Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

(I'm assuming full axiom of choice in this post)

Theorem: Let $V$ be a vector space and $\beta$ be a basis for $V$ and $S$ be a linearly independent subset of $V$. Then, there exists $H\subset \beta$ such that $H\cup S$ is linearly independent.

Is there a direct way to prove that "If a vector space has bases $\beta_1$&$\beta_2$, then $|\beta_1|=|\beta_2|$" as a corollary of the given theorem above?

I'm asking this because if $V$ is finite-dimensional, then this corollary can be directly proved from the "replacement theorem"(Dimension theorem for finite-dimensional vector space).

share|cite|improve this question

1) if $\,S\,$ is a maximal lin. ind. set (i.e., a basis) then $\,H=\emptyset\,$ makes the trick (in fact, it makes it in any case, so I'd assume we don't want trivial cases), otherwise

2) It can't be that $\,S\cup\{b\}\,$ is lin. dependent for all $\,b\in\beta\,$ , since we know that $\,S\cup\{b\}\,$ lin. dependent iff $\,b\in\operatorname{Span}{S}\,$ , but then we'd get that $\,\beta\subset\operatorname{Span }S\,$ and thus $\,S\,$ would be a generating set and thus maximal lin. ind.-- against the assumption--, so there must be some element $\,b\in\beta\;\;s.t.\;\;S\cup\{b\}\,$ lin. ind.

And yes: from this we get that any two basis of a vector space have the same cardinality, otherwise we'd get, using the above argument, a contradiction to the fact (either theorem or equivalent definition) that a basis is a maximal linear independent set, which in turn is equivalent to "a basis is a minimal generating set...)

Added on request: WLOG, assume $\,w_1\ge w_2\,$ and infinite cardinalities (as finite ones is rapidly dealed with, generally in basic linear algebra courses), and look at the basis $\,\beta_1\,$ and at the lin. ind. set $\,\beta_2\,$ .

According to the proven lemma, there exists $\,H\subset \beta_1\,\,\,s.t.\,\,\,\beta_2\cup H\,$ is lin. ind. Since $\,\beta_2\,$ is a maximal lin. ind. set this implies $\,H=\emptyset\,$ , but by the theorem mentioned in (2) in my answer (after "since we know") , this means


and we're done since $\,w_i=\left|\operatorname{Span}{\beta_i}\right|\,$ for infinite cardinalities , and thus $\,w_1\le w_2\,$

share|cite|improve this answer
I don’t think that Katlus is worried about proving the theorem at the top of the question; the question is whether that result can be used to give a direct proof that any two bases of $V$ have the same cardinality. Your third paragraph addresses the real question, but I think that you need to add a lot of detail to make it convincing. How exactly are you going to use the quoted theorem to show, for instance, that a vector space $V$ over $\Bbb R$ can’t have disjoint bases $\beta_1$ and $\beta_2$ of cardinalities $\omega_1$ and $\omega_2$, respectively? – Brian M. Scott Feb 20 '13 at 8:02
I'm adding something to my answer which, I believe, could have been done by the OP. – DonAntonio Feb 20 '13 at 14:20
Don't you mean $\beta \subset \operatorname{span}S$ rather than $\beta \subset S$? – dfeuer Jan 5 '14 at 6:44
Yup, thanks @dfeuer – DonAntonio Jan 5 '14 at 15:35

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.