Skip to content
← All questions

Why is rejecting just the candidates L=1L=1 and L=−1L=-1 not enough to prove divergence for an=(−1)na_n = (-1)^n?

Rejecting specific values like 1 or -1 only proves those particular numbers are not limits. To rigorously prove the limit does not exist, one must show that *every* possible real number LL fails the convergence definition.

Conditions

  • The sequence is defined as an=(−1)na_n = (-1)^n.
  • Proof requires ruling out all L∈RL \in \mathbb{R}.

Reasoning, step by step

  1. Assume a hypothetical limit LL exists.
  2. Apply the triangle inequality: 2=∣1−(−1)∣≤∣1−L∣+∣−1−L∣2 = |1 - (-1)| \le |1 - L| + |-1 - L|.
  3. Deduce that at least one of the distances ∣1−L∣|1-L| or ∣−1−L∣|-1-L| must be ≥1\ge 1.
  4. Choose ε=1\varepsilon = 1. The subsequence corresponding to the farther point will always have terms outside (L−ε,L+ε)(L-\varepsilon, L+\varepsilon).
  5. Conclude that no single LL can satisfy the condition for all sufficiently large nn.

Example

If we pick L=0L=0, both 1 and -1 are at distance 1. With ε=1\varepsilon=1, neither is strictly inside the band (−1,1)( -1, 1 ) if boundaries are exclusive, or they sit on the edge depending on strictness, but typically we use ε<1\varepsilon < 1 or rely on the fact that for any LL, one cluster is far away.

Common misconceptions

  • Believing that checking the visible oscillation points (1 and -1) covers all possibilities.
  • Confusing 'limit does not exist' with 'limit is infinite'.

Watch the explanation

Explore next

Answers are generated from source material and independently checked. Consult the original video or creator if something is unclear.