Outer measure μ* is already defined on every subset of ℝ. A student proposes that Lebesgue measure can be obtained by simply using μ* on all subsets of ℝ, since it already assigns a value to each one. What is the fundamental problem with this approach?
AOuter measure is only defined on open subsets of ℝ, so it cannot be used on arbitrary sets
BOuter measure fails to be σ-additive on all subsets of ℝ — for pathological (non-measurable) sets, disjoint union does not equal the sum of outer measures — so it is not a valid measure on the full power set
COuter measure always overestimates the true length of sets, so using it directly would give the wrong values
DRestricting to all subsets would work mathematically, but it would be inconsistent with the existing Borel σ-algebra
The core problem is that outer measure is subadditive but not σ-additive in general: for disjoint sets A and B, μ*(A ∪ B) ≤ μ*(A) + μ*(B), but equality can fail for pathological sets (the Vitali set is the classic example). A valid measure requires σ-additivity — without it, the 'measure' cannot consistently assign sizes. Carathéodory's theorem identifies the subcollection of sets where σ-additivity does hold, and shows this subcollection is a σ-algebra, solving the problem.
Question 2 Multiple Choice
A set E satisfies the Carathéodory condition: for every set A, μ*(A) = μ*(A ∩ E) + μ*(A ∩ Eᶜ). What does this condition capture geometrically?
AE has finite outer measure — it is a bounded set
BE contains no pathological sub-structure — all its subsets are measurable
CE 'splits' every test set A cleanly: the two pieces (inside E and outside E) have outer measures that add up correctly, with no loss or gain from subadditivity
DThe complement of E has the same outer measure as E, ensuring symmetry
The Carathéodory condition is a splitting test: E must divide every possible test set A into two pieces whose outer measures sum to exactly μ*(A). This is the geometric content — E acts as a 'clean divider' of the space. Pathological sets fail this test: they create phantom 'interference' where the pieces' outer measures sum to more than the whole. The condition is the precise formalization of 'behaves well under measurement,' and its power is that this collection of well-behaved sets turns out to form a σ-algebra.
Question 3 True / False
The collection of μ*-measurable sets is guaranteed to form a σ-algebra: it contains the empty set, is closed under complements, and is closed under countable unions.
TTrue
FFalse
Answer: True
This is the first main conclusion of Carathéodory's theorem, and it is non-trivial — it requires proof. Closure under complements is immediate from the symmetry of the condition (if E splits every A correctly, so does Eᶜ). Closure under countable unions requires careful argument, especially showing that the Carathéodory condition is preserved through countable operations. The fact that a condition defined on individual sets automatically yields closure under countably infinite operations is what makes the theorem powerful.
Question 4 True / False
The Carathéodory condition is merely a technical formality — any subcollection of sets on which outer measure is finitely additive would automatically form a σ-algebra.
TTrue
FFalse
Answer: False
The Carathéodory condition is specifically engineered to produce a σ-algebra, and this requires proof. A subcollection of sets where outer measure happens to be additive need not be closed under countable unions, complements, or intersections — σ-algebra structure requires all these. The Carathéodory condition's particular form (splitting every test set A, not just sets within the collection) is what forces σ-algebra closure. It is a precisely crafted condition, not an automatic one.
Question 5 Short Answer
Explain why outer measure alone is insufficient to construct Lebesgue measure, and what role the Carathéodory condition plays in resolving this.
Think about your answer, then reveal below.
Model answer: Outer measure is defined on all subsets of ℝ but fails to be σ-additive in general — for pathological sets, disjoint additivity breaks down. A measure requires σ-additivity, so we cannot use outer measure on all subsets. The Carathéodory condition identifies the 'good' sets: those that split every test set A into two pieces whose outer measures sum correctly to μ*(A). The theorem proves that these sets form a σ-algebra (so we have a domain closed under the operations measures need) and that outer measure restricted to this σ-algebra is σ-additive (so we have a genuine measure). The condition transforms an over-defined, non-additive function into a restricted, properly additive measure.
The key insight is that the fix is selective exclusion rather than definition repair: instead of fixing outer measure to be additive everywhere, we restrict it to the domain where it is already additive. The Carathéodory condition characterizes exactly that domain. For Lebesgue measure, this domain turns out to contain all Borel sets and more — it is the Lebesgue σ-algebra, which is strictly larger than the Borel σ-algebra. The construction is the template for building any measure from an outer measure.