Introduction

One of the central themes of analysis is the study of limits.

In calculus, we typically encounter limits of numbers:

$$a_n \to a$$

For example:

$$\frac{1}{n}\to0$$

Measure theory introduces a much richer question:

What does it mean for a sequence of functions to converge?

Suppose we have functions:

$$f_1,f_2,f_3,\ldots$$

and a limiting function:

$$f$$

What exactly should:

$$f_n \to f$$

mean?

Unlike sequences of numbers, there is not just one notion of convergence.

There are many.

Understanding the differences between them is one of the most important parts of measure theory.

The convergence theorems that follow later depend critically on which type of convergence is being used.

---

Why Multiple Notions of Convergence Exist

Consider the sequence:

$$f_n(x)=x^n$$

on:

$$[0,1]$$

For every:

$$0\le x<1$$

we have:

$$x^n\to0$$

However:

$$1^n=1$$

for all:

$$n$$

Thus the limiting function is:

$$f(x)=\begin{cases}
0,&0\le x<1\
1,&x=1
\end{cases}$$

At first glance, this seems straightforward.

However, many questions arise:

• How quickly does convergence occur?
• Does it occur uniformly?
• Does it occur almost everywhere?
• Does it preserve integrals?

To answer these questions, mathematicians developed several notions of convergence.

---

Pointwise Convergence

Definition

A sequence:

$$f_n$$

converges pointwise to:

$$f$$

if for every:

$$x\in X$$

we have:

$$\lim_{n\to\infty}f_n(x)=f(x)$$

This is the most intuitive form of convergence.

For each individual point, we examine the behavior of the sequence.

---

Example

Consider:

$$f_n(x)=\frac{x}{n}$$

on:

$$[0,1]$$

Fix any:

$$x$$

Then:

$$\frac{x}{n}\to0$$

Therefore:

$$f_n(x)\to0$$

for every:

$$x$$

Thus:

$$f_n\to0$$

pointwise.

---

Limitation of Pointwise Convergence

Pointwise convergence can be surprisingly weak.

Different points may converge at very different speeds.

Nothing guarantees that the convergence behaves uniformly across the entire domain.

This motivates a stronger concept.

---

Uniform Convergence

Definition

A sequence:

$$f_n$$

converges uniformly to:

$$f$$

if:

$$\sup_{x\in X}|f_n(x)-f(x)|\to0$$

as:

$$n\to\infty$$

This means that the largest error anywhere in the space approaches zero.

---

Intuition

Pointwise convergence says:

Every point eventually behaves.

Uniform convergence says:

The entire function behaves simultaneously.

Uniform convergence is much stronger.

---

Example

Consider:

$$f_n(x)=\frac{x}{n}$$

on:

$$[0,1]$$

We have:

$$|f_n(x)-0|=\frac{x}{n}$$

The largest possible value occurs at:

$$x=1$$

Thus:

$$\sup_{x\in[0,1]}\frac{x}{n}=\frac1n$$

Since:

$$\frac1n\to0$$

the convergence is uniform.

---

Pointwise But Not Uniform

Consider again:

$$f_n(x)=x^n$$

on:

$$[0,1]$$

We already know:

$$f_n(x)\to f(x)$$

where:

$$f(x)=\begin{cases}
0,&0\le x<1\
1,&x=1
\end{cases}$$

This convergence is pointwise.

However, it is not uniform.

Near:

$$x=1$$

the convergence becomes arbitrarily slow.

The maximum error never approaches zero uniformly.

Thus:

$$f_n\to f$$

pointwise but not uniformly.

---

Almost Everywhere Convergence

Measure theory introduces a new idea.

Sometimes a sequence fails to converge at a few exceptional points.

If those exceptional points form a measure-zero set, we often choose to ignore them.

---

Definition

A sequence:

$$f_n$$

converges almost everywhere to:

$$f$$

if:

$$f_n(x)\to f(x)$$

for all:

$$x$$

except on a set of measure zero.

We write:

$$f_n\to f \quad \text{a.e.}$$

---

Example

Suppose:

$$f_n(x)=0$$

for every:

$$x\neq0$$

and:

$$f_n(0)=n$$

Then:

$$f_n(x)\to0$$

for every:

$$x\neq0$$

Convergence fails only at:

$$x=0$$

Since:

$$m({0})=0$$

we obtain:

$$f_n\to0 \quad \text{a.e.}$$

---

Why Almost Everywhere Convergence Matters

Measure theory is largely concerned with integration.

Integration cannot detect measure-zero sets.

Therefore almost everywhere convergence is often sufficient for important results.

Many major theorems are stated using almost everywhere convergence rather than pointwise convergence.

---

Convergence in Measure

Another notion focuses on the size of the set where the approximation is poor.

---

Definition

A sequence:

$$f_n$$

converges in measure to:

$$f$$

if for every:

$$\varepsilon>0$$

we have:

$$\mu\left({x:|f_n(x)-f(x)|>\varepsilon}\right)\to0$$

as:

$$n\to\infty$$

---

Interpretation

For large:

$$n$$

the set where:

$$f_n$$

and:

$$f$$

differ significantly becomes very small.

Convergence in measure does not require every point to converge.

Only the badly behaved region must shrink.

---

Relationships Between Convergence Types

The different notions of convergence are related, but they are not equivalent.

Generally:

$$\text{Uniform Convergence}
\implies
\text{Pointwise Convergence}
$$

and:

$$\text{Pointwise Convergence}
\implies
\text{Almost Everywhere Convergence}
$$

when convergence holds everywhere.

However, the reverse implications usually fail.

---

Important Warning

Pointwise convergence alone does not allow us to interchange limits and integrals.

Many students initially assume:

$$\lim_{n\to\infty}\int f_n,d\mu=\int \lim_{n\to\infty}f_n,d\mu$$

whenever:

$$f_n\to f$$

pointwise.

This is false.

Measure theory’s great convergence theorems exist precisely because additional conditions are needed.

---

A Famous Example

Consider:

$$f_n(x)=n\mathbf{1}_{(0,\frac1n)}(x)$$

on:

$$[0,1]$$

For every:

$$x>0$$

eventually:

$$x>\frac1n$$

so:

$$f_n(x)=0$$

Therefore:

$$f_n(x)\to0$$

for every:

$$x>0$$

and also at:

$$x=0$$

if we define the indicator appropriately.

Thus:

$$f_n\to0$$

pointwise.

However:

$$\int_0^1f_n(x),dx=n\left(\frac1n\right)=1$$

for every:

$$n$$

Hence:

$$\lim_{n\to\infty}\int_0^1f_n(x),dx=1$$

while:

$$\int_0^1\lim_{n\to\infty}f_n(x),dx=0$$

The limit and integral are not interchangeable.

This example motivates the major convergence theorems that follow.

---

Why Convergence Is Central

Much of analysis is built around sequences.

We approximate:

• functions
• solutions of differential equations
• probability distributions
• operators

using sequences.

To ensure our approximations behave correctly, we must understand exactly how convergence occurs.

The next three lessons provide the tools that make these approximations rigorous.

---

Connection to Probability

Suppose:

$$X_n$$

are random variables.

Then:

• pointwise convergence becomes sample-path convergence
• almost everywhere convergence becomes almost sure convergence
• convergence in measure becomes convergence in probability

These concepts are foundational throughout modern probability theory.

---

Connection to Alain Connes

Much of functional analysis and operator theory revolves around convergence.

When Connes studies operator algebras, various notions of convergence appear:

• norm convergence
• strong operator convergence
• weak operator convergence

The idea that multiple types of convergence coexist begins right here in measure theory.

---

Key Concepts Learned

By the end of this lesson you should understand:

• Function sequences can converge in different ways.
• Pointwise convergence examines each point individually.
• Uniform convergence controls the largest error globally.
• Almost everywhere convergence ignores measure-zero exceptions.
• Convergence in measure focuses on shrinking bad sets.
• Uniform convergence is stronger than pointwise convergence.
• Pointwise convergence alone does not preserve integrals.
• Understanding convergence is essential before studying the major convergence theorems.

---

Looking Ahead

In the next lesson we prove the first great theorem of measure theory:

Lesson 12: The Monotone Convergence Theorem

This theorem provides the first rigorous condition under which limits and integrals can be exchanged, and it is one of the foundational results upon which all modern integration theory is built.