Introduction

In the previous lesson we studied a measure-preserving dynamical system:

$$\left(X,\mathcal F,\mu,T\right)$$

where:

$$T:X\to X$$

describes the evolution of the system.

Traditionally, one studies:

$$x,T(x),T^2(x),\ldots$$

the orbit of a point.

This is called the geometric viewpoint.

Around 1930, a revolutionary idea emerged.

Instead of studying points, study functions.

Instead of asking:

Where does a point move?

ask:

How does a function change under the dynamics?

This leads to the Koopman Operator.

The Koopman Operator is one of the most important ideas in ergodic theory because it transforms:

Dynamics → Linear Algebra

and eventually:

Dynamics → Functional Analysis → Operator Algebras

This is one of the first major conceptual steps toward Connes.

---

The Classical View

Suppose:

$$T:X\to X$$

is a dynamical system.

For a point:

$$x\in X$$

we observe:

$$x,T(x),T^2(x),\ldots$$

This orbit may be:

• periodic
• chaotic
• ergodic
• mixing

Directly studying orbits is often difficult.

---

A Different Perspective

Suppose:

$$f:X\to\mathbb R$$

is a measurable function.

Think of:

$$f$$

as measuring:

• temperature
• pressure
• density
• velocity

Instead of following points, follow measurements.

---

What Happens After One Step?

Initially:

$$f(x)$$

After applying:

$$T$$

the system moves to:

$$T(x)$$

The measurement becomes:

$$f(T(x))$$

This new function is:

$$f\circ T$$

This simple observation creates an operator.

---

Definition of the Koopman Operator

The Koopman operator:

$$U_T$$

acts on functions by:

$$U_Tf=f\circ T$$

In other words:

$$\left(U_Tf\right)(x)=f(T(x))$$

---

Why This Is Surprising

Notice:

$$T$$

may be highly nonlinear.

Yet:

$$U_T$$

is always linear.

---

Proof of Linearity

Suppose:

$$f,g$$

are functions and:

$$a,b$$

are constants.

Then:

$$U_T(af+bg)=(af+bg)\circ T$$

Thus:

$$U_T(af+bg)=a(f\circ T)+b(g\circ T)$$

which is:

$$aU_Tf+bU_Tg$$

Therefore:

$$U_T$$

is linear.

---

The Big Miracle

A nonlinear dynamical system produces a linear operator.

This allows us to apply:

• Linear Algebra
• Functional Analysis
• Spectral Theory

to study dynamics.

This idea transformed ergodic theory.

---

Example 1

Let:

$$X=[0,1]$$

and:

$$T(x)=2x\pmod1$$

Take:

$$f(x)=x$$

Then:

$$U_Tf(x)=f(T(x))$$

which becomes:

$$U_Tf(x)=2x\pmod1$$

The operator transforms one observable into another.

---

Example 2

Let:

$$f(x)=\sin(2\pi x)$$

Then:

$$U_Tf(x)=\sin(4\pi x)$$

The frequency doubles.

Dynamics becomes a transformation of functions.

---

Why Hilbert Spaces Appear

Recall:

$$L^2(X,\mu)$$

the space of square-integrable functions.

This is a Hilbert space.

Since:

$$U_T$$

acts on functions,

it naturally acts on:

$$L^2(X,\mu)$$

Thus dynamics becomes an operator acting on a Hilbert space.

This is a huge conceptual leap.

---

Measure Preservation Matters

Suppose:

$$T$$

preserves measure.

Then:

$$\mu(T^{-1}(A))=\mu(A)$$

for all measurable sets.

This implies:

$$|U_Tf|_2=|f|_2$$

for every:

$$f\in L^2(X,\mu)$$

---

Proof Sketch

Compute:

$$|U_Tf|_2^2=\int |f(T(x))|^2,d\mu(x)$$

Measure preservation allows a change of variables:

$$\int |f(T(x))|^2,d\mu(x)=\int |f(x)|^2,d\mu(x)$$

Therefore:

$$|U_Tf|_2=|f|_2$$

---

Unitary Operators

An operator satisfying:

$$|Uf|=|f|$$

is called unitary.

Thus:

Measure-preserving dynamics produces a unitary operator.

This is an extremely important fact.

---

Why Unitary Operators Matter

Unitary operators possess:

• Eigenvalues
• Eigenvectors
• Spectral decompositions

These are precisely the tools of quantum mechanics and functional analysis.

---

Eigenfunctions

Suppose:

$$U_Tf=\lambda f$$

Then:

$$f(T(x))=\lambda f(x)$$

Such functions are called eigenfunctions of the dynamical system.

They reveal hidden structure.

---

Example

For circle rotations:

$$T(x)=x+\alpha\pmod1$$

consider:

$$f_n(x)=e^{2\pi i n x}$$

Then:

$$U_Tf_n=e^{2\pi i n\alpha}f_n$$

Thus:

$$f_n$$

is an eigenfunction.

The corresponding eigenvalue is:

$$e^{2\pi i n\alpha}$$

---

Spectral Theory Appears

We can now study:

• Eigenvalues
• Spectra
• Decompositions

of:

$$U_T$$

instead of studying:

$$T$$

directly.

This is the birth of spectral ergodic theory.

---

Dynamics Through Spectra

Classical geometry studies points.

Spectral theory studies frequencies.

The Koopman operator transforms dynamics into a spectral problem.

This shift becomes one of the defining themes of twentieth-century mathematics.

---

Connection to Birkhoff’s Theorem

Recall:

$$\frac1n\sum_{k=0}^{n-1}f(T^k x)$$

Notice:

$$f(T^k x)=U_T^k f(x)$$

Thus Birkhoff averages become:

$$\frac1n\sum_{k=0}^{n-1}U_T^k f$$

The ergodic theorem is therefore a statement about operators.

---

von Neumann’s Ergodic Theorem

Before Birkhoff proved his famous theorem,

John von Neumann proved:

$$\frac1n\sum_{k=0}^{n-1}U_T^k f$$

converges in:

$$L^2$$

This result is one of the first great successes of operator methods in dynamics.

---

Why Analysts Love Koopman Theory

The Koopman viewpoint converts:

Dynamical System

Operator Theory

Orbit

Function

Iteration

Operator Power

Dynamics

Spectral Theory

Ergodicity

Operator Property

This translation is extraordinarily powerful.

---

Why This Matters for Quantum Mechanics

Quantum mechanics is built on:

• Hilbert spaces
• Unitary operators
• Spectral decompositions

The Koopman operator already contains all of these ingredients.

This is one reason ergodic theory and quantum theory developed side by side.

---

The Philosophical Shift

A major theme is emerging:

Classical mathematics studies points.

Modern mathematics studies operators.

The Koopman operator is one of the earliest places where this shift becomes visible.

---

Connection to von Neumann Algebras

Once we have operators acting on:

$$L^2(X,\mu)$$

we can begin forming algebras generated by these operators.

These become:

• Operator algebras
• von Neumann algebras

The road to Connes has now officially begun.

---

Connection to Alain Connes

Connes’ early work was deeply rooted in:

• Ergodic theory
• Operator algebras
• Dynamical systems

The Koopman operator provides one of the first bridges between:

measure-preserving transformations

and

operator-algebraic structures.

Many von Neumann algebras arise from exactly this construction.

In a very real sense, the Koopman operator is one of the first places where classical measure theory begins transforming into noncommutative geometry.

---

Key Concepts Learned

By the end of this lesson you should understand:

• The Koopman operator is defined by:

$$U_Tf=f\circ T$$

• Every dynamical system induces a linear operator.
• Measure-preserving transformations produce unitary operators.
• Koopman operators act naturally on:

$$L^2(X,\mu)$$

• Eigenfunctions reveal hidden dynamical structure.
• Spectral theory can be used to study dynamics.
• Birkhoff’s theorem can be reformulated using operators.
• The Koopman operator forms one of the first bridges from ergodic theory to operator algebras.

---

Looking Ahead

Measure Theory Lesson 38: Spectral Theory of Dynamical Systems

Next we study one of the deepest ideas in ergodic theory:

Can we classify a dynamical system by the spectrum of its Koopman operator?

This question leads directly into spectral theory, Hilbert spaces, unitary operators, and the mathematical framework that ultimately inspired much of von Neumann’s and Connes’ work. From this point onward, we begin moving decisively from measure theory into operator theory.