence principle in deducing new physical phenomena. Later,
when he had formulated the full general theory, Einstein had another
explanation: light, like matter, moves in the straightest possible paths
through spacetime. But because spacetime is curved near massive objects,
those paths are not the true straight lines of tenth-grade geometry.
Incidentally, Einstein’s final form of general relativity predicts a bending of
light twice as great as his earlier equivalence-principle argument would
imply.
So the path of a light ray should bend as the light passes a massive
body. To test this idea, astronomers sought to measure the bending of light
passing close to the most massive body in our Solar System—the Sun. The
idea was to view one or more stars at a time when their light had to pass
close to the Sun on its way to Earth (Figure 15.3). Astronomers could then
measure changes in the stars’ apparent positions as compared with
observations made when the Sun was not in the same general direction as
the stars. From those changes they could calculate the angle by which the
Sun’s gravity had deflected the light.
There’s a problem with this approach: To view stars that appear in the
sky near the Sun, one has to observe the stars in the daytime, but stars aren’t
visible in the bright daytime sky. So the only way to make this observation
is during a total eclipse of the Sun. Ideally, one would like an eclipse that
occurred at a time when many bright stars would appear near the Sun. Now,
Einstein made his final calculation of the bending of light in 1916.
Fortunately, an eclipse was to occur on May 29, 1919, and on that date it
happens that many bright stars appear near the Sun. So the 1919 eclipse
would provide an excellent test of Einstein’s prediction.
Fig. 15.3 (a) Light from a distant star comes directly to Earth. (b) When the Sun and star are in the
same region of the sky, the Sun’s gravity deflects the starlight. Observers on Earth then see the star at
a different apparent position in the sky. Dashedlines mark the directions observers would look to see
the star. The deflection shown is greatly exaggerated.
Gravitational Lenses
Like the precession of Mercury’s perihelion, the bending of light remained
through most of the twentieth century an obscure and very subtle effect
significant only in helping confirm the general theory of relativity. Einstein
himself saw greater possibilities, suggesting that massive bodies deep in
space might provide much more dramatic bending of light—enough to
produce distorted and even multiple images of more distant objects. Figure
15.4 shows how this works. Here, a massive galaxy lies between Earth and
a quasar. Quasars are distant objects so bright they can outshine an entire
galaxy, yet small enough that they appear as pointlike sources of light (more
on quasars later). Light from the quasar bends so much in passing the
galaxy that an observer on Earth can see the quasar by looking in either of
two directions—above or below the galaxy in Figure 15.4. The result is two
images of the same object! In a telescopic photograph, these appear as two
distinct objects at different positions on the photo. Actually, things get more
complicated; in three dimensions the situation of Figure 15.4 would result
in the quasar’s image being smeared into a circular ring, provided the
quasar were directly behind the galaxy and the galaxy perfectly circular.
Absent this perfectly symmetric situation, the usual result would be several
images of the same quasar.
Fig. 15.4 Gravitational lensing. A massive galaxy lies between Earth and a distant quasar. Light from
the quasar takes several paths around the galaxy, resulting in multiple images of the quasar as viewed
from Earth. Dashed lines show two different directions in which the quasar is visible, resulting in two
distinct quasar images. Additional images,or a continuous ring, would appear in three dimensions.
It was not until the 1970s that astronomers firmly identified such
gravitational lenses. By now, numerous examples of galaxies lensing
quasars have been discovered. One of the best known is the “Einstein
cross,” a constellation of four images of the same quasar, shown in Figure
15.5a. In other examples, distant galaxies appear smeared out and distorted
as their light passes a nearer galaxy (Figure 15.5b). Study of these images
gives information about the lensing object as well as the distant galaxies.
Fig. 15.5 Hubble Space Telescope images showing gravitational lensing. (a) The Einstein Cross
includes four images of the same quasar, gravitationally lensed by a massive galaxy. The central
bright spot is the nucleus of the galaxy, with the four quasar images around it. Image quality is
limited by the resolution of the Hubble Telescope. (b) Here an entire cluster of galaxies acts as a
gravitational lens for more distant galaxies, producing distorted, arclike images. Some of these are
multiple images of the same distant galaxy.The lensing galaxies are visible as the larger, brighter
objects. (Credits: (a) NASA, J.Westphal, W. Keel; (b) NASA, A. Fruchter, and the ERO team [STScI,
ST-ECF].)
Warping Time
In general relativity, gravity is the geometry of spacetime. So it should
come as no surprise that time itself is affected by the presence of massive
objects that warp spacetime. Once again it was the equivalence principle
that led Einstein to this new phenomenon. He imagined the situation of
Figure 15.2, but this time with the flashlight shining from floor to ceiling.
In the frame of Figure 15.2a, with gravity absent, the light reaches the
ceiling unchanged; in particular, it’s the same color as it was at its source.
According to the equivalence principle, that must also be the case in the
equivalent freely falling frame of Figure 15.2b. To an observer on Earth,
though, the room is accelerating downward as the light makes its journey
from floor to ceiling. So when the light reaches the ceiling, the room is
moving faster than it was when the light was emitted. That means an
observer at the ceiling should see a Doppler shift, toward higher frequency
and bluer color. But the equivalence principle says that doesn’t happen, so
there must be a redshift, associated with the presence of gravity, that cancels
the blueshift. Observers who aren’t accelerating downward would still see
this redshift even though there would be no blueshift. So if I’m above Earth
or some other gravitating body, and I observe light emitted at that body,
then when I receive the light it’s redshifted—that is, it has a lower
frequency—compared to what an observer right at the light source would
see. Gravitational redshift is one name for this phenomenon, but a deeper
name is gravitational time dilation.
Why “time dilation”? Because the vibrations of a light wave, like any
other periodically repeating phenomenon, provide a measure of time. In
fact, our most accurate clocks—the atomic clocks that establish world time
standards—use as their ticks the frequency of light waves emitted by
particular atoms. So in the frequency and wavelength of light waves we
have a measure of time itself. If I’m looking down at a source of light, the
light reaching me has a lower frequency than it would if I were right beside
the source. That is, the time interval between crests of the light wave is for
me stretched out. But light frequency is simply a measure of the underlying
passage of time; therefore, time intervals between any events occurring
down at the light source also appear to me stretched out. That includes the
interval between ticks of an ordinary clock; in other words, if I look down
on a clock, I see it running slow compared to a clock right next to me.
That’s gravitational time dilation. As with the time-dilation phenomenon we
encountered in special relativity, gravitational time-dilation isn’t about light
or clocks—it’s about time itself. Time runs at different rates depending on
where you are in relation to a massive gravitating object.
Is the lower clock really running slow? If I go down there and stand
next to it, its timekeeping will seem perfectly normal. Of course; the clock
is in a perfectly legitimate frame of reference and so it works normally. But
compared with the clocks of an observer located higher up, the lower clock
really does run slow. This time-dilation effect, unlike the time dilation of
special relativity, is not reciprocal. That is, a clock lower down really does
keep time more slowly compared with one higher up, and the clock higher
up really is running fast compared with the lower clock.
Does gravitational time dilation actually happen? It does and it’s been
measured. Here on Earth the effect, like all manifestations of general
relativity, is subtle and difficult to detect. However, in a famous 1960
experiment, physicists managed to measure the difference in timekeeping
rates of clocks at the top and bottom of a 74-foot tower at Harvard
University. Their clocks were atomic nuclei emitting radiation whose
changes in wavelength could be detected with exquisite precision. The
result—a shift of only about a thousandth of a trillionth of the original
wavelength—verified general relativistic time dilation for the weak gravity
at Earth’s surface. In the 1971 experiment I introduced in Chapter 1,
scientists flew atomic clocks around the world and compared their
timekeeping with an atomic clock left behind. The combination of special
relativistic time dilation associated with relative motion and reduced
gravitational time dilation from higher altitude were fully consistent with
relativity. (See “Around the World Atomic Clocks,” in the Further
Readings, for details of this experiment.) Today, the Global Positioning
System (GPS) times signals from a constellation of orbiting satellites to
provide precise locations anywhere on Earth. So accurate is GPS that if the
satellites’ atomic clock times weren’t corrected for gravitational time
dilation, the system would soon be off by a matter of miles!
As always with general relativistic effects, it’s in the astrophysical realm
that gravitational time dilation is most obvious. Even the Sun’s weak
gravity produces a measurable effect. Decades before the earthbound
Harvard experiment, astronomers had measured the gravitational redshift of
light from a white dwarf star, which boasts the Sun’s mass crammed into
the size of the Earth. At the surface of such a dense object, gravitational
time dilation is some 30 times greater than that of the Sun. Finally, today’s
astrophysicists routinely observe substantial time dilation in the strong
gravity around neutron stars and their even more bizarre cousins, the black
holes.
Black Holes
What goes up must come down, right? No! Throw a ball straight up as hard
as you can and it eventually slows, stops, and returns. If you could throw it
fast enough—for an object thrown from Earth’s surface, “fast enough” is
about 7 miles per second—the ball would have enough energy to escape
Earth’s gravity altogether and would travel outward forever without
stopping. That’s because gravity weakens so rapidly with increasing
distance that escape to an infinitely great distance does not require infinite
energy. The 7 miles per second you’d need is called the escape speed for
Earth’s surface. Although the human arm can’t propel anything at escape
speed, rockets can. Spacecraft traveling to the outer planets, for example,
leave Earth’s vicinity at greater than escape speed. Pioneer and Voyager
spacecraft even exceed escape speed for the Sun, meaning they’ll
eventually leave our Solar System and spend the foreseeable future drifting
through the galaxy.
What determines escape speed from a given location? Ultimately, it’s
the strength of gravity at that point. For the surface of a planet or star, that’s
set by the mass and size of the object. Cram more mass into a given-size
object and escape speed goes up. Shrink an object of fixed mass and again
escape speed goes up. So imagine compressing Earth to ever-smaller sizes.
Escape speed from the surface of the shrinking planet rises from its current
7 miles a second to ever-higher values. It gets harder and harder to “throw”
a spacecraft or other object forever outward, but with sufficient energy and
advanced technology, it remains possible. Possible, that is, until escape
speed reaches the ultimate value, namely, the speed of light. For Earth, that
would happen when the entire planet was a little under an inch in diameter.
That’s right—you, me, Mount Everest, New York City, all the water in the
oceans, all the continents, the liquid and solid cores of the planet—all
crammed into a space smaller than a Ping-Pong ball. If this seemingly
impossible compression occurred, then we would have an object so dense
that not even light could escape. That’s a black hole.
Our incredible shrinking Earth, compacting until its escape speed
approaches the speed of light, finally provides a solid definition of the terms
“strong” and “weak” gravity that I’ve been using throughout this chapter.
Strong gravity exists where escape speed is an appreciable fraction of the
speed of light, c. Weak gravity means escape speed is far less than c. Earth’s
and Sun’s escape speeds, at 7 and 380 miles per second, respectively, are far
less than the 186,000-mile-per-second speed of light. Gravity everywhere in
our Solar System is weak. At the surface of a typical neutron star, however,
escape speed is about two-thirds that of light. This is strong gravity! Shrink
that neutron star even a little bit and it will collapse to a black hole with
escape speed c—the ultimate in strong gravity.
A black hole is a remarkable object. Our current understanding of
physics suggests that once an object has been squeezed to black-hole size,
there’s no force in the Universe that can prevent its further collapse to a
single point of infinite density. This infinite conclusion may change
somewhat when we finally learn how to merge general relativity with
quantum physics, the theory that describes matter at the atomic and
subatomic scales. Even so, black holes will remain objects in which matter
is compressed to a near point of incredible density. Surrounding this point is
a spherical surface called the event horizon, which bounds the region within
which the escape speed exceeds the speed of light. No light can escape from
within this region, making it a true horizon. Those of us on the outside can
never see in, past the horizon. There’s simply no way for us to get
information about events occurring within the horizon, hence the name
event horizon.
Because light can’t escape a black hole, and since no material object can
go faster than light, that means nothing whatsoever can escape the hole.
That fact makes black holes remarkably simple objects. From the outside,
black holes exhibit very few distinguishing properties. Most significant is
their mass—the total amount of matter and energy that has fallen across the
horizon. It doesn’t matter whether that mass-energy was in the form of
stars, planets, people, interstellar dust, mice, water, light, or whatever. Once
it’s across the horizon, we can’t know anything about it and so all that
matters is the total mass. That mass determines the size of the event horizon
and the gravitational influence the hole has on the surrounding Universe. If
the infalling matter has electric charge, the black hole, too, will be charged,
and the charge will be felt outside the horizon. If the infalling matter has
rotational motion, the black hole will itself be spinning in a way that
influences spacetime outside the horizon. But that’s it: mass, spin, and
electric charge are the only properties that distinguish black holes.
People often picture a black hole as sucking up all the matter in its
vicinity. That’s a misconception, because a black hole’s gravitational
influence is the same as that of any other object with the same mass. Far
from the hole, matter will orbit in essentially Newtonian elliptical orbits
determined by the hole’s mass alone. If Earth suddenly collapsed to a black
hole, for example, the Moon would be completely unaffected and would
continue in its orbit about the Earth-mass black hole. It would not suddenly
be sucked in any more than the Moon or a satellite is sucked to Earth by the
planet’s gravity. The only objects that strike Earth are those that are on a
collision course with our planet or are close enough that Earth’s gravity
deflects them toward a collision. The same is true with a black hole; only
matter that comes very close to the event horizon actually falls through it
and because typical event horizons are very small, such a course is quite
improbable. That means an isolated black hole will swallow matter at a
rather low rate. On the other hand, a hole surrounded by a dense
aggregation of matter—as in a binary star system or near a galactic center—
will generate a substantial inflow of matter. More on this when we consider
real black holes out there in the cosmos.
The notion of a black hole behaving like a cosmic vacuum cleaner does,
however, have some merit. That’s because the event horizon is a one-way
street; matter that crosses the horizon can never re-emerge. So a black hole
only grows in mass. Again, it doesn’t do so by inexorably pulling in
everything around it; rather, whatever happens to fall into the hole simply
doesn’t get out.
Actually, even that conclusion has to be tempered. In a remarkable
quantum-physics process first envisioned by Stephen Hawking, black holes
can actually lose mass by evaporation involving particles created in the
vacuum just outside the event horizon. For astronomical-size black holes,
this process is so feebly slow as to be completely negligible, but it might
play a role in the very long-term evolution of the Universe.
Fig. 15.6 What a black hole in a binary system might look like. The massivestar in the center is
distorted by the strong tidal forces of the black hole. Gas from the star is drawn to the vicinity of the
hole, where it swirls around in a disk before finally disappearing into the hole itself. Friction in the
gas generates such high temperatures that the gasproduces copious x-ray emission.
Ripples in Spacetime
Drop a rock into a pond and circular ripples spread across the water. Only
later, when the ripples have reached it, can a distant point on the pond
“know” about the rock’s plunge. Disturb spacetime, perhaps through a
violent collision between black holes, and what happens? Special relativity
assures us that distant points can’t know instantaneously about this event;
indeed, that’s one of the reasons Einstein knew that Newton’s theory of
gravity couldn’t be right. As on the pond, “ripples” in spacetime itself
propagate outward, carrying information about the violent event at their
center. What’s a ripple in spacetime? Simply a change in the curvature of
spacetime—a change that moves outward from the disturbance that initiates
it. Einstein’s general relativity predicts the existence of these ripples in
spacetime; they’re called gravitational waves. General relativity also shows
that gravitational waves travel at the familiar speed of light, c.
Gravitational waves are an entirely new phenomenon predicted by
general relativity. Detecting these waves would provide an independent
confirmation of Einstein’s theory, and might give us a novel window on the
cosmos. How can we detect them? With their peaks and troughs,
gravitational waves would stretch and compress spacetime itself. We could
detect the spatial part of that stretch and compression by measuring the
associated motion of physical objects. Because mass is what responds to
gravity—that is, to spacetime curvature—we’ll have better luck with
massive objects. For some decades researchers around the world have
attempted to detect gravitational waves by using huge aluminum bars
equipped with exquisitely sensitive motion detectors. The bars would be set
into vibration by a passing gravitational wave. To eliminate vibrations
induced by trucks, scientists walking by, and other mundane causes, a
typical experiment involves identical setups located thousands of miles
apart. The only events considered real candidates for gravitational waves
are those that trigger both detectors. Although these experiments have
produced a few intriguing signals, none to date has passed muster as a true
gravitational-wave detection.
This may all change soon, however, as a new generation of
gravitational-wave detectors becomes operational. Abandoning the massive
cylinders of first-generation detectors, gravitational-wave researchers are
now turning to interferometry—the method pioneered by Michelson in his
famous experiment with Morley—to measure precisely the distance
between two widely separated objects. Recall from Chapter 6 that a slight
change in the travel time for light along one arm of the Michelson–Morley
apparatus (shown in Figure 6.2) would result in a shift in the observed
interference pattern. Michelson and Morley hoped to find changes
associated with differences in the speed of light; for gravitational-wave
detection, we’re looking for changes in the distance from beam splitter to
mirror as a spacetime ripple goes by.
In the United States, the Laser Interferometer Gravitational-Wave
Observatory (LIGO) consists of two complete interferometers, one in
Washington state and one in Louisiana. Each consists of a Michelson-type
apparatus with two perpendicular arms 2.5 miles long. These instruments
can measure changes in the lengths of these 2.5-mile paths of less than a
trillionth the diameter of a human hair, and with that sensitivity they should
be able to detect gravitational waves produced in the supernova explosions
that result in neutron stars and black holes; in collisions of black holes and
neutron stars; and in the Big Bang explosion that began our Universe.
Similar detectors are being built around the world, and collectively they
will give astrophysicists a new type of “telescope” for observing hitherto
unseen events in the cosmos.
Even more ambitious is the proposed Laser Interferometer Space
Antenna (LISA), a Michelson-type apparatus whose “arms” will consist of
spacecraft forming a triangle 3 million miles on a side (that’s 400 times
Earth’s diameter, or 3 percent of the Earth–Sun distance!). Sensitive to
changes in that distance on the order of a billionth of an inch, LISA should
“see” gravitational waves from beyond our galaxy, including those
generated by the massive black holes at the centers of other galaxies.
Although gravitational waves have yet to be detected directly,
astrophysicists nevertheless have one piece of convincing evidence for their
existence. This is the Taylor–Hulse binary pulsar, which I introduced early
in the chapter for its obvious general-relativistic orbit precession. Recall
that the orbital period of this binary neutron star admits very precise
measurement and that this period is changing slowly with time. Why
changing? Because the neutron stars’ orbits are shrinking. Why shrinking?
Because the stars are losing energy, gradually spiraling closer just as a
satellite in orbit near Earth slowly loses altitude through friction with the
upper atmosphere. But there’s no atmospheric friction in the binary pulsar.
Instead, the neutron stars lose energy through an unseen process that leaves
its fingerprint in the slowly decaying orbital motion. That process is the
generation of gravitational waves. As the massive neutron stars swing round
in their close orbit, the energy they expend in disturbing spacetime—energy
that’s carried away as the yet-undetected ripples of gravitational waves—is
lost from their orbital motion. Years of meticulous observation show that
the binary pulsar is losing energy at just the rate that general relativity
predicts. So although we haven’t “seen” gravitational waves from the
binary pulsar, we’re quite sure they’re being produced and are responsible
for the observed orbital changes. Incidentally, their meticulous observations
of the binary pulsar and its general relativistic implications earned Taylor
and Hulse the 1993 Nobel Prize in Physics.
EINSTEIN’S UNIVERSE
• • •
Einstein’s Blunder
Einstein himself was among the first to apply general relativity to
cosmology, and what he found was unsettling. The Universe, according to
the simplest formulation of general relativity, couldn’t be static; it had to be
expanding or contracting. But prevailing wisdom held that the Universe was
static, having existed forever unchanged in its overall features. Now, in the
mathematical development of general relativity, there arose a number called
the cosmological constant, whose value seemed to be arbitrary. Absent any
reason to the contrary, the most sensible choice is to set this number to zero,
giving the simplest formulation of the theory. That choice is what made
Einstein’s universe expand or contract, so he introduced a nonzero
cosmological constant of just the right value to keep the Universe static.
Einstein’s cosmological constant represented a sort of repulsive force acting
on the largest scales, preventing the Universe from collapsing under its own
gravity. Nothing then known about gravity suggested such a repulsion, so
Einstein was a little uneasy with his cosmological constant. But it was not
inconsistent with the general theory, so he accepted the cosmological
constant as necessary to make his theory fit what seemed to be the real
Universe.
Other physicists also explored relativity-based models for the Universe.
In the early 1920s, the Russian Aleksandr Aleksandrovich Friedmann found
solutions to the equations of general relativity that described an evolving
universe beginning in a very dense state and then expanding at an ever
slowing rate. Friedmann’s results revealed two distinct kinds of possible
universes: those in which expansion continues forever, albeit at an ever
slower rate; and those in which the expansion eventually halts and the
universe subsequently contracts to an eventual high-density crunch. These
two possible fates are intimately linked to the overall spacetime geometry
of the Universe. A forever expanding Friedmann universe has negative
curvature, meaning its spacetime is shaped like the four-dimensional analog
of a saddle or a pass between mountain peaks, and is infinite in extent (see
Figure 16.1a). A universe that eventually collapses has positive curvature,
like the surface of a sphere (Figure 16.1b). It’s closed back on itself and is
finite in extent. Yet, like a sphere’s surface, it has no edge. Between these
two types of Friedmann universe is a dividing case. Overall, it’s flat
(although spacetime in such a universe would still be curved locally in the
vicinity of massive objects), infinite, and will just barely expand forever.
Fig. 16.1 (a) A universe with negative curvature would be analogous to thissaddleshaped surface in
three dimensions. (b) A positively curved universe would be analogous to a three-dimensional
sphere.
A Theory of Everything?
General relativity provides our best understanding of the Big Picture of the
Universe as a whole, but can general relativity tell us everything? The
answer is no. That’s because physicists have not yet been able to reconcile
general relativity with quantum physics, the theory that describes matter on
atomic and smaller scales. Special relativity and quantum physics were
reconciled decades ago, giving us a powerfully accurate description of the
behavior of matter at small scales that reveals some entirely new atomic
phenomena required by the Principle of Relativity. A reconciliation of
general relativity and quantum physics, though, faces deep conceptual
problems. That’s because the essence of quantum physics is quantization,
meaning that the “stuff” of the Universe—from particles of matter to energy
itself—comes in discrete “chunks” rather than being continuously
subdividable. You can have one electron, but not half of one. You can have
a “chunk” of light energy of a given color (called a photon), but you can’t
have less. It’s this essential graininess that ultimately dictates the strange
rules governing the quantum world.
We, and the things we interact with in everyday life, are so large that we
don’t notice quantization. A glass of water contains so many individual
water molecules that the quantization of the water into molecules doesn’t
seem to make a difference; the water might just as well be a continuous
fluid. A light bulb or even a candle flame emits so many photons that they
might as well constitute a continuous stream of energy. All this is even
more true for the planets, stars, and galaxies that make up the astrophysical
Universe. As a practical matter, astrophysics’ description of the Universe is
at such a large scale that the reconciliation of quantum physics and general
relativity is usually unimportant. Put another way, the curvature of
spacetime is generally significant only on scales vastly larger than the size
of atoms or elementary particles. But we can imagine situations where this
is not true, situations where spacetime is so tightly curved that even
something as small as an elementary particle is big enough to experience
spacetime curvature. What absurd situations would those be? One is the
singularity at the center of a black hole. There, general relativity predicts
that spacetime curvature becomes infinitely sharp. Before that true
singularity is reached, effects of quantum physics must come into play.
Another example is the very early Universe, at the start of the Big Bang—
specifically, the time before about 10–43 of a second from the beginning.
(That’s 1/10 000 000 000 000 000 000 000 000 000 000 000 000 000 000 of
a second, a time known as the Planck time.) At that point the Universe was
so dense that even its geometrical structure would have to be described
using quantum physics. Because we don’t know how to reconcile quantum
physics and general relativity, we can’t say much about conditions at the
center of a black hole or in the early Universe before the Planck time.
Ultimately, our knowledge is incomplete.
Why are the two pillars of modern physics—general relativity and
quantum physics—so seemingly irreconcilable? Because general relativity
is, at its heart, a continuous theory. It envisions a spacetime that is
continuously divisible. That means we can divide a meter or an inch into
ever smaller lengths, without limit. Similarly, we can divide the 1-second
interval between ticks of a clock into as many ever tinier sub-intervals as
we wish. For time intervals much longer than the incredibly tiny Planck
time introduced in the previous paragraph and for distances much longer
than the Planck length—or the distance that light travels in one Planck time
(about 10–35 of a meter)—the idea of a continuous spacetime is a very
good approximation to reality. As we approach the Planck length and time,
though, spacetime itself must exhibit quantization. There must be
fundamental, indivisible units of length and time on the order of the Planck
units. What does this mean? No one knows for sure. Because the quantum
world is seething with submicroscopic events, physicists often picture
quantized spacetime as a spongelike structure with a complicated and
everchanging geometry, in which wormholes and bridges continually form
and dissolve.
Will we ever reconcile general relativity and quantum physics,
producing a successful theory of quantum gravity? Probably. Some
physicists believe we’re close, with a group of theories collectively called
string theory or M-theory, which envision the fundamental entities of nature
not as particles but as tiny looplike strings. Vibrations of these strings
represent the different elementary particles of physics. Because the strings
have finite size, even though they’re indivisible, they manage to sidestep
problems with gravity at the Planck scales. Furthermore, some versions of
the theory suggest particles called gravitons that would quantize gravity in
the same way the photon is a quantized chunk of electromagnetic energy.
No string theory yet conforms to the observed set of elementary particles
and all require a remarkable complication—the inclusion of at least six
more dimensions in addition to the familiar three dimensions of space and
one of time. Furthermore, no one has yet figured out how to test
experimentally most predictions of string theories. For some physicists,
string research is nevertheless getting us close to a “Theory of Everything,”
explaining the behavior of the entire Universe from the smallest to the
largest scales. To others, string research seems a fruitless mathematical
exercise. Time will tell.
Out of the Black Hole: Inflation, Multiple
Dimensions, and Parallel Universes
Although no one has yet figured out how to merge general relativity and
quantum physics into a consistent theory, we do have some hints of how the
quantum realm might influence relativity. The British physicist Stephen
Hawking—occupant of Newton’s former chair at Cambridge University—
has shown how quantum effects lead to a subtle “glow” of radiation from
the spacetime around a black hole. That radiation ultimately saps the black
hole of its energy—equivalently, by E = mc2, of its mass—causing the hole
to “evaporate” and eventually disappear altogether. For the massive black
holes that astrophysicists have discovered, that process will take far longer
than the present age of the Universe. However, tiny black holes that might
have formed early in the Big Bang would already have vanished through
this mechanism of Hawking radiation.
Quantum physics shows that the emptiness we call vacuum isn’t quite
empty, but is seething with ghostly particle–antiparticle pairs that burst into
a fleeting existence and promptly annihilate. Hawking radiation arises when
one of these particles falls into a nearby black hole, leaving its partner no
one to annihilate with. The lonely partner then looks like a particle that’s
emerged from the hole. Although that isn’t quite the case, the black hole
has, in a sense, created the new particle and thus has given up some of its
own mass-energy.
Not only is the seemingly empty vacuum actually the site of frenetic
physical activity, but the current theory of elementary particle physics—the
so-called standard model—also suggests the possibility of a “false
vacuum,” a high-energy state that might decay rapidly to a true vacuum.
Decay of the false vacuum can lead to exponentially rapid inflation of
spacetime. Detailed theories of the Big Bang suggest that a period of just
such inflation occurred in the first fraction of a second of the Universe.
During this brief instant, the Universe grew in size by many orders of
magnitude—a process that flattened out any curvature and resulted in a
Universe whose large-scale spacetime geometry is essentially the flat,
Euclidean geometry that you learned in tenth grade.
It’s the prospect of inflation that leads to a remarkable possibility I
mentioned in Chapter 1: a baby universe could bud from a parent and then
grow by inflation to become a full-blown universe in its own right.
Repeated endlessly through eternity, this process would give rise to a
complex Multiverse of which our Universe is but one branch.
Attempts to merge relativity with quantum physics often seem to lead us
beyond the three familiar dimensions of space and one of time. Today’s
eleven-dimensional string theories are only the latest instance of
multidimensional theories of everything. Einstein himself spent much of his
later life working, in vain, on five-dimensional theories that he hoped would
combine his general relativity with quantum physics.
Extra dimensions allow for the remarkable possibility that there may be
parallel universes lying very close to ours but separated by an extra
dimension that we don’t directly perceive. Stanford physicist Savas
Dimoupolos and colleagues have argued that a parallel universe might be
only a few millimeters (about a tenth of an inch) away from ours (Figure
16.2). We think we know nothing of it because all processes in that universe
are confined to its three dimensions of space and one of time. But not quite!
According to Dimoupolos, the quantized particles of gravitational influence
—the gravitons—would be unique in that they could traverse the extra
dimension. So, while ordinary matter, light, and other forms of energy
would be confined to that nearby but hidden universe, its gravitational
influence, as Figure 16.2 shows, would not. If this idea is correct, then what
we infer to be dark matter in our Universe may actually be ordinary matter
in the parallel universe! We perceive only its gravitational influence, not the
matter itself, and so of course it seems dark to us. Strange as this parallel-
universe idea is, no one has yet disproved it, and experiments with gravity
over millimeter scales might just reveal that extra dimension.
Fig. 16.2 Parallel universes? Each plane is a two-dimensional analog of a universe with three space
dimensions, populated with galaxies. The planes shown are millions oflightyears in extent, but
they’re separated by a fraction of an inch in an extra dimension that we don’t directly perceive. Thin
arrows represent electromagnetic interactions(e.g., light beams) that are confined to the individual
universes. Gravity (thick arrows) extends into the extra dimension, so each universe feels the other’s
gravitational effect. (Galaxy photos courtesy of NASA.)
Because spacetime can be curved, it’s also possible that the parallel
universe isn’t really a distinct universe but rather a part of ours that’s very
far away in the ordinary dimensions but close in the extra dimension. If
that’s true, then it’s possible to imagine a short wormhole connecting our
neighborhood to the distant realm of the parallel universe. Maybe that’s
what the machine in Carl Sagan’s Contact is all about.
Intuiting Relativity
We’ve now come full circle, back to some of the stranger ideas I introduced
briefly in the first chapter. Those ideas, and other new concepts ranging
from the relativity of simultaneity to black holes and gravitational waves,
become possible when we abandon rigid, absolute space and time and
replace them with the flexible, curving spacetime of Einstein’s relativity.
Einstein’s special and general theories of relativity reveal a Universe far
richer and stranger than anything in our commonsense experience. In this
book I’ve taken you rather thoroughly through the special theory, trying to
convince you of its validity on rigorous logical grounds. Yet I’ve
continually reminded you that you’ll never have a fully natural intuition for
relativity. I don’t either, and I don’t believe Einstein did. That’s because
none of us has experienced in everyday life the conditions that show up the
difference between Einsteinian relativity and Newtonian common sense.
With general relativity I’ve had to leave things a lot murkier. I’ve given
you the briefest motivation for the fundamental underpinnings of the theory,
ultimately leading to the idea of gravity as the geometry of spacetime. In
these final chapters I’ve described some of the consequences of general
relativity—especially significant where gravity is strong—but I haven’t at
this level been able to give a lot of motivation for those consequences, let
alone any kind of intuitive feel for them.
So you can understand relativity logically, and, if you work with it
enough, gain a comfortable familiarity. For the nonscientist that’s possible
with the simpler concepts and mathematics of special relativity, and for the
mathematically brave a thorough understanding of general relativity is
possible too. But in neither case will you have much natural, commonsense
intuition about relativity. Is such intuition at all possible? I believe it would
be if we grew up experiencing the Universe in the full richness that
relativity describes. If, as a baby, you crawled at speeds approaching that of
light, then your common sense would be fully consistent with special
relativity. You would have no misleading notions about absoluteness of
space and time measurements, and reference frame–dependent lengths,
times, and even simultaneity would be the norm for you. You wouldn’t be
surprised when a friend went out for a high-speed jog and came back a
member of a younger generation. In short, special relativity would make
perfect sense to you because it would be part of your regular experience.
If, in addition to being a relativistic crawling baby, you were so large
that your body directly experienced the curvature of spacetime, then the
geometrical nature of gravity would also be intuitively obvious to you. Your
geometry teacher, drawing on a blackboard so big that spacetime curvature
was obvious, would never deceive you with Euclidean nonsense; it would
be obvious that the angles of a triangle add to other than 180 degrees and
that parallel lines might meet (or might not, depending on the sign of the
local spacetime curvature). You would experience directly that these
geometrical effects arose in the presence of matter and energy, and your
understanding of gravity would naturally be that of Einstein. But you’re
small in relation to the spacetime curvature in your neighborhood and you
move slowly relative to things important in your life; thus, for you,
relativity can never seem intuitive. You can, however, grasp the simple
principle at the heart of all relativity—that motion doesn’t matter, or that
there’s no preferred frame of reference—and despite the failure of your
intuition, you can understand intellectually the remarkable consequences of
this principle and appreciate the wonderfully rich Universe it engenders.
APPENDIX
TIME DILATION
• • •
Here I’m going to use high-school math to convince you that the formulas
for time dilation and length contraction do indeed follow directly from the
Principle of Relativity. You can take those formulas on faith and skip this
appendix, but if you’re a stickler for logical consistency, then reading this
will give you what you need.
C2 = A2 + B2.
You also learned that you can do the same thing to both sides of an equation
and it’s still a valid equation. So let’s take the square root of both sides, to
get the length of the hypotenuse, C:
C = √A2 + B2.
I’ve marked this value on Figure A.1.
Fig. A.1 The Pythagorean theorem gives the hypotenuse of a right triangle in terms of the lengths of
the other two sides.
You also learned, probably even before high school, that distance =
speed × time. After all, that’s just what speed means: go 50 miles per hour
for 2 hours, and you’ve gone 100 miles. Distance equals speed times time.
Time Dilation
Now you’re all equipped to understand time dilation, quantitatively. Figure
A.2 is a modified version of Figure 8.3, the light box that I used to
introduce time dilation. I’ve emphasized the light path by making it a solid
line and dimming everything else. In Figure A.2a the box is again shown in
a reference frame where it’s at rest, and it’s clear that the light goes a round-
trip distance 2L between source, bouncing off the mirror, and returning to
the source. We’ve called the time in this reference frame t’ (“t prime”), and
the light’s speed is, of course, the speed of light, c. So the formula distance
= speed × time becomes
2L = ct’.
Fig. A.2 The light box of Figure 8.3, shown with various quantities used to derive the time-dilation
formula
The total light path is twice this distance, the time the light takes is t, and—
here’s where relativity comes in—the speed of the light is c in this reference
frame as well. So distance = speed × time now gives
1
2 √ /4v2t2 + L2 = ct.
Now, we want to know the time t. Unfortunately it’s on both sides of this
equation and tangled up in a square root. To get it out, square both sides:
1
4( /4v2t2 + L2)= c2t2, or v2t2 + 4L2 = c2t2.
Almost done! Divide both sides by c2 and take the square root of both sides.
The result is
But 2L/c is just the time, t’, measured in the frame at rest with respect to the
box [see Equation (A2)]. So we have
t’ = t√1 − v2/c2.
Aberration of starlight The change in apparent position of a star, due to Earth’s orbital motion.
Used to show that Earth did not drag ether with it.
Acceleration The rate at which an object’s motion changes. Acceleration includes changes in speed
or direction.
Accretion disk The disk-shaped cloud of matter swirling around and into a black hole.
Action at a distance The Newtonian view that influences, especially gravity, reach instantaneously
from a gravitating object to more distant objects. The action-at-a-distance picture is inconsistent
with special relativity.
Antimatter Matter consisting of particles with properties that are exactly the opposite of ordinary
matter. Antielectrons (positrons), for example, are like electrons but with positive charge.
Antimatter is created in energetic interactions involving elementary particles or high-energy
electromagnetic radiation. When they meet, matter and antimatter annihilate in a burst of energy.
Beam splitter A device that splits a beam of light into two beams traveling on different paths. The
simplest beam splitter is a mirror with insufficient reflective material.
Big Bang The cosmic explosion that began the Universe.
Big Crunch In theories that propose an oscillating Universe that expands and then contracts, the Big
Crunch is the ultimate state of contraction before the next expansion. In the Big Crunch, all
matter would be compressed to near-infinite density.
Black hole An object so massive yet so compact that not even light can escape its gravity.
Cosmological constant A number Einstein introduced into the general theory of relativity so the
theory would predict a static Universe. Einstein abandoned the cosmological constant when
observations showed that the Universe was in fact expanding. Discoveries at the end of the
twentieth century suggest the constant might be necessary after all.
Cosmology Study of the origin, evolution, and large-scale structure of the Universe.
Critical density The minimum average density necessary for the Universe to expand forever instead
of eventually collapsing.
Dark matter Unseen material believed to make up over 90 percent of the Universe’s mass.
Doppler effect The increase (or decrease) in the frequency of waves—sound or light—when the
source of the waves is moving toward (or away from) the observer.
Electric charge A fundamental property of matter that is at the basis of all electric and magnetic
interactions.
Electromagnetic induction The phenomenon whereby a changing magnetic field produces an
electric field.
Electromagnetic wave An electromagnetic phenomenon in which changing electric and magnetic
fields continually generate each other, producing a wave of electromagnetism that travels with the
speed of light, c. Electromagnetic waves include radio waves, microwaves, infrared, visible light,
ultraviolet, x-rays, and gamma rays.
Electron A fundamental particle in nature. Carries a negative electric charge and little mass.
Elsewhere The elsewhere of an event consists of those events that cannot influence or be influenced
by the given event.
Energy One of the two manifestations of mass-energy, the fundamental “stuff” of the Universe.
Energy takes many forms, including the energy of motion, gravitational energy, heat energy,
electromagnetic energy, etc. Without energy there would be no motion, no activity whatsoever.
Epicycle Smaller circular paths of the planets, imposed on their larger circular orbits and required in
early cosmological models to account for retrograde motion.
Escape speed The speed necessary for an object to escape to an infinitely great distance from a
gravitating body. At Earth’s surface, escape speed is about 7 miles per second; at the horizon of a
black hole, it’s the speed of light.
Ether Hypothetical substance proposed by nineteenth-century physicists as the medium of which
light waves are a disturbance.
Event Something that happens, characterized by where it occurs in space and when it occurs in time.
Event horizon The region around a black hole at which escape speed becomes the speed of light.
Nothing—no material object, no light, and no information—can escape from within the event
horizon.
Field An invisible influence in the space surrounding a massive object (gravitational field), an
electrically charged object (electric field), or a moving charged object (magnetic field).
Force A push or pull, either by an obvious agent or an invisible influence like gravity, electricity, or
magnetism. In Newtonian physics, forces cause changes in motion.
Frame of reference The surroundings that share one’s state of motion; the physical setting that
establishes one’s point of view for making measurements.
Free fall, free float Terms applied to a situation in which an object moves under the influence of
gravity alone.
Frequency The number of times per second that a periodic occurrence, such as the oscillation of a
wave, takes place.
Future In relativity, the future of an event consists of all those events that the given event can
influence.
Galilean relativity The principle, known since the time of Galileo and Newton, that the laws of
motion provide no way to distinguish different states of uniform motion. In other words, the
question, Am I moving? is meaningless as far as the laws of motion are concerned.
General relativity Einstein’s 1916 theory that describes gravity as the curvature of spacetime.
Geodesic The straightest possible path through a space or spacetime that may itself have curvature.
Great circles are the geodesics on Earth’s surface.
Gravitational lens Any massive body whose gravitation, or spacetime curvature in general relativity,
bends light from more distant objects. The result may be brightened, distorted, or multiple
images.
Gravitational time dilation The phenomenon whereby time passes more slowly near a gravitating
object. Also called gravitational redshift.
Gravitational wave A wave disturbance in the structure of spacetime itself. Gravitational waves
originate in the acceleration of massive objects and propagate at the speed of light.
Hawking radiation A phenomenon associated with quantum physics in the presence of a black hole,
whereby particle–antiparticle pairs come spontaneously into existence just outside the hole. When
one of the pair falls into the hole, the other is left alone and appears as if it were radiation
emerging from the hole. Eventually, Hawking radiation saps black holes of their mass-energy.
Interference A wave phenomenon whereby two waves meeting at the same place simply add. In
constructive interference, the wave disturbances are in the same direction (e.g., crests meet crests,
troughs meet troughs) and the effect is a strengthened wave. In destructive interference, crests
meet troughs and the overall wave is diminished.
Inertial reference frame Any reference frame in which the law of inertia is obeyed. In practice, the
only inertial frames are those in free fall (or free float).
Length contraction The reduction in the length of an object as measured by an observer with respect
to whom the object is moving.
Light A form of electromagnetic radiation visible to the human eye. More loosely, in this book, the
term is often used to designate any electromagnetic radiation.
Light-year The distance light travels in 1 year.
Luxon A particle that travels at exactly the speed of light relative to any uniformly moving reference
frame. The photon, or quantized bundle of electromagnetic wave energy, is one familiar example.
Maxwell’s equations The four equations developed in the 1860s by James Clerk Maxwell that
describe all phenomena of classical electricity and magnetism.
Mechanics The study of motion, one of the major branches of physics.
Medium The substance of which a given wave is a disturbance. Air, for example, is the medium for
sound waves, and water for water waves. Light waves do not require a medium.
Michelson interferometer A device that uses interference of light waves traveling on two
perpendicular paths to detect minute changes in speed or distance between the two paths.
Michelson–Morley experiment The famous 1887 experiment in which Michelson and Morley failed
in their attempt to detect Earth’s motion through the ether, despite using a Michelson
interferometer more than adequate to the task. The failure of this experiment paved the way for
Einstein’s relativity.
Multiverse A multibranched system of multiple universes, proposed by some cosmologists as
representing the overall structure of all that exists.
Muons Subatomic particles, created by cosmic rays high in the atmosphere and heading Earthward at
nearly the speed of light, whose radioactive decay confirms time dilation.
Neutron star An object resulting from the collapse of a massive star and composed almost entirely
of neutrons, with mass about that of the Sun’s crammed into a sphere only a few miles across.
Newton’s laws of motion Three laws describing the relation between motion and force. The first
states that an object not subject to any force continues in uniform motion. The second states that
an object’s acceleration is proportional to the force applied to it and that for a given force the
acceleration is less for a greater mass. The third law states that forces come in pairs; if one object
exerts a force on another, then the second exerts a force of equal strength, but opposite direction,
back on the first object.
Orbit The path described by an object moving under the influence of gravity alone.
Past In relativity, the past of an event consists of all those events that can influence the given event.
Photon A particlelike bundle of energy that, in quantum physics, is associated with electromagnetic
waves.
Planck length The tiny length—about 10-35 meter—at which quantum physics should affect the
nature of space itself.
Planck time The tiny time—about 10-43 second—at which quantum physics should affect the nature
of time itself. To study happenings at the Planck time or length scale will require a successful
merging of quantum physics with general relativity.
Positron Antiparticle to the electron.
Precession The gradual shift in the orientation of a planet’s orbit. General relativity predicts that
precession should occur but Newtonian gravitational theory does not. Precession has been
detected in Mercury’s orbit and in binary pulsars.
Present In relativity there is no such thing as a universal present, which would consist of events near
and far that are simultaneous with what’s happening here and now. Instead, the present of a given
event consists, strictly speaking, of that event alone.
Principle of Equivalence The statement that it is impossible to distinguish the effect of gravity from
acceleration and the absence of gravity from free fall.
Principle of Relativity The statement that the laws of physics do not depend on one’s frame of
reference. Thus absolute motion is a meaningless concept; only relative motion is meaningful.
Proper length The length of an object as measured in a frame of reference in which the object is at
rest.
Pulsar A rapidly spinning neutron star that emits a searchlightlike beam of electromagnetic radiation,
observed as a sequence of pulses as the beam sweeps repeatedly by Earth.
Retrograde motion The occasional reversal of direction that planets undergo when their motion
through the sky is viewed from Earth.
Simultaneous events Events that occur at the same time. In relativity, events that are simultaneous in
one frame of reference need not be simultaneous in other frames that are moving relative to the
first.
Spacetime The union of space and time into a single four-dimensional structure.
Spacetime interval The four-dimensional “distance” between events in spacetime. The interval has
the same value regardless of one’s frame of reference.
Special relativity Einstein’s 1905 theory based on the principle that the laws of physics are the same
in all reference frames in uniform motion. The restriction to uniform motion is what makes this
the special theory.
Standard model The theory that describes the different elementary particles and their interactions.
String theory A contemporary theory that envisions fundamental entities not as particles but as
looplike strings. What we consider elementary particles would be different modes of vibration of
these strings. String theory may have the potential for unifying quantum physics with general
relativity, providing a theory of everything.
Tachyon A hypothetical particle capable of traveling faster than the speed of light, but not at light
speed or slower. Tachyons have not been detected and their existence might wreak havoc on
traditional notions of cause and effect.
Tardyon Any particle that travels at less than the speed of light relative to any uniformly moving
reference frame. All known matter is composed of tardyons.
Time dilation A relativistic effect wherein the time between two events is shorter on a clock present
at both events than it is when measured by two separate clocks in a reference frame where the
events take place at different positions. Time dilation is sometimes described with the phrase
“moving clocks run slow,” but this is relativistically incorrect language for reasons described in
the text.
Twins paradox A phenomenon resulting because time dilation occurs on both the outbound and
return legs of a round-trip journey. A twin who makes such a journey returns to her starting point
younger than her stay-at-home twin.
Brian, Denis. Einstein: A Life. New York: John Wiley & Sons, 1996.
A modern and thoroughly candid account of Einstein’s life, among the first to exploit the Einstein
Archives and Einstein documents made widely available in the 1980s.
Gott, J. Richard. Time Travel in Einstein’s Universe: The Physical Possibilities of Travel through
Time. Boston: Houghton Mifflin, 2001.
Gott, a Princeton physicist, not only elaborates relativity’s clear implication that time travel to the
future is possible but also explores the more speculative prospect of time travel to the past—a
prospect that some physicists find increasingly worthy of serious study.
Greene, Brian. The Elegant Universe: Superstrings, Hidden Dimensions, and the Quest for the
Ultimate Theory. New York: W. W. Norton, 1999.
Columbia University theorist Brian Greene gives a thorough and thoroughly readable account of
string theory, the leading candidate for a “theory of everything” that would unite relativity and
quantum physics. Greene’s own contributions to the field are substantial, and he writes with an
insider’s deep knowledge and yet with the skill of one who can communicate complex ideas to
nonspecialists.
Guth, Alan H. The Inflationary Universe: The Quest for a New Theory of Cosmic Origins. Reading,
MA: Addison Wesley, 1997.
Guth is one of the originators of the inflationary universe idea, which resolves many of the
conundrums raised by the Big Bang theory. This lively, accessible, yet authoritative account of
our modern understanding of the origin of the Universe mixes solid science with the history of
and personalities at the forefront of cosmology.
Hafele, J. C., and R. E. Keating. “Around the World Atomic Clocks: Predicted Relativistic Time
Gains” and “Around the World Atomic Clocks: Observed Relativistic Time Gains.” Science 177
(July 1972): 166–70.
This account of a real “twins” experiment done with highly accurate atomic clocks leaves no
doubt about the reality of time dilation.
Livingston, Dorothy Michelson. The Master of Light. New York: Charles Scribner’s Sons, 1973.
This biography by Michelson’s daughter provides an insightful account of the Michelson–Morley
experiment and its history, but also spares no detail of Michelson’s often turbulent personal and
professional life.
Pais, Abraham. Einstein Lived Here. New York: Oxford University Press, 1994.
A thematic, nonmathematical introduction to Einstein’s life and thought. Author Abraham Pais, a
physicist and Einstein scholar, has written extensively about Einstein’s science and life. His
earlier book, Subtle Is the Lord: The Science and the Life of Albert Einstein (Oxford University
Press, 1982), delves deeply and quantitatively into Einstein’s scientific work.
Paterniti, Michael. Driving Mr. Albert. New York: Dell Publishing, 2000.
Ever wonder what happened to Einstein after his death? His brain, at least, goes on—sloshing
about in a Tupperware container! This amusingly improbable travelogue chronicles the true story
of a cross-country drive with Einstein’s brain and its eccentric steward.
Pyenson, Lewis. The Young Einstein. Boston: Adam Hilger Ltd., 1985.
A scholarly account of Einstein from childhood through the development of both relativity
theories; particularly strong on the cultural and scientific milieu in which Einstein arose and on
scientists’ and mathematicians’ reactions to Einstein’s work.
Smolin, Lee. Three Roads to Quantum Gravity. New York: Basic Books, 2001.
A self-confessed optimist, physicist Lee Smolin believes we are only years away from unifying
general relativity and quantum physics to produce a “theory of everything.” In this lively and
very contemporary book, he outlines three approaches that show promise.
Taylor, Edwin F., and John Archibald Wheeler. Spacetime Physics, 2d ed. New York: W. H. Freeman,
1992.
Written for sophomore-level college physics students, this book presents special relativity in a
lively and entertaining style that emphasizes the underlying simplicity and fundamental principles
of the theory. Some sections of the book are reasonably math-free and provide deep insights into
relativity. If you don’t mind a little algebra, Spacetime Physics will greatly enhance your
understanding of Einstein’s theory.
Thorne, Kip. Black Holes and Time Warps: Einstein’s Outrageous Legacy. New York: W. W. Norton,
1994.
This definitive history of black holes and related phenomena of general relativity will convince
you that black holes really do exist. Written in the first person by a leading relativity researcher,
the book will reward a persevering reader with a really thorough but nonquantitative
understanding of black holes in particular and relativity in general.
Will, Clifford. Was Einstein Right? Putting General Relativity to the Test. Basic Books, A Division of
HarperCollins, 1986.
This highly readable book gives a thorough look at both the classical and fairly contemporary
tests of general relativity. Written for the nonscientist, it’s a fascinating blend of physics,
personalities, and history.
Wambsganss, Joachim. “Gravity’s Kaleidoscope.” Scientific American 285 (November 2001): 64–71.
A nice account of gravitational lenses that shows how they work and the multiple uses to which
astrophysicists are now putting them. Great illustrations, too!
INDEX
• • •
aberration of starlight, 66
acceleration, 30
accretion disk, 223
action at a distance, 45, 175, 193
Ampère, André-Marie, 43
antimatter, 156
antiparticle, 156
Apollo 13, 184
Aristotle, 20
beam splitter, 72
Bell, John, 161
Big Bang, 3, 231
binary neutron star, 207
binary pulsar, 207
gravitational waves from, 227
binary star, 57
containing black hole, 223
Biot, Jean-Baptiste, 43
black hole, 10, 206, 218–25
in binary star system, 223
in galactic centers, 224
properties, 218
Brahe, Tycho, 23
causality, 142–43
clock, 100
atomic, 7, 121, 215
constructive interference, 39
Copernicus, Nicolaus, 13, 22, 63
cosmic microwave background, 233
cosmological constant, 230
cosmology, 229
Coulomb, Charles Augustin, 42
critical density, 232
dark energy, 233
dark matter, 232
destructive interference, 39
Dimoupolos, Savas, 237
Doppler effect, 208
double star, 57
E=mc2, 154
Earth
motion relative to ether, 62
Eddington, Arthur, 211
Einstein, Albert
association with nuclear weapons, 158
childhood years, 78
college years, 79
and EPR paradox, 161
“greatest blunder,” 231
1905 papers, 81
Nobel Prize, 81
electric charge, 42, 171
electric current, 42
electric field, 47
electricity
static, 42
electromagnetic induction, 44
electromagnetic spectrum, 51
electromagnetic wave, 49
speed of, 50
electromagnetism, 43
consistency with special relativity, 175
electron, 44
elsewhere, 143
energy
conversion to matter, 156
equivalent to matter, 155
kinetic, 159
and momentum, 171
in wave, 37
epicycle, 21
EPR paradox, 161
escape speed, 21
ether, 56
properties of, 56
ether drag, 64
Euclidean geometry, 190
event, 96
on spacetime diagram, 145
event horizon, 218
experiment, definition of, 14
Faraday, Michael, 44
field
electric, 47
gravitational, 45
magnetic, 47
Fitzgerald, George, 77
force, 20
frame of reference, 32
inertial, 177
Franklin, Benjamin, 42
free fall, 184
free float, 184
frequency
of electromagnetic wave, 51
Friedmann, Aleksandr, 230
future, 140
Galilean relativity, 32
Galileo, 14, 24
and Tower of Pisa, 177
general theory of relativity, 174
geodesic, 195
geometry
Euclidean, 190
in general relativity, 192
Global Positioning System, 216
Gott, J. Richard, 4
GPS. See global positioning system
gravitational field, 45
gravitational lens, 9, 212–14
gravitational redshift, 215
gravitational time dilation, 215
gravitational waves, 225
detection of, 226
graviton, 160
gravity, 27
Einstein’s view, 192
in general relativity, 192
Newton’s view, 28
strong, 218
weak, 203
great circle, 191
Guth, Alan, 6
Jupiter, 24
Kepler, Johannes, 23
kinetic energy, 159
magnetic field, 47
magnetism, 42
relation to electricity, 172
Manhattan project, 159
Marconi, Guglielmo, 49
Mari´c, Mileva, 79
Mars Rover, 141
mass-energy
conservation of, 156
mass-energy equivalence, 155, 199
matter
conversion to energy, 156
equivalent to energy, 155
Maxwell, James Clerk, 48
Maxwell’s equations, 48
and ether, 60
mechanics, 31
medium, for wave, 36
Mercury, 204
Michelson, Albert, 70
Michelson interferometer, 70
Michelson-Morley experiment, 70–76, 117
apparatus, 72
conceptual description, 70
Einstein’s knowledge of, 70
result, 76
microlensing, 214
microwave oven, 11
Minkowski, Hermann, 168
missing mass, 214
momenergy, 171
Morley, Edward, 75
motion
natural state, 20, 23, 26, 196
relative, 32, 91
retrograde, 21
study of, 19
uniform, 176
Mount Washington, 119
M-theory, 235
Multiverse, 5, 237
muon, 119
neutrino, 160
neutron star, 206
Newton, Isaac, 14, 27
Newtonian gravitation
problems with, 175
Nobel Prize, 81
nuclear power
and E=mc2, 157
observing
versus seeing, 105
Oersted, Hans Christian, 43
orbit
circular, 29
elliptical, 23
of Mercury, 204-5
pair-creation, 156
Parable of the Surveyors, 165
parallel universe, 237
past, 140
perihelion, 205
photon, 160, 234
Planck length, 235
Planck time, 235
Podolsky, Boris, 161
positron, 156
precession, 204
present, 139
Principle of Equivalence, 178–83
principle of Galilean relativity, 32
Principle of Relativity, 81
and c as maximum speed, 149
implications for speed of light, 83
proper length, 116
Ptolemy, 21
pulsar, 207
tachyon, 160
tardyon, 161
Taylor, Edwin, 165
Taylor, Joseph, 207
tennis, 11, 176
Theory of Everything, 236
theory of invariance, 164
theory of relativity
alternate name for, 164
tidal forces, 190
time
commonsense notions, 95
Newton’s views on, 94
relativity of, 94
time dilation, 96–108
applied to space travel, 111
formula, 106
gravitational, 215
mathematical derivation, 241
muon experiment, 119
timelike interval, 168
time loop, 4
time travel, 8, 125–26, 220
Tower of Pisa, 177
twins paradox, 122–25
in general relativity, 197
Venus, 12
wave, 36
electromagnetic, 49
wavelength, 41
of electromagnetic wave, 51
weak gravity, 203
weightlessness, 184
Wheeler, John Archibald, 165
white dwarf, 217
worldline, 148
of Earth, 202
wormhole, 7, 238
Young, Thomas, 39
COPYRIGHT
• • •
Wolfson, Richard.
Simply Einstein : relativity demystified / Richard Wolfson.— 1st ed.
p. cm.
Includes bibliographical references and index.
ISBN 0-393-05154-4 (hardcover)
ISBN 978-0-393-24218-8 (e-book)
1. Relativity (Physics)—Popular works. I Title: Relativity demystified.
II. Title.
QC173.57 .W65 2003
530.11—dc21
2002002984
W. W. Norton & Company, Inc.
500 Fifth Avenue, New York, N.Y. 10110
www.wwnorton.com