Questions: The Axiom of Choice and Its Equivalences

Question 1

A mathematician wants to prove that every vector space has a basis. Which statement correctly describes the role of the axiom of choice in this proof?

• A. AC is not needed; the proof constructs a basis by listing all vectors in order
• B. AC is needed via Zorn's lemma: the poset of linearly independent sets has chains with upper bounds, so a maximal element (a basis) exists
• C. AC is needed only for infinite-dimensional spaces over uncountable fields
• D. AC is only needed if the vector space is well-ordered, which requires a separate assumption

Answer: (B)

The standard proof uses Zorn's lemma: consider the poset of all linearly independent subsets of the vector space, ordered by inclusion. Every chain (totally ordered subcollection) has an upper bound — its union, which is also linearly independent. Zorn's lemma (equivalent to AC) then guarantees a maximal linearly independent set exists, and maximality implies it spans the space, so it's a basis. Without AC, it is consistent with ZF that vector spaces with no basis exist. Option C is wrong because even countable-dimensional spaces over ℚ use this argument.

Question 2

Which of the following is NOT equivalent to the axiom of choice over ZF set theory?

• A. The well-ordering theorem: every set can be well-ordered
• B. Zorn's lemma: every chain-complete poset has a maximal element
• C. The axiom of regularity: every nonempty set has an element disjoint from it
• D. The multiplicative principle: any Cartesian product of nonempty sets is nonempty

Answer: (C)

The axiom of regularity (also called the axiom of foundation) is an independent ZF axiom that prevents sets from containing themselves — it is provable from ZF without any choice principle and is not equivalent to AC. By contrast, the well-ordering theorem, Zorn's lemma, and the multiplicative principle are all provably equivalent to AC over ZF: each can be derived from the others. The surprising content of AC's equivalences is precisely that logically very different-sounding statements turn out to say exactly the same thing.

Question 3

The axiom of choice is provably true from the other Zermelo-Fraenkel axioms.

• T. True
• F. False

Answer: False

This is a fundamental result in set theory. Gödel showed in 1938 that AC is consistent with ZF — you cannot derive a contradiction by adding it. Cohen showed in 1963 (via forcing) that ¬AC is also consistent with ZF — you cannot disprove AC from ZF either. Together these results establish that AC is independent of ZF: it can neither be proved nor refuted from the other axioms. This means mathematicians genuinely choose whether to work in ZFC (with choice) or explore alternatives like the axiom of determinacy (AD), which contradicts AC.

Question 4

Zorn's lemma guarantees that a maximal element exists in any partially ordered set.

• T. True
• F. False

Answer: False

Zorn's lemma has a crucial hypothesis: the poset must have the property that every chain (totally ordered subset) has an upper bound within the poset. Without this condition, maximal elements need not exist — for example, the integers ordered by ≤ have no maximal element and no bounded chains. Zorn's lemma says: IF every chain has an upper bound THEN a maximal element exists. The hypothesis is not automatic and must be verified in each application. This is precisely why applying Zorn's lemma to a given poset requires real work: you must prove the chain-boundedness condition, not just assert it.

Question 5

Why is the axiom of choice considered nonconstructive, and why does this matter mathematically?

Model answer: AC asserts that a choice function exists for any collection of nonempty sets, but it provides no recipe for constructing it. It says 'there exists a way to pick one element from each set' without specifying which element or how to find it. For finite collections, explicit choices can always be made; the nonconstructive character appears only for infinite collections of arbitrary sets. This matters because AC proves existence without exhibiting an example. Results that rely on AC — the existence of a vector space basis, a well-ordering of the reals, a non-measurable set — are existence theorems with no constructive witness. Constructive mathematicians who require proofs to provide explicit objects reject AC for exactly this reason.

The independence of AC from ZF means there are mathematical worlds (models of ZF+¬AC) where vector spaces may have no basis, where the reals cannot be well-ordered, and where all sets of reals are Lebesgue measurable. AC's nonconstructive character is what enables these strange but consistent alternative universes.