The Central Problem of Measure Theory

At first glance, measure theory seems simple.

We have a set:

$$X$$

and we want to assign sizes to subsets of $$X$$.

For example:

$$X = [0,1]$$

and we would like to assign lengths to all subsets of the interval.

However, something surprising happens.

Not every subset can be assigned a sensible measure.

In fact, mathematicians proved that there exist subsets of the real numbers for which no reasonable notion of length can exist.

This means that if we want a consistent theory, we must be selective.

We need to decide:

Which subsets are allowed to be measured?

The answer is the sigma-algebra.

---

Why Ordinary Collections of Sets Are Not Enough

Suppose we decide to measure the interval:

$$[0,1]$$

and all of its subintervals.

Examples:

$$[0,\frac{1}{2}]$$

$$[\frac{1}{4},\frac{3}{4}]$$

$$(0,\frac{1}{3})$$

This seems reasonable.

But then we may want to combine intervals.

For example:

$$[0,\frac{1}{4}] \cup [\frac{3}{4},1]$$

should also be measurable.

Similarly:

$$[0,1] \setminus [0,\frac{1}{2}]$$

should be measurable.

Likewise, infinite unions should be measurable.

If:

$$A_1,A_2,A_3,\ldots$$

are measurable, then surely:

$$\bigcup_{n=1}^{\infty} A_n$$

should also be measurable.

Therefore our collection of measurable sets must be closed under certain operations.

---

The Idea of Closure

A collection of sets is called closed under an operation if performing that operation never takes us outside the collection.

Example:

The even integers are closed under addition.

Since:

$$2+4=6$$

and:

$$6$$

is still even.

Measure theory needs a collection of sets that remains stable under operations such as:

• complements
• unions
• intersections

especially countably infinite ones.

---

Definition of a Sigma-Algebra

Let:

$$X$$

be a set.

A collection of subsets:

$$\mathcal{F}$$

is called a sigma-algebra if the following three properties hold.

---

Property 1

The whole space belongs to the collection.

$$X \in \mathcal{F}$$

If we can measure smaller pieces, we should certainly be able to measure the entire space.

---

Property 2

Closed under complements.

If:

$$A \in \mathcal{F}$$

then:

$$A^c \in \mathcal{F}$$

where:

$$A^c = X \setminus A$$

This means that whenever a set is measurable, everything outside it is also measurable.

---

Property 3

Closed under countable unions.

If:

$$A_1,A_2,A_3,\ldots \in \mathcal{F}$$

then:

$$\bigcup_{n=1}^{\infty} A_n \in \mathcal{F}$$

This is the defining feature that distinguishes sigma-algebras from simpler structures.

The Greek letter sigma refers to countable sums and countable operations.

---

Why Countable Unions?

One might ask:

Why not require closure under arbitrary unions?

The answer is that measure theory is built around countable processes.

Most of analysis relies on:

• sequences
• infinite series
• limits

all of which are countable.

Countable operations provide enough power for analysis while avoiding many pathological problems.

---

Immediate Consequences

The three axioms imply many additional properties.

---

The Empty Set Is Measurable

Since:

$$X \in \mathcal{F}$$

and sigma-algebras are closed under complements:

$$X^c = \emptyset$$

Therefore:

$$\emptyset \in \mathcal{F}$$

---

Closed Under Countable Intersections

Suppose:

$$A_1,A_2,A_3,\ldots \in \mathcal{F}$$

Then:

$$A_1^c,A_2^c,A_3^c,\ldots \in \mathcal{F}$$

Taking a countable union:

$$\bigcup_{n=1}^{\infty} A_n^c \in \mathcal{F}$$

Taking a complement again:

$$\left(\bigcup_{n=1}^{\infty}A_n^c\right)^c \in \mathcal{F}$$

Using De Morgan’s Law:

$$\left(\bigcup_{n=1}^{\infty}A_n^c\right)^c = \bigcap_{n=1}^{\infty}A_n$$

Therefore:

$$\bigcap_{n=1}^{\infty}A_n \in \mathcal{F}$$

So countable intersections automatically become measurable.

---

Example 1: Trivial Sigma-Algebra

Consider:

$$X={1,2,3}$$

Define:

$$\mathcal{F}={\emptyset,X}$$

This satisfies all three axioms.

Therefore it is a sigma-algebra.

It contains only the absolutely necessary sets.

---

Example 2: Power Set

The power set of:

$$X$$

is the collection of all subsets of $$X$$.

It is denoted:

$$\mathcal{P}(X)$$

For:

$$X=\{1,2,3\}$$

the power set is:

$${\emptyset,\{1\},\{2\},\{3\},\{1,2\},\{1,3\},\{2,3\},X}$$

Since all possible subsets are included, closure is automatic.

Therefore:

$$\mathcal{P}(X)$$

is always a sigma-algebra.

---

Example 3: A Non-Trivial Sigma-Algebra

Let:

$$X=\{1,2,3,4\}$$

Consider:

$$\mathcal{F}=\{\emptyset,\{1,2\},\{3,4\},X\}$$

Check the axioms:

Complements:

$$\{1,2\}^c=\{3,4\}$$

Countable unions:

$$\{1,2\}\cup\{3,4\}=X$$

Everything remains inside the collection.

Thus this is a sigma-algebra.

---

Why Sigma-Algebras Matter in Probability

Suppose we toss a coin.

The sample space is:

$$\Omega=\{H,T\}$$

A probability measure assigns probabilities to events.

What are the events?

They are exactly the elements of a sigma-algebra.

For example:

$$\mathcal{F}=\{\emptyset,{H},{T},\Omega\}$$

The probability measure is defined on:

$$\mathcal{F}$$

not directly on individual outcomes.

Thus modern probability is built upon sigma-algebras.

---

Why Sigma-Algebras Matter in Measure Theory

A measure is not defined on every subset.

Instead it is defined on:

$$\mathcal{F}$$

where:

$$\mathcal{F}$$

is a sigma-algebra.

This leads to the formal definition:

A measure space consists of:

$$\left(X,\mathcal{F},\mu\right)$$

where:

• $$X$$ is the underlying space
• $$\mathcal{F}$$ is a sigma-algebra
• $$\mu$$ is a measure

This triple is the fundamental object of measure theory.

---

The Hidden Reason Sigma-Algebras Exist

Without sigma-algebras, paradoxes appear.

In the early twentieth century, mathematicians discovered bizarre subsets of the real line known as non-measurable sets.

These sets cannot consistently be assigned lengths.

The most famous example is the Vitali set.

Sigma-algebras protect measure theory from such pathologies.

They define a universe of sets on which measurement behaves correctly.

---

A Geometric Interpretation

Think of a sigma-algebra as a collection of “observable” sets.

Measure theory says:

Not every imaginable subset can be measured.

Only those belonging to the sigma-algebra are visible to the measuring process.

This viewpoint becomes extremely important later.

In probability:

• measurable sets are observable events.

In quantum theory:

• measurable quantities become observables.

In Connes’ work:

• the notion of observability becomes encoded in operator algebras.

Thus the sigma-algebra is the first hint that not everything in mathematics is necessarily measurable.

---

Key Concepts Learned

By the end of this lesson you should understand:

• Not every subset can be measured.
• A sigma-algebra specifies which sets are measurable.
• A sigma-algebra must contain the whole space.
• It must be closed under complements.
• It must be closed under countable unions.
• Countable intersections follow automatically.
• Probability measures and ordinary measures are defined on sigma-algebras.
• Measure spaces are triples of the form:

$$\left(X,\mathcal{F},\mu\right)$$

---

Looking Ahead

In the next lesson:

Lesson 4: Measurable Spaces

we will study the pair:

$$\left(X,\mathcal{F}\right)$$

by itself and understand why a space equipped with a sigma-algebra is the true starting point of modern measure theory.