Lesson 8 : Baryogenesis

Notes from Prof. Susskind video lectures publicly available

on YouTube

1
Introduction

Baryogenesis means the creation of matter particles. Its

study is more specifically concerned with explaining the
excess of matter over antimatter in the universe.

The conditions which explain the imbalance of particles and

antiparticles are called nowadays the Sakharov conditions 1 .

In the previous chapter, we studied several landmarks in the

temperature of the universe. Going backwads in time, we
met the decoupling time when the universe become trans-
parent. The temperature was about 4000°.

Further back at T ≈ 1010 degrees, photons could create

electron-positron pairs. Protons had already been created
earlier.

Even further back at T ≈ 1012 or 1013 , photons could create

proton-antiproton pairs. More accurately they could create
quark-antiquark pairs, because in truth it was too hot for
protons to exist as such. But, at that temperature, let’s say
by convention that when we talk about a proton we really
talk about three quarks. It does not make any difference in
the reasonings.

1. Named after Andrei Sakharov (1921 - 1989), Russian nuclear

physicist. Sakharov published his work on this in 1967 in Russian
and it remained little known until the 1980s. In the meantime, the
same conditions were rediscovered in 1980 by Savas Dimopoulos and
Leonard Susskind who were trying to answer the question of why
the number of photons, which is a measure of the entropy of the
universe, exceeds by eight orders of magnitude the number of protons
or electrons. Thus for a time the conditions were known in the West
as the Dimopoulos-Susskind conditions.

2
So at that early time, when the temperature was, say, 1013
degrees, the universe was a hot soup of photons, protons,
antiprotons, electrons and antielectrons, also named posi-
trons. Their numbers 2 are denoted respectively

Nγ Np Np̄ Ne Ne+

There was a thermal equilibrium and these five numbers

were of the same order of magnitude.

But they were not exactly equal. There was a slight excess
of protons over antiprotons, and a slight excess of electrons
over positrons. Furthermore, the universe being electrically
neutral, these two excesses exactly compensated each other.
So we had

Np − Np̄ = Ne − Ne+ (1)

As the universe expanded and progressively cooled, the

average energy of photons decreased. First the creation
of proton-antiproton pairs ceased. But since the expansion
and cooling of the universe was relatively slow, the annihi-
lation of protons and antiprotons could continue until there
remained essentially only the excess protons. They are the
protons we see in our universe today.

A little later, when the temperature went below 1010 de-

grees, the same process happened for electrons and posi-
trons. The creation of electron-positron pairs ceased, but

2. Notice that, although we don’t make it explicit with the nota-

tions, these numbers depend on time. When the temperature cooled
Np and Np̄ decreased, but their difference stayed fixed. The same
comment applies to Ne and Ne+ .

3
the annihilation of electrons and positrons continued, even-
tually leaving only the excess electrons. They are the elec-
trons we see today.

Then there was a period during which electrons and pro-

tons began to bind into atoms. But photons were energetic
enough to maintain a part of the electrons and protons sepa-
rated. In other words, the universe was ionized. In turn, the
charged particles were still scattering photons and maintai-
ning a thermal equilibrium.

Finally, when the temperature decreased to about 4000°, de-

coupling took place. Photons could no longer ionize atoms.
The scattering of photons stopped. The universe became
transparent. And the photons were frozen into a blackbody
thermal spectrum, even though they were no longer being
scattered. Moreover they continued to cool as a(t) conti-
nued to increase, but in a separate process from matter
particles.

Over the period from the early soup of photons, protons,

antiprotons, electrons and positrons until today, the num-
bers that did not change are

Nγ Np − Np̄ Ne − Ne+

More accurately, the first stayed about the same order of

magnitude from then until now. And the last two are also
simply Np and Ne of today’s universe.

Today we observe that

Np
≈ 10−8 (2)
Nγ

4
Therefore when the temperature was high enough for pho-
tons to create protons and antiprotons, we had

Np − Np̄
≈ 10−8 (3)
Nγ

or equivalently, since Nγ was of the same order of magnitude

as Np + Np̄ ,

Np − Np̄
≈ 10−8 (4)
Np + Np̄

A natural question is :

Why is there today so much more photons than protons or

electrons ?

But the question is really upside down. Rather than why

there is so much photons, it should be why there is so few
protons and electrons.

There are basically three conditions, stated by Sakharov,

such that if they are satisfied – and they are satisfied –
then there will be an imbalance of matter over antimatter.

Then, once we explain the imbalance of matter over anti-

matter, the real question becomes the magnitude of it.

Let’s pause to make a general comment on the questions

we ask in cosmology, because it often comes up. Generally
speaking we might wonder : is what we are asking a legi-
timate question ? Perhaps the answer is simply : well, the
world started that way, end of story.

5
However, it focuses our attention a lot when we have a
number – in this case 108 photons per proton. This number
needs an explanation. The excess in itself was a fact about
which we could have said : that is just the way it is. But
once there is a number, and in particular if the number is
some oddball figure like ten to the minus eighth, we start
to think maybe this needs an explanation.

Sometimes the question is formulated as : why is the entropy

of the universe today so large compared to the number of
protons ? The reason is that roughly speaking, for a black-
body thermal spectrum, the entropy is simply the number
of photons. The blackbody radiation carries entropy. It is
thermal. And up to some proportionality factor, the en-
tropy of blackbody radiation is just the average number of
photons in a gas.

When people speak about the entropy of the universe, in

many contexts what they are talking about is simply the
number of cosmic microwave background photons. That is
what they mean by it.

So, as explained in the last lesson, and recalled above, in the

very early universe, when the temperature was a thousand
billion degrees, there was a huge number of protons and a
huge number of antiprotons. And they were basically equal
to each other. Or, multiplying by three, the number quarks
and the number of antiquarks were approximately the same.

The same is true for electrons and positrons, their numbers

being approximately the same. And all these numbers were
equal to the number of photons. In other words,

6
Nγ Np Np̄ Ne Ne+

were all about equal.

The protons, antiprotons – or more accurately quarks and

antiquarks –, electrons, positrons and photons, were in ther-
mal equilibrium at some very high temperature.

Then the universe cooled, fixing first of all Np , because the

photons could no longer create quark-antiquark pairs, and
then fixing Ne for similar reasons. So, when the universe
cooled, the photons were leftover.

They eventually decoupled from matter because, when pro-

tons and electrons bound up to form atoms, the particles of
matter became neutral and no longer scattered radiation.
The universe became transparent. The photons just hung
around, and it is those we see today in the CMB.

Let’s emphasize : before the present situation, the protons

and antiprotons annihilated each other. Somewhat later so
did electrons and positrons – the usual name for antielec-
trons. The universe expanded fairly slowly so there was
plenty of time for them to find each other and by and large
annihilate each other.

All that was left over was a slight excess of protons, with
positive charges, and correspondingly of electrons, with ne-
gative charges.

So from "why are there so many photons ?", the question

became "why is there this tiny excess of 10−8 protons, when

7
measured in terms of the number of photons ?" Why isn’t
it just zero.

Already in the very early universe we had

Np − Np̄
≈ 10−8 (5)
Nγ

The number of photons, Nγ , being approximately the same

at that time, in order of magnitude, as the number of pro-
tons plus antiprotons, Np + Np̄ , equation (3) can be rewrit-
ten as

Np − Np̄
≈ 10−8 (6)
Np + Np̄

So there is a number to compute. Why does this small num-

ber appear ? Where does it come from ? Is it possible, within
a theory, to compute it from more fundamental principles ?

We don’t know the reason for the number. That is because

we don’t have a complete theory. But almost any theory
that we write down, which explains it, gives a small num-
ber.

So we will talk about the kinds of conditions that are neces-

sary – and, it turns out, sufficient – to make an imbalance
of matter over antimatter.

Now, is it matter of antimatter ? How come it didn’t come

out antimatter over matter ? That is largely a definition.
The thing we call matter is the thing we are made out of.

8
Baryonic excess

Let’s begin with a hypothesis which is really believed to

be wrong, but we will come to it. It concerns the baryon 3
number. What does the baryon number mean ?

If there are only protons, the excess baryon number in the

world can be defined in terms of quarks and antiquarks. It
is the number
1
B = (Nq − Nq̄ ) (7)
3

where Nq is the number of quarks in the world, and Nq̄ the

number of antiquarks.

Why this factor 1/3 ? Because originally the baryon number

was defined in terms of protons and neutrons, and not in
terms of quarks. One proton was given the baryon number
1, and so was one neutron. Then physicists established in
the 1960s that a proton and a neutron were each made out
of three quarks. So one quark received the baryon number
1/3.

There are other kinds of objects that carry baryon numbers

besides protons and neutrons. But they are all unstable.
Even the neutron is unstable. Nevertheless there exist other
kinds of objects. We can mostly focus by thinking about
quarks themselves if we like.

The statement that there is a baryonic excess is the sta-

tement that there were in the early universe more quarks
3. Baryon means matter particle, or heavy particle. It is built from
the Greek word barýs which means heavy.

9
than antiquarks. The numbers of quarks and antiquarks se-
parately were about the same as the number of photons 4 .
So the question is how it got that way ?

Let’s suppose for the moment that baryon number is like

electric charge. One of the things about electric charge is
that it is conserved. It doesn’t change with time.

Now baryon number is not really like electric charge. Elec-

tric charge is the source of Coulomb forces 5 . Long-range
electric fields which create long-range electrostatic forces.
Baryon number itself is not a source of any kind of Cou-
lomb type force.

Of course the protons are electrically charged. Therefore

they make conventional Coulomb forces between each other.
They make electric fields. The neutrons are not electrically
charged. They don’t make electric fields. So what we would
say is that it is the charge of the proton, not the baryon
number of it, which is creating any kind of long-range field.
And baryon number itself may truly be concerned. But it
is not exactly like electric charge. It doesn’t exhibit this
tendency to make long range forces.

So, suppose it is conserved. Then if there was ever an ex-

cess, let’s say in the beginning of the universe, whatever
that means, then there will always be an excess. And that
excess will be sort of frozen in. If you change the number of

4. We are talking about orders of magnitude. Therefore the num-

ber of protons and the number of quarks can be both about "equal"
to the number of photons, even though there are three times as many
quarks as protons. In orders of magnitude they are the same.
5. Named after Charles-Augustin Coulomb (1736 - 1806), French
physicist best known for his study of electrostatic forces.

10
quarks, you must change the number of antiquarks by the
same amount, if baryon number is conserved.

Moreover, experimentally baryon number appears to be

highly conserved. Nobody has ever seen a proton disappear.
We can talk more about experiments which search for the
decay of protons and so forth. But in first approximation
in our world protons are extremely stable.

Suppose the proton was to decay. What could it decay to ?

It must decay into things which are lighter than itself. And
it must decay to something which has a positive electric
charge. So, if we want to assume that whatever it decays to
is stable, there is really only one thing that it could decay
to. It is a positron and something electrically neutral.

A proton could disappear and become a positron. That

conserves electric charge. But it doesn’t conserve energy. A
positron is much much lighter than a proton. But it would
compensate by giving off a neutral particle. What kind of
neutral particles are around ? Photons. So a photon is a
prime candidate.

Figure 1 : Possible decay of a proton.

11
Thus a decay possibility for the proton would be a photon,
γ, and an electron antiparticle, that is a positron, e+ . Let’s
draw it as a kind of Feynman diagram, figure 1. That is a
possible thing that could happen. And we don’t really know
any deep fundamental reason why it can’t happen.

Maybe it does happen. We are going to talk about whether

it happens. In fact we think it happens, but for whatever
reason, as mentioned, nobody has ever observed the decay
of a proton yet.

Questions / answers session

Question : Couldn’t the proton decay into a positron and a

neutrino ?

Answer : No, it cannot decay into a positron and a neu-

trino. That is no good. The reason is that a neutrino is a
fermion. So is a positron. And two fermions make a boson.

However the proton is a fermion. So whatever the proton

decays to, it must decay into something with an odd num-
ber of fermions.

Q. : If the baryon number was conserved, it would corres-

pond to a symmetry of some sort. Is there a natural candi-
date for that symmetry ?

12
A. : Yes. Given any conserved quantity, you can always
make up a symmetry 6 . But then you can ask : well, al-
right, is it really a symmetry ? And the way to test it, is
to ask whether the baryon number is conserved. So it is a
little bit circular.

But yes, if baryon number was conserved it would corres-

pond to a symmetry. And if it is not conserved, it doesn’t
correspond to a symmetry.

Q. : Is particule annihilation, like a proton and an antipro-

ton annihilating each other into pure radiation energy, a
decay ?

A. : No, we don’t call that a decay. It is not the decay of a

proton. It is just called annihilation of a proton.

But the important point here is that the baryon number

doesn’t change. Three quarks and three antiquarks came
together. Before the annihilation the sum of the baryon
numbers is zero 7 . And afterwards it is just a bunch of pho-
tons, whose baryon number is still zero.

6. A symmetry is a differentiable transformation of the degrees of

freedom describing a physical system which leaves the Lagrangian in-
variant. See chapter 7 of volume 1 of the collection The Theoretical
Minimum for Noether’s theorem about the relation between symme-
tries and conservation laws. For example, the conservation of electric
charge corresponds to global gauge invariance of the electric field.
7. If a particle has baryon number n, its corresponding antiparticle
has baryon number −n.

13
C, P and T
and Sakharov first condition

So for whatever reason – and we don’t really know the rea-

son – we haven’t observed the decay of a proton as shown
in figure 1.

Incidentally, the standard model does permit such a decay.

In fact it not only allows it, it insists on it at some level.

Now, one thing we know about it immediately is that the

decay rate of the proton cannot be very large. Our protons
have been around for 13.8 billion years and they haven’t
disappeared on us.

If the decay time of the proton were a microsecond or even

faster, whatever particle physics might permit, that proton
just wouldn’t be here anymore. So the lifetime of the proton
is at least very very long.

It fact the lifetime of the proton is much longer than is

necessary for them to still be here. The mean lifetime of
the proton is more than 1032 years. So it is very long. And
we don’t completely understand why. In fact we don’t un-
derstand why at all. But we take it as a fact, at least a
temporary fact.

So let’s assume for the moment that the proton doesn’t

decay. That implies that the total baryon number in the
universe – counting antiquarks as negative –, or let’s just
say in box, doesn’t change. Therefore, the only theory of the
excess B, in equation (7), that we could have is that it was
built in from the very beginning. And it still survives today.

14
That would be a consequence of baryon number conserva-
tion. We are not saying that it is true. We are saying that
it is a consequence of baryon number conservation, that B
is the same today as it was at the very beginning.

Now there are other transformations which were thought to

be symmetries of nature. They are called respectively C, P
and T , and stand for the following
• C = Charge conjugation
• P = Parity, or mirror reflection
• T = Time reversal

To say that C is a symmetry is the statement that if par-

ticles and antiparticles are interchanged, we cannot tell the
difference. Of course we can tell the difference between a
proton and an antiproton, because we are made of protons.
But suppose I was made of protons, and you were made of
antiprotons – don’t get too close to me –, and somebody
showed me a proton. I would say "yes, that is a proton".
And if somebody showed you an antiproton, you would have
exactly the same response, in the sense that you would say
"yes, that is the particle I’m made of". You would see the
same thing that I saw.

P has nothing to do with economics or fairness. It is mirror

reflection. If there is a particle which exhibits an orien-
tation, like a corkscrew spinning clockwise when it moves
forward, or like a person with a long right arm and a short
left arm, then it had been assumed that there would be a
particle with the opposite orientation, i.e., with the corks-
crew image, spinning counterclockwise when it also moves
forward. The statement that P holds is also called left right

15
symmetry, or simply parity.

Finally T , the last putative symmetry that we are concer-

ned with in this lesson, corresponds to running the film of
nature backwards. Consider any process that can happen
in nature. We can even restrict ourselves to laboratory pro-
cesses, and not worry about the whole universe. If we take a
movie of it and run it backwards, it should still be a process
which is physically possible, which can happen.

These three transformations were thought to be, naturally,

symmetries of nature, until Cronin and Fitch 8 discovered
CP violation in 1964. That was only three years before Sa-
kharov put out his theory of baryogenesis.

You could object that the time reversal T doesn’t seem true
in the real world of course. The second law of thermody-
namics tells us that things get only worse. But that is true
only in probabilities. It has to do with averaging over large
numbers of possibilites. At the true microscopic level, we
don’t average things out. So any process that happens in
nature was thought to be such that its time reversal was
another possible process.

Now, there is a mathematical result in quantum field theory

which states that the combination of C, P and T is a sym-
metry. It is not an axiom or a fact from observation, but
a theorem – which we are not going to prove in this lesson
– that was already known before the 1960s. It follows from
the basic structure of quantum mechanics and relativistic
field theory.

8. James Cronin (1931 - 2016) and Val Fitch (1923 - 2015), Ame-
rican particle physicists.

16
This theorem doesn’t say that charge conjugation – the
particle-antiparticle interchange – is a symmetry. It doesn’t
say that parity reflexion is a symmetry, nor that time rever-
sal is a symmetry. But it says that the product of all three
of them, CP T , is a symmetry. And this is for fundamental
mathematical reasons. It would lead to mathematical in-
consistencies if it was not a symmetry.

What does CP T symmetry mean ? It means that if we take

any process, replace every particle by its antiparticle, reflect
it in a mirror and run it backward 9 , it is still a possible pro-
cess in nature. We have to do all three in order to be sure
that it is a symmetry.

Let’s go back to the time when the three transformations

were thought to be symmetries. And, since people focussed
mostly on that, let’s consider the product CP . It means
changing every particle to its antiparticle and changing the
orientation of space to its reflexion in a mirror. So CP was
thought to be a good symmetry of nature.

If you believe in that, then you might ask : Well, gee whiz,
if there is a complete symmetry between particles and an-
tiparticles – after the change of orientation too, but that
went along for the ride –, why should it be that at the very
beginning there was an imbalance of one versus the other ?

Now, nobody can tell you that there isn’t an imbalance

9. When we change the sign of t we also have to take the complex
conjugate of the wave function representing the time evolution of the
state-vector of the system. This can be seen for instance with the
time-dependent Schrödinger equation. See chapter 4 of volume 2 in
the collection The Theoretical Minimum, on quantum mechanics.

17
that just dates back to the very beginning. But you might
also wonder : What is going on here ? The laws of physics
seem to be completely symmetric between the two kinds of
things, particles and antiparticles, and yet for some reason
there was this small imbalance of size 10−8 . That doesn’t
sound right.

The modern theory of baryogenesis begins with the idea

that there was a balance. It says that particles and antipar-
ticles were balanced – again, not for any good reasons. It
just assumes that, in the initial conditions we started with,
there was no bias toward particles or antiparticles. It can
actually be justified in some frameworks, but we are not
concerned with that.

Then how is it possible that it got into an imbalance ? The

only way it is possible for it to get into an imbalance is if
the conservation of baryon number is not correct. In other
words, if processes can happen in nature in which a proton
becomes a positron, that is a violation of baryon conser-
vation which allows the baryon number of the universe to
change.

It would be the first requirement for a theory of baryoge-

nesis, based on the assumption that the starting point was
balanced between the two, explaining how we are going
to wind up with an excess of quarks over antiquarks, or
baryons over antibaryons – baryons meaning protons and
neutrons. We must have a mechanism which violates the
conservation. So that was Sakharov first condition :
1) Baryon number violation.

Violation means violation of the conservation law. And the

18
process of a proton becoming a positron and a photon is an
example, if it happens in nature.

Let’s comment on this first condition. If baryon number

conservation is not a good conservation law of physics, then
it must be of very very weakly broken one. As we said, pro-
tons are very old. They didn’t disappear, so they must be
very old. So whatever mechanism to change baryon num-
ber, such as the kind of decay shown in figure 1, it must
have an extremely small probability per unit time.

In fact every current fundamental theory, unified or not, if it

couples to gravity, violates baryon conservation. Therefore
you can ask : In the known theories, why is the proton so
stable ? The answer goes something like this. Any theory
has a Feynman diagram explaining the decay of the proton
as in figure 2.

Figure 2 : Decay of a proton, with intermediate particles, and in

particular a transient heavy mass M .

The proton comes in from the left, meaning from early.

Out goes a positron and a photon. But somewhere in the
guts of the Feynman diagram there are all kinds of par-
ticles. Let’s not specify exactly which ones are in there.

19
But among them, it is assumed that there is one or more
particles which are very very heavy.

In other words, the decay of the proton requires a particle

type which is very very heavy in the process – particles
which have not been discovered yet.

One of the reasons to believe this is that the standard mo-

del by itself, with its ordinary known particles, does not
permit this decay to happen. So in order to make it happen
you would have to have new additional particles that were
not part of the standard model. And that certainly means
that they are heavier, because they haven’t been discovered
yet.

But imagine making them very heavy. For instance 1016

GeV, that is 1016 heavier than a proton, which is not an
unnatural number for heavy particles. Then what is true is
that the kind of process in figure 2 is extremely unlikely.
By extremely unlikely we mean that the quantum mecha-
nical amplitude 10 for it is squashed by inverse powers of
the heavy mass.

Let’s call the heavy mass M . Then the Feynman diagram

will contain a factor 1/M 2 , just because it is so hard to
make a heavy particle.

The heavy particle won’t last very long because it has to

melt into the rest of the Feynman diagram. But, irrespec-

10. The reader is invited to go back to volume 2, in the collection

The Theoretical Minimum, on quantum mechanics, to review what is
the amplitude of a given observable state for a system, and its relation
to probabilities.

20
tive of its lifetime it has to exist for a wink during the
decay, and, if the particle is heavy enough, then the decay
will have a very very small probability. That is how modern
unified theories explain the stability of the proton.

We don’t know if it is right. But it is a tentative explana-

tion. The mechanism requires transient extra particles that
are sufficiently heavy to squash the probability for the de-
cay to happen.

Now let’s take a second look at the factor 1/M 2 which re-
duces the amplitude and the probability of decay. If for
some reason the proton is given a lot of energy E some-
how 11 , then, without going into the quantum calculations
about the Feynman diagram, the factor becomes
1
(8)
(M − E)2

In other words, it may well be a much smaller denominator

dividing the amplitude. We will see in a moment where that
extra energy comes from, or can come from.

In the ordinary world, when we look at a proton sitting

around, there is nothing to give it that enormously high
kick E. It doesn’t have a huge amount of energy. Protons,
when they collide or anything else, don’t have huge amounts
of energy. That is the reason why we begin with 1/M 2 . And
theories of this type can explain why the protons are stable.

Of course you could object that they don’t explain any-

thing. They just formulate the hypothesis that transient
11. That is an extra energy on top of its rest mass energy M c2 . And
in our calculations, we have chosen units such that c = 1.

21
particules, which are necessary for the proton decay, must
be very heavy.

As we said, all known theories – and I will venture to say

any theory that ever will exist – allow for the possibility of
baryon violation.

So let’s agree tentatively that baryon violation is not taboo,

that processes such as shown in figure 2 can happen, they
are not forbidden by any fundamental law of physics. It is
just an accident of the particle spectrum that the proton
is as stable as it is. Let’s take that as an assumption, or a
working hypothesis.

Baryon violation by itself is not sufficient to give us an ave-

rage excess of protons over antiprotons. Why not ? Because
for every process like in figure 1, which reduces the baryon
number in the world by 1, since there is 1 baryon unit on
the left and 0 on the right, there is the charge conjugate
process, figure 3.

Figure 3 : Charge conjugate process of the proton decay.

An antiproton comes in. A true electron goes off. And the

antiparticle of the photon is the photon itself. And that

22
charge conjugate decay cancels a −1 baryon, that is to say,
eliminates an antiproton, of baryon number −1, to replace
it by an electron and a photon, both of baryon number 0.

So, if we believe in particle-antiparticle symmetry, then, for

every process like in figure 1, a process like in figure 3 will
happen with the same probability. And, on the average,
there will be as many proton decays as antiproton decays.

But as a consequence, if we started out with an equal po-

pulation of quarks and antiquarks, of baryons and antiba-
ryons, and we tried to rely only on the violation of baryon
number, it would not be a very efficient way to create an
excess.

Incidentally, in the very early universe, there were lots of

electrons, positrons and photons around. And it was also
perfectly possible for the opposite processes to go. A posi-
tron and a photon can come together and make a proton.
An electron and a photon can come together and make an
antiproton. So there is a statistical balance of things going
on.

It is only statistical that there would be as many protons

and antiprotons, because these processes are statistical pro-
cesses.

But the imbalance of protons over antiprotons, that we

measure, cannot be a statistical effect. The impossibility
comes from an elementary argument in probability theory.
Let’s go through it.

The baryon excess is about 10−8 times the number of ba-

23
ryons in the very early universe. How many protons are
there in the entire known observable universe today ? The
answer is about 1080 .

We can view the excess of baryons as the result of N head

or tail tosses where N is a number even larger than 1080 .
On average that will produce N/2 baryons and N/2 antiba-
ryons. But these won’t be the exact numbers. There will be
an excess of one over the other due to statistical variation.
Probability theory tells us that this excess has an order of
magnitude of square root of N 12 .

For example, if you flip a coin 1000 times, you will get on
average 500 heads and 500 tails, but not√ exactly. The stan-
dard deviation of the difference will be 1000, that is about
32. And the difference of heads and tails will almost surely
be less than three times the square root of 1000, that is 100.

So the difference between baryons and antibaryons in abso-

lute value, divided by N , if it was due to a statistical effect,
should be less than three times the square root of 1080 di-
vided by 1080 , that is three times 10−40 at the most.

We conclude that the excess of 10−8 that we measure is way

too big to be the result of a statistical variation around zero.

The other observation that would be extremely hard to jus-

12. Consider a random variable X that can take the value +1 with
probability p, and the value −1 with probability (1 − p). Then the
expectation of X is 2p − 1. In the case p = 1/2, this is of course
zero. The variance of X is 4p(1 − p). If Y is the sum of N independent
random variables like X, pthen the variance of Y is 4N p(1 − p). And its
standard deviation
√ is 2 N p(1 − p). If p = 1/2, this gives a standard
deviation of N .

24
tify, if it was only a statistical variation, is why it is the same
everywhere we look in the universe ? You would expect a
patch over here with protons, a patch over there with anti-
protons, etc.

What is the experimental evidence that no neighboring or

even distant galaxies are antigalaxies ? Astronomers point
out that there is good evidence. If the population of galaxies
were sort of symmetric between galaxies and antigalaxies,
then we would expect cosmic rays, and especially very high-
energy cosmic rays, which are thought to be cosmic in ori-
gin, to have as many nuclei as antinuclei. However, we do
see for example helium nuclei in cosmic rays, but nobody
has ever seen an antinucleus in cosmic rays.

We do see antiprotons, but that is fairly easy to explain.

Even if there weren’t any antiprotons in cosmic rays, when
a high-energy particle hits the atmosphere it can make an-
tiprotons. But it is extremely difficult to make an antinu-
cleus. That would take an incredible piece of luck, a very
high energy collision with the atmosphere creating an an-
tinucleus. So all this is consistent with the hypothesis that
there are no antigalaxies out there, and that the antipar-
ticles we do observe are created in collisions with the at-
mosphere.

The complete absence of antinuclei strongly corroborates

the idea that the universe is not equally populated with
galaxies and antigalaxies.

So there is something to explain. The excess of baryons over

antibaryons is not statistical, and we need to explain it.

25
Let’s finally mention another thing we don’t see. If our ga-
laxy for example was made of particles and the Andromeda
was made of antiparticles, then in the region in between we
would expect to find both particles and antiparticles. For
example there are plenty of electrons circulating in the re-
gion between galaxies. We would expect there would also
be plenty of positrons. And then we would observe lots of
positron-electron annihilations.

Positron-electron annihilations are very easy to detect. They

produce pairs of photons of very definite energy. These
pairs of photons could be observed. Now, it is not that we
don’t see any electron-positron annihilation, but we don’t
see nearly enough of it to account for the possibility that
some neighboring galaxies would be antigalaxies.

So it is almost certainly ruled out that there are antiga-

laxies out there. And it is certainly ruled out that there is
an equal population of them.

That closes our series of arguments for saying that the ba-
ryon excess that we obseve is not a statistical effect.

CP and Sakharov second condition

Particle-antiparticle symmetry tends to suggest rather stron-

gly that the universe was created symmetrically, although
it doesn’t prove it.

So the next element of the argument, in order to account for

the fact that the excess of protons over antiprotons that we

26
observe cannot simply be a statistical effect, is that charge
conjugation symmetry – i.e. the idea that the particles and
antiparticles are symmetric in the laws of physics – must
fail.

If we have charge conjugation symmetry, C, or charge conju-

gation times parity, CP , or charge conjugation together
with anything else, any symmetry that involves interchan-
ging particles and antiparticles, then we have a very hard
time explaining the imbalance that we observe.

Another way to say it is that if we only allow baryon vio-

lation – i.e. Sakharov first condition – that is not enough
to explain the magnitude of the imbalance. That will just
give us the statistical effect 13 .

We need something in the laws of physics to bias it to-

ward particles rather than antiparticles, in other words so-
mething, in the baryon violation, which makes one of the
decays, in figures 1 and 3, more probable than the other.

The implications of all that are that you need violations of

particle-antiparticle symmetry, in particular the so-called
CP symmetry, but basically just particle-antiparticle sym-
metry.

Again, is there particle-antiparticle asymmetry in the world ?

Yes, there is. We know with absolute certainty experimen-
tally that particles and antiparticles don’t behave the same
13. Violation of baryon number conservation doesn’t imply viola-
tion of symmetry. We investigated the consequences of Sakharov first
condition while keeping symmetry. And we showed that it lead to a
wide discrepancy between the proton-antiproton imbalance permitted
by statistical fluctuations and that observed.

27
way. The examples are hard to come by, but once you have
one or two or three, you know that the laws of physics are
not symmetric between particles and antiparticles.

The simplest example to explain is the so-called B meson.

It is a particular kind of particle made of a quark and an an-
tiquark 14 . There are actually four kinds of B mesons, made
of an up quark and a bottom antiquark, denoted ub̄, or a
down quark and a bottom antiquark, db̄, or a strange quark
and a bottom antiquark, sb̄, or finally a charmed quark and
a bottom antiquark, cb̄.

To simplify the argument, let’s focus on the B meson made

of the u quark and the b̄ antiquark, bound into ub̄. This
particle is simply denoted B+ .

B+ has an antiparticle. You just interchange the quark and

the antiquark. Instead of ub̄, you consider the assemblage
of ū and b into ūb. This is the anti B meson denoted B− .

Both B+ and B− are electrically neutral. But they are not

their own antiparticles.

Each of these particles decays into two other particles.

B+ −→ particle 1 + particle 2
B− −→ antiparticle 1 + antiparticle 2

To know the decay particles is not important for us. The

only important thing to remember is that it is a process in
14. There are six types of quarks respectively named up, down,
charmed, strange, top and bottom, and denoted by the letters u, d, c,
s, t and b. And there are the six corresponding antiquarks, which are
denoted with the same letters with a bar on top.

28
which a particle B+ decays into two particles, versus the
process in which the antiparticle B− decays into the corres-
ponding two antiparticles.

Now the rates for these decays to happen are measurable.

They are measured. And they are different ! One of the de-
cays is about two-thirds more important than the other.
And so in this case it is a fairly gross violation of symmetry
in the particle-antiparticle interchange.

It is definitely a real effect. There are lots of indirect mea-

surements of it. It has been known for a long time that
particle-antiparticle symmetry in reality is not a symmetry.

Once it is established that there exist fundamental pro-

cesses in nature, buried deep inside Feynman diagrams some-
where, that are imbalanced between particles and antipar-
ticles, it is also no longer the case that the decay of the
proton and the decay of the antiproton, into respectively a
positron and a photon, and an electron and a photon, as
shown in figure 4 below have to have equal probability.

There is a rule however that the total half-life of the proton

and of the antiproton have to be exactly the same. It is a
theorem in relativistic field theory.

But the way that the proton and the antiproton decays can
be different from each other, while their total half-lives are
the same, is that there is more than one possible way for
the proton to decay.

29
Figure 4 : Possible decays of a proton and an antiproton.

A proton can also decay for example into an antimuon µ+

and a photon. And an antiproton can decay into a muon 15
µ− and a photon.

What the theorem says is that if you consider all the pos-
sible ways that the proton can decay, and you calculate its
total rate of decay, it must be exactly the same as the to-
tal rate of decay of the antiproton. But it doesn’t say that
decay rates of the processes shown in figure 4 must be the
same.

15. The muon is an elementary particle similar to the electron, with

the same negative electric charge, but 200 times more massive. The
electron mass is 0.5 M eV , while that of the muon is approximately
100 M eV . Remember that the mass of the proton is ≈ 1 GeV .

30
The theorem doesn’t say that any particular decay has to
have a symmetry. So in particular, if at some fundamental
level the charge conjugation symmetry is violated by so-
mething in the theory, then it will allow the two decays in
figure 4 to be asymmetric, and the decay rates to be dif-
ferent.

It not only allows it, it basically insists that the two decays
in figure 4 be not the same.

Once that is true, it says that there is a bias somewhere in

the laws of physics – a bias toward either protons or an-
tiprotons, or matter vs antimatter. And that is something
which is absolutely necessary to add to baryon number vio-
lation to give it a directionality, that is to give it a push in
one direction rather than the other.

So Sakharov second condition is CP asymmetry 16 .

As said, Sakharov’s paper was published only three years af-

ter the discovery of CP violation. That is remarkably short.

Let’s list the two conditions we already have

1) Baryon number violation.
2) CP asymmetry.

Any theory that has been developped for particle physics

always contains the first condition, including the standard
model. It insists, as a theoretical statement, that baryon
number violation must happen.

16. In this lesson, we don’t distinguish parity separately from

particle-antiparticle inversion, so for us C and CP are the same.

31
Experimentally, we know that CP is violated. In other
words, particle-antiparticle interchange times parity has been
discovered not to be a symmetry.

Now there is one remaining condition. But before going into

it, let’s do a questions / answers session.

Questions / answers session (2)

Q. : If the total decay rates of protons and antiprotons are

the same, then why don’t we have equal numbers of protons
and antiprotons ?

A. : It comes from the fact that the µ particle is heavier

than the electron.

To explain it, I need to explain first why the decay of the

proton into a positron and a photon can happen rapidly,
even though I said that the proton is very stable. It has to
do with the environment.

How do we overcome the incredible stability of the proton ?

Remember that its mean lifetime is something like about

1033 or 1034 years or longer. The universe has only been
around for about 1010 years. So on that scale the proton is
very stable. So why don’t we get to ignore the asymmetry
between proton and antiproton decay ?

The reason has to do with the fact that the universe in its
very early stages was very hot.

32
Because it was very hot, the protons or the quarks in par-
ticular – we can substitute quarks for protons in figure 2,
it doesn’t matter – were constantly engaged in very high
energy collisions.

The high energy collisions meant that the protons moving

in the plasma, that is the very hot gas or hot primordial
soup, had lots of energy. How much energy ? It depends on
the temperature.

But at some high enough temperature, where the average

extra excitation energy E of the proton is high enough, the
transient heavy particles shown in figure 2 no longer sup-
press the proton decay efficiently. We saw in equation (8)
that the coefficient, which reduces the probability of decay,
itself decreases from M 2 to (M − E)2 , thereby increasing
the probability of the decay, or of the baryon violation.

In other words, to put it short, a proton in an environment

where it isn’t constantly being knocked around and having
an enormously large excess energy E of some kind of ano-
ther is very stable.

But when it’s heated up to a high enough temperature, the

transient heavy mass M shown in figure 2 is no longer an
important factor, and the proton decay will happen quickly.

So if we go back to the very early universe when the tempe-

rature was very hot, then the proton and antiproton decays
can happen.

Now the statement that the total decay rate is the same for

33
protons and antiprotons is a statement about zero tempe-
rature. It is a statement about a proton at rest in an en-
vironment where it has no excess energy. When it is being
kicked around and it has extra energy then that is not ne-
cessarily true.

In fact it is generally not the case. In an environment which

has some energy around, which is kicking the protons around,
the decay rates don’t have to be the same. So that is not a
problem.

The problem has to do with the CP T symmetry.

Time-reversal and Sakharov third condition

We saw that there is no particle inversion symmetry. But

there is a symmetry – at least in all known quantum field
theories, string theories, any theory we know how to write
down, which have relativity and quantum mechanics built
into them : it is charge conjugation times reflection times
time reversal, that is CP T .

Again what that means is

1. you exchange every particle for its antiparticle,
2. you reflect in a mirror,
3. and you run the film backward.

That is a symmetry of all theories.

34
What does that mean ? That means, among other things,
that in thermal equilibrium, where the universe is just a
pot static and hot, forward times is the same as backward
time. In a universe in thermal equilibrium backward and
forward in times are the same thing ; there is no asymme-
try of time-reversal ; there is T symmetry.

But if the universe has T symmetry, and it also has CP T

symmetry, then it must have CP symmetry. CP just means
particle change to antiparticle for our purposes.

In other words, if particle goes to antiparticle combined

with time goes to minus time is a good symmetry – and
that, we know, is always true because it is a mathematical
theorem of quantum field theory –, and the world has no
bias toward one direction of time vs the other, then we are
stuck again. We are back to having a particle-antiparticle
symmetry. And we cannot have a proton antiproton excess
in thermal equilibrium.

It is actually a theorem, which has been known for a long

time, that says that in thermal equilibrium we cannot have
an excess of protons over antiprotons. The thermal equili-
brium will necessarily come to a configuration with equal
number of protons and antiprotons.

But at early times the universe was not in thermal equili-

brium. It was expanding fast – fast enough that it could not
be considered in thermal equilibrium. If it is not in equili-
brium that means that forward time and backward time are
different. That is the reason : if the universe is expanding
fast, forward time and backward time are not the same.

35
So in a rapidly expanding phase of the universe, we don’t
have to worry about time symmetry. It is definitely not
symmetric. If time symmetry is broken, just by the rapid
expansion of the universe, then we are in business. We have
enough asymmetry of all possible kinds to explain the mat-
ter antimatter imbalance.

The baryon violation allows a change in the baryon num-

ber. The CP violation allows a directionality for it. And
now the third condition is that the universe be out of equi-
librium.

This third condition simply means the universe is expan-

ding fast enough that running the thing backwards does not
look like the original thing. It is not enough for the universe
to be slowly expanding. It has to be expanding sufficiently
rapidly so that all the microscopic processes don’t have the
time to adjust themselves to the equilibrium configuration.
But whatever they are, it means that backward and for-
ward times are different.

Movies of the universe must allow us to tell which way is

forward in time and which way is backward in time 17 , by
the rapid expansion and cooling. The fact that at early
times the universe has cooled rather suddenly is enough
to completely ruin the time reversal symmetry. And if the
time reversal symmetry is ruined, the CP symmetry is also
necessarily ruined.

17. Notice that if we make a movie of the movements of particles

in a hot static chamber, or even a chamber whose volume expands
slowly while a piston moves, and we show it to spectators, they will
not be able to tell whether we are showing them the movie with the
time running forward or running backward.

36
We can finally list the three conditions of Sakharov :
1) Baryon number violation.
2) CP asymmetry.
3) Out of equilibrium

All three are believed to be really satisfied in the real world.

They are also believed to be sufficient. If of all three of these

are true, it would be a complete accident if there was not
some excess created.

The problem is that nobody knows enough about the phy-

sics of the early universe and the physics of very high energy
collisions, the physics of very hot temperature, the nature
of the particles that are in here, the details of what drives
the CP violation and so forth, to be able to calculate the
imbalance of protons to antiprotons.

So we know the three ingredients necessary and proba-

bly sufficient to explain an imbalance. It is Sakharov three
conditions. But the ingredients needed to make a compu-
tation to show that

(Nq − Nq̄ ) = 10−8 (Nq + Nq̄ ) (9)

that is out of reach. We don’t know how to do that.

That is the status of this particular problem. It is the pro-

blem of baryogenesis. And it will await a much more detai-
led theory of both early cosmology and particle physics at
very high energy.

We are now finished with baryogenesis.

37
Questions / answers session (3)

Q. : Doesn’t the fact that we have a measurement of 10−8

put some constraints on what these theories can look like ?

A. : It does. But they are hard to use.

Nobody, to my knowledge, has ever used that number in a

really effective way to constrain things. There are too many
variables, hundreds of parameters. We just don’t know en-
ough.

Q. : What is the connexion between the positrons and

electrons numbers that you mentioned previously, and the
asymmetry in protons and antiprotons, or quarks and an-
tiquarks ?

A. : Every time a proton disappears a positron appears 18 .

Figure 5 : Decay of a proton into a positron and an electron.

18. There can be other decays, for example into an antimuon µ+

and a photon, but the reasoning is the same.

38
So the loss in baryon number is made up for by an in-
crease in something with a positive electric charge. Conver-
sely anything that creates a proton must either cancel a
positive charge or create another negative charge too.

As long as all the processes we are looking at conserve elec-

tric charge, there is no alternative to the statement that if
the baryon number shifts one way, the electron excess must
shift the other way.

Q. : We hear claims that, in the standard model, the CP

violation is not big enough to explain the baryonic excess.
What to make of them ?

A. : I’m aware of those. And I have no reason not to believe

them. I just am a little bit skeptical that anybody really
knows how to do the calculation.

My involvement in baryogenesis ended with stating the

three conditions, which I later discovered to have been sta-
ted thirteen years earlier by Sakharov.

Since then, dozens of people started to try to make com-

putations based on standard models. Stephen Wolfram was
one of them. He invented Mathematica, I think, to do the
calculations.

In my opinion there is no way that they are going to be able

to do this without, as I said, a much more detailed theory
of both early cosmology and particle physics at very high
energy, because there are just too many unknowns about

39
the early universe and so forth.

But it may be correct. And I don’t know. I haven’t followed

this story for a long time. It may be that we know that CP
violation in the standard model is too weak to drive the
excess.

Once you admit CP violation into physics, though, there

is no reason not to expect that at high energies the ratio
could be tens or hundreds of times larger than 10−8 . Once
you have opened the door to the imbalance, you don’t have
really much control over it.

Q. : Has the 10−8 number always been the same number ?

A. : The 10−8 is today’s number for the ratio of baryon

number or protons to photons.

In the early universe, there was a lot of annihilation bet-

ween protons and antiprotons. But before the temperature
fell low enough, it was compensated by the creation of new
protons and antiprotons. There was an equilibrium.

As I said, the five numbers

Nγ Np Np̄ Ne Ne+

were all about equal. But there appeared a slight excess of

protons over antiprotons, and an exactly opposite excess of
electrons over positrons. The excess is slight, but big en-
ough not to be explainable by statistical fluctuations.

40
At very high temperature, there are very high-energy pho-
tons. Two photons come together and they create one way
or another a proton and an antiproton. These photons must
together have at least 2 GeV of energy.

As long as the temperature is high enough, these processes

are going on backward and forward in equilibrium.

But then, when the temperature falls below a certain thre-

shold, there simply aren’t enough high energy photons around
to create protons and antiprotons. Protons and antiprotons,
however, continue to annihilate each other creating pho-
tons.

Those photons still have a lot of energy. But they go out

into the soup and their energy gets lowered by coming to
equilibrium with the lower temperature background stuff.

So as the temperature goes down the number of available

high-energy photons decreases and you can’t make the proton-
antiproton pairs anymore. But the other way around still
goes on.

Thus there is a point at which the annihilation is no longer

compensated and eventually all the protons and antipro-
tons get eaten up, except for the excess of one kind over
the other.

Once the excess reached a certain value in the early uni-

verse, it then never changed. Np −Np̄ stayed the same. This
difference is today’s number of protons in the universe. So
the ratio of 10−8 could be measured in the past, but as an

41
excess of protons over the total number of protons and an-
tiprotons, or equivalently over the total number of photons,
because in orders of magnitude the two denominators are
the same.

Q. : When did the process of creation of protons and anti-

protons start ?

A. : That was quite early, when the temperature was in the

gigaelectron-volts for the energy of the photons.

It began within a second or so after the initial inflation

which created space, and which is our next subject in this
lesson and the following one. So it is at the left extremity
of figure 6, right after inflation.

Figure 6 : Chronology of the universe.

Source : NASA/WMAP Team.

42
Q. : Why are the photons today not as energetic as when
they were creating pairs of protons and antiprotons ?

A. : Today’s photons – those of the CMB, which are the

overwhelming majority of photons in the universe – are
those that participated in the creation of proton-antiproton
pairs and also electron-positron pairs. They remained as nu-
merous in order of magnitude. But they are no longer as
energetic.

Remember that radiation energy density decreases like 1/a4 ,

where a is the scale factor, while the volume of the universe
increases like a3 .

Q. : If we need these transient very massive particles in the

decay of the proton, figure 2, does that mean the standard
model is still incomplete ?

A. : Yes, it does.

That is what is known about the matter-antimatter imba-

lance in the universe. And, as I said, I don’t think it is going
to be too soon before we know a lot more and in particular
are able to calculate the 10−8 ratio from theory.

The next topic we want to move onto is inflation, or in-

flationary universe. We shall do some groundworks in the
remainder of this lesson. And next lesson will be entirely
devoted to it.

43
Inflationary universe

The inflationary universe idea was put forward to try to

explain another of those observations which were puzzling.
The question was why is the universe so terribly homoge-
neous and isotropic. It became a rather critical issue when
the cosmic microwave background was discovered.

The CMB quickly became a rather high precision affair.

In short order the CMB wavelength distribution had been
observed to correspond to the blackbody curve ; its tempe-
rature had been measured carefully ; it was known to be
about 2.7 degrees Kelvin. Over the years, the fit between
the CMB and the blackbody curve became incredibly pre-
cise. Today the error bars are microscopic.

So the temperature was very well defined. The CMB looked

like a Planck distribution, that is like a blackbody distri-
bution almost exactly – exactly as far as we could tell.

But, moreover, it was the same in every direction. Thus

suddenly this rough idea that the universe is isotropic be-
came a high precision idea.

In truth, it took years to establish this scenario. I’m conden-

sing history. But, by now in any case, the idea that the uni-
verse is isotropic is a very high precision fact, and can be
the starting point for further theoretical developments. So
we can ask why is the universe so isotropic ?

If today it is isotropic, it must have been isotropic very

early. There is no particular reason why an anisotropy, a
lumpiness in the distribution, would decrease with time.

44
In fact, it is quite the opposite : lumpiness tends to increase
with time, because of gravity. Gravity tends to take lumps
and increase their magnitude.

Consequently, the universe must have started very early

being extremely homogeneous and isotropic 19 . When I say
very early, I mean at the time before there were galaxies,
before there were stars and planets, at the time when the
blackbody photons originated. In other words, at the de-
coupling time the universe was extremely isotropic.

It was of course known that it couldn’t be completely iso-

tropic and homogeneous. If it was exactly homogeneous, it
would stay homogeneous. Anything which is exactly uni-
form and allowed to evolve will stay uniform.

But the universe is not uniform. It is full of galaxies and

full of clusters of galaxies. It has a lumpiness.

That lumpiness, that we see now, clearly was much smaller

in magnitude to begin with, for the simple reason that lum-
piness tends to increase with time. As said, if you start with
a world which is completely uniform of course it stays that
way. But let’s suppose there was a little bit of overdensity
in some fairly big region. The density was just a tiny bit
bigger than in the neighboring regions. What happens ?

In a gravitational theory what happens is the opposite of

what happens in other kinds of theories.
19. Homogeneous and isotropic is a conventional wording. Logically
isotropy does not imply homogeneity – humankind could happen to be
at the center of the universe –, whereas homogeneity implies isotropy.
So to say the universe is homogeneous would be sufficient. We even
often use isotropic to mean homogeneous.

45
In other kinds of theories what tends to happen is that the
overdensity will diffuse out and be eliminated. For example
we have an overdensity of ink dropped into water. Over
time the drop diffuses out and then disappears, the density
of ink becoming homogeneous.

In gravity the opposite happens. And the argument is very

simple. If we have an overdensity of matter in some region,
because gravity is universally attractive it will tend to at-
tract the stuff around it. It will pull into the region the
stuff outside, decreasing the density outside and increasing
it inside, figure 7.

Figure 7 : Evolution under gravity of slightly overdense regions

into aggregates of matter.

In other words, it is kind of a runaway situation where a

little bit of inhomogeneity will tend to reinforce itself. So if
we have, distributed in space, regions with overdensity, un-
derdensity, overdensity, underdensity and so forth, in some
pattern, the tendency of gravity will be to suck stuff out

46
of the underdense regions and put it in into the overdense
regions, thereby magnifying the degree of inhomogeneity.

So the fact that we see inhomogeneity today does not mean

that the universe very early was as inhomogeneous as it is
today. It must have been much less inhomogeneous, just by
running the argument backward.

In fact, rather early, cosmologists were able to estimate,

by running the theory backward, just how much inhomoge-
neity there was at the time of decoupling in order that the
galaxies could nucleate out of that inhomogeneity. In other
words the picture is that the universe was very homoge-
neous, with just a little bit of inhomogeneity. There were
little ripples, little bits of excess, depletion, excess, deple-
tion of some kind.

The overdense regions eventually collapsed by the mecha-

nism shown in figure 7. They formed galaxies and clusters
of galaxies. And the underdense regions formed voids. And
that is what we see today.

Jim Peebles 20 and other early cosmologists in the sixties

estimated how much inhomogeneity was necessary at the
time of decoupling.

Here is the way we quantify it. We characterize the lumpi-

ness by the ratio
δρ
(10)
ρ

20. Phillip James Edwin Peebles (born in 1935), Canadian-

American physicist and theoretical cosmologist.

47
where ρ is the density, and δρ is roughly the mean excess
density in a lump relative to the background. So the ratio
is the fractional overdensity in a typical overdense region
relative to the density itself.

That is the dimensionless measure of how much inhomo-

geneity there was. And it was pretty early recognized that
this number had to be

δρ
≈ 10−5 (11)
ρ

More precisely, so to speak, it had to be in a range between

10−4 and 10−5 .

Here we were again in the same kind of situation as with the

excess of matter over antimatter. Did the question "Why
was the universe almost perfectly homogeneous ?" really re-
quire an answer ? Perhaps it was enough to say : "Well, it
just started that way." But now there was a number in ad-
dition to just saying the universe was almost homogeneous.

The universe was almost perfectly homogeneous, with a

specific numerical magnitude to its inhomogeneity. And once
you have a specific numerical magnitude, you want to know
why that is the magnitude.

Now, it is not that we learned why this is the magnitude –

we did not. We do not really understand why 10−5 is the
right number. But the existence of this number focused our
attention on two questions. The first is : Why is the uni-
verse almost perfectly homogeneous ? Later we will worry
about the second : Why is it not exactly homogeneous ?

48
The first-order fact was that it was very very homogeneous.
The explanation of that today – which seemed rather far-
fetched when it was put forward by Alan Guth 21 in 1980
– was that the universe simply expanded by many orders
of magnitude. And of course when something expands it
stretches.

If you have an inhomogeneous universe with lumpiness in

it, and you stretch it out enough, you will make it homoge-
neous, at least on the scales that are relevant. So the idea
was the universe inflated, which means expanded. In par-
ticular it exponentially expanded for some period of time,
and stretched itself out so much that it flattened itself.

It is literally like blowing up a balloon. When it is deflated,

the balloon has a crinkled shape. We blow it up and it
stretches out and flattens out, figure 8.

Figure 8 : Inflating a balloon flattens out its crinkles.

21. Alan Harvey Guth (born in 1947), American theoretical physi-

cist and cosmologist.

49
Of course that is an analogy which we shouldn’t take too
far. But it may help feel how streching leads to an homo-
geneization of density.

Questions / answers session (4)

Q. : Can you explain how the universe can be homogeneous

if at the same time it has lumpiness ?

A. : It is a matter of scale. If we look around this room, it

does look homogeneous at all. But on an average over large
distances, for all practical purposes it is homogeneous.

It is like the surface of the Earth. It is full of hills and moun-

tains and valleys and so forth, and as a geologist you want
to explain these geological phenomena. But on the whole
the Earth is a very smooth ellipsoid.

The universe is the same. It may have lumpiness, galaxies

and so forth, and yet be considered homogeneous when loo-
ked at on a sufficiently large scale.

Q. : If this inhomogeneity of the universe is a random pro-

cess causing lumps to appear, wouldn’t we expect much less
overall homogeneity ?

A. : The initial inhomogeneity and the appearance of lumps

is not a random process in the sense you think.

50
You are suggesting that if it were random we might expect
that somewhere out there there would be large overdensi-
ties. There would be some statistical probability that we
might find a large lumpiness.

But it is not random 22 . And we are going to talk about the

pattern of δρ/ρ.

Basically it is like this

Figure 9 : Inhomogeneities in one dimension.

Now squeeze this down until it is of size 10−5 relative to the

background. On the whole it is very smooth. Nevertheless
there are wiggles in it. The sizes of the wiggles are very
small.

Suppose, furthermore, they are homogeneously, isotropi-

cally distributed through space. There are no tremendously
large lumps out there. They are all very small in magnitude
relative to the average. It is more or less like a very very
smooth Earth with ripples on it.

We will see what the implications are.

22. Most paradoxes and popular riddles about randomness come

from a deficient or mistaken specification of what is the experiment to
be replicated, and what is the random variable to be measured.

51
Q. : What was the temperature like during inflation ? How
did it decrease and what role did it play ?

A. : During the inflationary period temperature was not an

important aspect of things. It was not a terribly relevant
factor.

It was the exponential expansion which was the single most

important thing.

We shall go through the basic equations of inflation in detail

in the next lesson. To finish the present lesson, let’s review
some preliminaries about friction that will be useful.

Friction and viscosity

In the context which will occupy us, when we speak of fric-

tion we are speaking of viscosity. We are thinking about
something like a stone falling through a viscous fluid.

What does friction or viscosity have to do with anything ?

It has everything to do with everything. But to finish this
lesson and prepare for the next one, we are just going to
study elementary friction, the elementary equation of fric-
tion. We shall write it down. It is very simple. Why are we
doing that ? Because it will come up.

Consider a stone falling through water or honey, figure 10.

It is falling due to a force which is a gradient of potential
energy, with a negative sign.

52
Figure 10 : Stone falling through a viscous fluid.

We are going to call its height φ, which is a stupid name for

the height but it is not a stupid name for a field. And as you
might expect what we are going to be talking about is fields.

Nevertheless, the analogy is φ is the height of the stone.

And there is a force on the stone which is related to its
potential energy. In this case it could just be the potential
energy of gravitation. But let’s call it V (φ) and not worry
about where it comes from.

Let’s assume the force is downward, which means V (φ) in-

creases in the upward direction. The force exerted on the
stone due to the potential energy is the derivative of V with
respect to φ, with a minus sign.

dV
F =− (12)
dφ

Now let’s write Newton’s equation for the stone, if there is

only that force. For simplicity we take the stone to have unit

53
mass. The generic equation F = ma in this case becomes

dV
− = φ̈ (13)
dφ

The force on the left-hand side is minus dV /dφ. And on

the right-hand side, the mass is 1 and the acceleration is φ̈,
that is the second derivative of φ with respect to time.

This is just F = ma. There is nothing special there. In

a uniform force field, which would mean the dV /dφ is a
constant, the stone will fall with a uniform acceleration. So
it will simply pick up speed, and continue to accelerate. Al-
right.

But that has ignored the viscosity of the fluid that the stone
is moving through, which let’s say is something like honey.
When we take this into account, that means that there is
another force on the left-hand side of equation (13).

The other force is zero if the object is at rest. The viscosity

exerts no force on an object at rest. It is the motion through
the fluid which creates the force. And so the extra force is
going to depend on the velocity. The larger the velocity, the
higher the force due to viscosity.

For simple fluids the force due to viscosity – which we some-

times simply call "the viscosity" – is actually proportional
to the velocity itself. So, adding this force, and reversing
sides in equation (13), we get

dV
φ̈ = −γ φ̇ − (14)
dφ

54
The term γ φ̇ has a minus sign in front of it because visco-
sity opposes motion. The coefficient γ is called the drag.

The potential energy increases with height, so, in equation

(14), the term −dV /dφ is a force pulling downward. The
stone is falling. Therefore φ̇ is negative, and −φ̇ is positive.
Thus the force from viscosity is pointing up, opposing that
from potential.

In the beginning, when the stone starts to fall, φ̇ is 0. So ba-

sically the stone starts out accelerating exactly as it would
without the viscosity term.

But then φ̇ will increase. And, unless as we move down the

force from potential gets bigger and bigger, because φ̇ is
increasing the force from viscosity will eventually balance
dV /dφ. At that point forces cancel, the acceleration be-
comes nil, and the corresponding φ̇ is called the terminal
velocity.

In particular, if the downward force is constant, for example

like the force of gravity which is uniform near the surface
of the Earth, then it is just F , and, if we take F to be
negative, equation (14) simplifies into

φ̈ = −γ φ̇ + F (15)

Then the terminal velocity satisfies

0 = −γ φ̇ + F (16)

or equivalently

55
F
φ̇ = (17)
γ

The terminal velocity has the same sign as F . Hence if F

is negative, the terminal velocity is negative. And this is
consistent, of course, with the fact that the stone is falling
through the honey.

That is the basic theory of friction in a nutshell. And what

does it do ? It slows things down.

What happens in particular if V (φ) has a shallow slope ?

To view it comfortably let’s plot V (φ) as in figure 11.

Figure 11 : Potential V (φ) as a function of φ.

Even though φ is the height, in this figure we represent it

on the horizontal axis in order to view V (φ) on the vertical
one. So we represented a potential which has a very small

56
gradient, or what we can call a shallow slope. Then the
stone can be imagined, as time passes, rolling slowly along
this curve 23 .

If not only the potential has a shallow slope but the visco-
sity coefficient γ is large, then the stone falling along the
potential energy curve of figure 11 will take a long long time
to roll down the hill.

That is the situation we will want to be in. And φ will not

be the position of a stone. It will be the value of a field.
But we will want to be in a situation where a field evolves
very slowly because of a lot of friction. And that will drive
inflation.

Let’s begin to use this model. And we will continue in the

next lesson.

Classical field theory

What we are going to consider is classical field theory. The

universe is filled with some field. The field is going to be a
scalar field.

Now where does this come from ? What scalar field ? It was
simply made up in order to be able to explain the isotropy
and homogeneity of the universe.

23. Remember that Galileo, to study comfortably falling objects,

let them rolling on wood planks with only a moderate slant. That
way he created only a small horizontal gradient.

57
At the time that this was created, it was a long shot rather
crazy idea, but it seems to be right. Here is the idea : in
addition to the electromagnetic field, the gravitational field
and all the usual fields, there is one more field in the uni-
verse. It is a scalar field. It is called the inflaton 24 and is
denoted φ.

Why inflaton ? Because it has to do with the inflation of

the universe when it began.

We are going to assume that φ is pretty much uniform in

space. We could assume that it is not uniform in space.
But, as the space expands, it will tend to stretch out the
variations in φ. So let’s just assume for simplicity that φ is
uniform in space.

Let’s think about the energy stored in a scalar field of this

type. Does the reader remember what the energy density
of a scalar field is ? We studied it in volume 3 of the col-
lection The Theoretical Minimum on special relativity and
classical field theory. It contains a kinetic term with a time
derivative. There are other terms which have to do with
gradients in space. So the energy density begins with

φ̇2
E= + terms related to gradients in space (18)
2

The term φ̇2 /2 is not the kinetic energy in the sense of mo-
vement in space but in the sense of due to time dependence.

As said, in general there are also differential terms in the

energy which have to do with gradients in space. Gradients
24. Note the spelling and pronounciation : inflaton, not inflation.

58
in space also store energy.

But since we have assumed that the field is homogeneous,

i.e. not varying in space, which we can justify – we will do
that later –, there are no gradients in space. Therefore the
only term involving derivatives is the one akin to a kinetic
energy.

Then the other thing that can be on the right-hand side

of equation (18) is a potential energy density V (φ). So the
final form of the energy density is

φ̇2
E= + V (φ) (19)
2

V (φ) is just a thing that is made up : different values of

the field have different energies. It doesn’t have to do with
derivatives, either in time or in space. Just in having a field
of a given magnitude there is an energy density associated
with it. It is called the field potential energy 25 .

It is not by accident that we use the same notation as in

the previous section. Here φ is a field. Before it was the
coordinate of a stone.

Now let’s follow the field in an expanding box. As usual

we look at a box that is expanding with the universe. How
big is the box at any given time ? The box has volume a3 ,
where a is the scale factor of expansion of the universe.

25. Pay attention to the fact that it is a density. Its units is energy
per volume.

59
To get the energy of the field in the box, we use the energy
density given by equation (19), and we multiply it by a3 .
" #
φ̇ 2
E = a3 + V (φ) (20a)
2

By a slight abuse of notation, in equation (19) E was an

energy density, whereas now in equation (20) it is a plain
energy.

But a is time dependent. So equation (20) is better expres-

sed as
" #
2
3 φ̇
E = a(t) + V (φ) (20b)
2

So we have a time-dependent expression for the energy in

the "expanding unit box" 26 .

You can think of equation (20), if you like, as formally,

mathematically being the same as the as the energy of a
particle : kinetic energy plus potential energy, except with
a coefficient in front which depends on time – something
you wouldn’t ordinarily write down.

We might have a funny situation where the mass of a par-

ticle might depend on time. But ordinarily we wouldn’t
write down, for the energy of a particle, an equation like
equation (20).

26. Remember the fictitious grid which we introduced in chapter 1.

The grid-unit is by definition 1. But its real value in meters is a(t).

60
Nevertheless that is what we have. We have an energy, ki-
netic plus potential.

Questions / answers session (5)

Q. : What would happen if φ depended not only on time

but also on space ?

A. : The general form of a scalar field is represented in

figure 12.

Figure 12 : General form of a scalar field, depending on time and

space.

In general, a field depends on time and space. If there was a

space-dependence, there would also be terms in the energy
equation (20), in the brackets, which have to do with gra-
dients in space. As I said, gradients in space also store
energy.

61
But we assume for the moment that the field is uniform in
space. If it is not, eventually it will be because the expan-
sion of the universe will stretch it out. It will iron out, so to
speak, the variations in space. Go back to figure 8, to see
the comparison with an inflated balloon.

So that is reasonable to assume that, after a period of time,

there are no gradients of the field in space.

Q. : So, just on a conceptual level, we are thinking of asso-

ciating this field with a quantum space of states ?

A. : Yes. But here we are not doing quantum field theory,

but classical field theory.

Furthermore we assume that the inflaton field is uniform in

space. Therefore all we need to do to find the total energy
is to multiply by the volume of the box, which is time de-
pendent.

Q. : Is there anything physical we can think of which would

correspond to this time-dependent and space-uniform field
φ, and that would help us figuring it out and figuring out
its evolution ?

A. : Well, I’m so used to it that I think of it as sort of ob-

vious. But we are going to find out. We have to solve the
equations of motion.

62
Your question is like asking before hand what will happen
to a particle whose energy is

φ̇2
E= + V (φ) (21)
2

Is it moving up or moving down ?

That will depend on the initial conditions. It will depend

on how long we wait. And it will especially depend on the
sign of dV /dφ, namely the sign of the force.

And whatever it will be doing, it will be doing it every-

where uniformly in space simultaneously.

Now we can back off that, and study what happens when
it is not uniform in space. But this is the easy problem for
us to study.

Now, how do we find the equation of motion when we know

the kinetic energy and we know the potential energy ? There
are various ways we could do it. But the most efficient way
is through Lagrange’s equations.

Let’s write the lagrangian. As the reader remembers, in a

simple situation like here, where the energy is made of two
clearly separated terms, the lagrangian is the difference bet-
ween the kinetic energy and the potential energy 27 .

27. Of course, when talking of a field, we are talking first of all of

the density of energy at a point in time and space. In due course, we
shall integrate it to get a plain energy.

63
So we pick up our copy of volume 1 in the collection The
Theoretical Minimum on classical mechanics, and we rea-
dily write
" #
φ̇ 2
L = a(t)3 − V (φ) (22)
2

where φ is like the coordinate – what we called the degree

of freedom. And equation (22) is the lagrangian.

The only new thing that wouldn’t be there for an ordinary

particle is this coefficient a(t) to the cube in front.

From there, let’s work out Lagrange’s equations. When

there is no space-dependence, Lagrange’s equations for a
field are simply

d ∂L ∂L
= (23)
dt ∂ φ̇ ∂φ

The calculation begins with

∂L
= a(t)3 φ̇ (24)
∂ φ̇

Then we have to take its time derivative. Equation (23)

rewrites

d dV
a(t)3 φ̇ = −a(t)3 (25)
dt dφ

The derivative of V with respect to φ can be written as a

plain derivative, as opposed to a partial one, because there

64
are no other independent variables driving V .

We could call the derivative dV /dφ, the force, or more ac-

curately minus the force. But notice that there is the extra
explicit time-dependent coefficient a(t)3 appearing on both
sides of equation (25).

Therefore it doesn’t quite look like Newton’s equations. It

would be Newton’s equations if a was constant. But a is
not constant. So let’s work it out and see what it is.

For the left-hand side, differentiating with respect to time

the product, we get

a3 φ̈ + φ̇ 3a2 ȧ (26)

If, on the right-hand side, we write −dV /dφ = F , we get

for equation (25)

a3 φ̈ + 3a2 ȧ φ̇ = a3 F (27)

It is very tempting, since it appears on both sides, to divide

by a3 . Let’s do it

ȧ
φ̈ + 3 φ̇ = F (28)
a

What is ȧ over a called in cosmology ? The Hubble constant.

Remember that it is a constant in space, but not necessarily
in time. It can be and usually is time-dependent.

So equation (28) now takes the form

φ̈ + 3 H φ̇ = F (29)

65
We recognize the same form as equation (14). Hence, we
made the junction with our preliminary work on friction
and viscosity :

Equation (29) is exactly the same equation as that of a stone

falling through a fluid with a viscosity coefficient, i.e. a drag,
γ = 3H, and there is a force which is minus the gradient
of V .

Remember that the potential energy – or energy density to

be precise – V (φ) is a made up thing which we added to
the energy of the inflaton field φ, equation (19).

So our model for the way this field evolves can be envisioned
by just supposing that φ was the position of a particle on
a hill, where the height of the hill was V , and where the
gradient of the altitude of the hill was the force −dV /dφ,
figure 13.

Figure 13 : Inflaton field φ and evolution of the universe.

The evolution of the universe can be compared to that of

66
a ball rolling down the hill, except that there is a viscous
drag force proportional to the Hubble expansion rate. That
is why we went through the exercise of studying elementary
classical friction.

If the Hubble friction is strong, that is, if H is large, and if

the hill in figure 13 is reasonably flat, not terribly flat, just
moderately slanted, then our model is like a ball moving
through motor oil on a cold day in Wisconsin.

If just left to its own devices without the friction term, the
ball might roll down the hill in a few seconds. But with a
large friction it could take years to roll down the hill, de-
pending on the magnitude of the friction.

So the term 3H φ̇ is called the cosmic friction term in the

equation of motion of the scalar field φ, equation (29). It
has the simple effect of slowing down the evolution of the
system and keeping the ball from rolling down the hill.

In the next lesson, we are going to use this to study the cos-
mology of a universe which contains a field like φ. In other
words, we are going to look at the Friedmann’s equation
with an energy density which is given by

" #
φ̇2
E = a(t)3 + V (φ) (30)
2

We shall study how the universe expands and evolves un-

der the influence of an energy density resulting from a field
which is slowly slowly slowly rolling down the hill.

67
That is the phenomenon of inflation, i.e. the way the uni-
verse responds to this very small slowly moving inflaton
field φ.

68