Measure Spaces — Mathematics Knowledge System

Overview

A measure space is the fundamental structure of measure theory: a set X equipped with a σ-algebra ℱ and a function μ : ℱ → [0, ∞] that assigns a 'size' to every measurable set in a consistent way. This framework generalises length on ℝ, area in ℝ², counting on discrete sets, and probability on sample spaces, all within a single axiomatic system.

Intuition

Imagine a weighing scale for subsets of X. The σ-algebra tells you which subsets you are 'allowed' to weigh, and the measure μ gives each allowed subset a non-negative weight. Countable additivity — the key axiom — says the total weight of disjoint pieces equals the sum of the individual weights, just as physical mass behaves when you cut an object into separate pieces.

Formal Definition

Definition

A measure space is a triple (X, ℱ, μ) where X is a set, ℱ is a σ-algebra on X, and μ is a measure: a function satisfying the following axioms.

\mu : \mathcal{F} \to [0, \infty]

μ maps measurable sets to extended non-negative reals

\mu(\emptyset) = 0

Axiom 1 — null empty set

\text{If } A_1, A_2, \ldots \in \mathcal{F} \text{ are pairwise disjoint, then } \mu\!\left(\bigsqcup_{n=1}^{\infty} A_n\right) = \sum_{n=1}^{\infty} \mu(A_n)

Axiom 2 — countable additivity (σ-additivity)

Notation

Notation	Meaning
$(X, \mathcal{F}, \mu)$	Measure space
$\mu(A)$	Measure of a set A ∈ ℱ
$\text{a.e.}$	Almost everywhere — a property holding except on a set of measure zero
$\mu \ll \nu$	μ is absolutely continuous with respect to ν

Properties

Monotonicity

A \subseteq B \implies \mu(A) \leq \mu(B)

Condition: A, B ∈ ℱ

Countable subadditivity

\mu\!\left(\bigcup_{n=1}^{\infty} A_n\right) \leq \sum_{n=1}^{\infty} \mu(A_n)

Condition: A₁, A₂, … ∈ ℱ (not necessarily disjoint)

Continuity from below

A_n \nearrow A \implies \mu(A_n) \nearrow \mu(A)

Condition: Increasing sequence Aₙ ↗ A in ℱ

Continuity from above

A_n \searrow A,\; \mu(A_1) < \infty \implies \mu(A_n) \searrow \mu(A)

Condition: Decreasing sequence Aₙ ↘ A in ℱ with μ(A₁) < ∞

Worked Examples

Define μ(A) = |A| (cardinality of A) for A ⊆ ℕ, with μ(A) = ∞ for infinite A.
$\mu(A) = \#A$
μ(∅) = 0 since ∅ has no elements.
$\mu(\emptyset) = 0 \checkmark$
For pairwise disjoint A₁, A₂, … ⊆ ℕ, the elements of their union are exactly the elements of each Aₙ, with no repetition, so total count is the sum of individual counts.
$\mu\!\left(\bigsqcup_n A_n\right) = \sum_n \mu(A_n) \checkmark$

Answer: (ℕ, 2^ℕ, μ_#) is a valid measure space.

Practice Problems

Difficulty 6/10

Prove the inclusion–exclusion formula: for A, B ∈ ℱ with μ(A), μ(B) < ∞, μ(A ∪ B) = μ(A) + μ(B) − μ(A ∩ B).

Difficulty 7/10

Show that if μ(Aₙ) = 0 for all n ∈ ℕ, then μ(∪ₙ Aₙ) = 0.

Difficulty 8/10

Give an example showing that continuity from above fails without the assumption μ(A₁) < ∞.

Common Mistakes

Common Mistake

Confusing σ-additivity with finite additivity

σ-additivity requires the property to hold for countably infinite disjoint collections. Finite additivity alone does not imply σ-additivity, and many pathological set functions satisfy the latter but not the former.

Common Mistake

Assuming continuity from above always holds

Continuity from above requires μ(A₁) < ∞. Without this, the property can fail spectacularly (e.g., counting measure on tail sets of ℕ).

Common Mistake

Treating null sets as 'nonexistent'

A null set has measure zero but may be non-empty (e.g., the Cantor set has Lebesgue measure zero but is uncountable). Properties holding 'almost everywhere' still allow failure on null sets.

Quiz

Historical Background

Henri Lebesgue's 1902 thesis introduced the measure-space framework to solve the deficiencies of the Riemann integral — in particular, the inability to integrate many pointwise limits of functions. His key insight was to measure the pre-image of intervals rather than partition the domain, yielding a far more powerful integration theory.

1902
Lebesgue introduces his measure and integral
Henri Lebesgue
1910s
Carathéodory develops outer measure and the extension theorem
Constantin Carathéodory
1933
Kolmogorov axiomatises probability theory using measure spaces
Andrei Kolmogorov

Summary

A measure space (X, ℱ, μ) consists of a set, a σ-algebra, and a countably additive non-negative function μ on ℱ.
Key derived properties include monotonicity, countable subadditivity, and continuity from below/above.
Continuity from above requires μ(A₁) < ∞; this hypothesis cannot be dropped.
A measure space is finite if μ(X) < ∞ and σ-finite if X is a countable union of finite-measure sets.
Null sets (sets of measure zero) play a central role — properties holding off null sets hold 'almost everywhere'.

References

BookFolland, G.B. — Real Analysis, 2nd ed. (1999), §1.1–§1.2
BookRoyden, H.L. & Fitzpatrick, P.M. — Real Analysis, 4th ed. (2010), Chapter 17
WebsiteWikipedia — Measure (mathematics)