Questions: Basis for a Topology

Question 1

In ℝ with the standard topology, which of the following correctly describes the set (0,1) ∪ (3,5)?

• A. It is open and is itself a basis element (an open interval)
• B. It is not open because it is not a single connected interval
• C. It is open (as a union of two basis elements) but is not itself a basis element
• D. It is open only if (0,1) and (3,5) jointly satisfy the second basis condition

Answer: (C)

The standard basis for ℝ consists of open intervals (a,b). The topology is all unions of such intervals. The set (0,1) ∪ (3,5) is open — it is a union of two basis elements — but it is not an open interval itself, so it is not a basis element. This illustrates the key point: basis elements generate the topology through unions, but most open sets are not basis elements.

Question 2

The second condition for a basis — that for any point x ∈ B₁ ∩ B₂ there exists B₃ ∈ ℬ with x ∈ B₃ ⊆ B₁ ∩ B₂ — is required because:

• A. It ensures that basis elements are pairwise disjoint, preventing overlap
• B. It guarantees the generated topology is closed under finite intersections, as all topologies must be
• C. It prevents the basis from generating the indiscrete topology {∅, X}
• D. It ensures every basis element is an open set in some existing topology on X

Answer: (B)

A topology must be closed under finite intersections. When we define the topology as all unions of basis elements, we need to verify that the intersection of two such open sets is also a union of basis elements. The second condition is precisely what ensures this: for any point x in B₁ ∩ B₂, a basis element B₃ containing x and contained in the intersection exists, so the intersection can be expressed as a union of such B₃'s.

Question 3

A basis for a topology is analogous to a basis for a vector space: both are smaller generating sets from which the full structure (topology or vector space) is recovered by a combination operation — union in topology, linear combination in linear algebra.

• T. True
• F. False

Answer: True

This analogy is precise and useful. Just as every vector is a linear combination of basis vectors, every open set is a union of basis elements. Neither the topology nor the vector space needs to be specified element-by-element once a basis is given — the basis determines it completely. The analogy helps explain why bases are so useful: working with a few well-chosen generating sets is far more tractable than the full structure.

Question 4

In a metric space, nearly every open set in the metric topology is an open ball B(x, r) for some center x and radius r.

• T. True
• F. False

Answer: False

Open balls form a BASIS for the metric topology, meaning every open set is a union of open balls — but most open sets are not themselves open balls. For example, in ℝ with the Euclidean metric, (0,1) ∪ (2,3) is open (it is a union of open intervals, which are open balls in ℝ) but is not itself a single open interval. Confusing 'basis element' with 'open set' is one of the most common errors when first learning topology.

Question 5

State the two conditions a collection ℬ of subsets of X must satisfy to be a basis for a topology on X, and explain why the second condition is necessary.

Model answer: (1) Coverage: every point x ∈ X belongs to at least one basis element B ∈ ℬ. (2) Intersection condition: for any B₁, B₂ ∈ ℬ and any point x ∈ B₁ ∩ B₂, there exists B₃ ∈ ℬ with x ∈ B₃ ⊆ B₁ ∩ B₂. The second condition is necessary to ensure the generated topology — defined as all unions of elements of ℬ — is actually a topology, i.e., closed under finite intersections. Without it, the intersection of two 'open sets' (unions of basis elements) might not be expressible as a union of basis elements, violating the topology axioms.

The coverage condition ensures every point is in some open set (so X itself is open). The intersection condition is the non-trivial part: it is exactly what makes the union-of-basis-elements construction self-consistent as a topology.