1

Say we are considering some vector space $V$ and some non-zero unit vector $x$. Considering the one-dimensional vector space $\langle x \rangle = \text{span}(x)$, we know it admits a complementary subspace $H$ (per Zorn's lemma). I was wondering what is a sufficient argument then to establish that $H$ is a hyperplane then, as we can't just look at the sums/differences of dimension.

Would it be enough to claim that that $x + H$ provides a basis for the quotient space $V/H$ (and so it's one dimensional)? It's clear we can readily demonstrate this by using the definition of the internal direct sum, but I just wanted to check that my reasoning is correct as I am unfamiliar with working in this setting. Thanks for the help!

kodiak
  • 520
  • Please [edit] the question to provide your definition of "hyperplane". – Ethan Bolker Feb 11 '24 at 15:15
  • @EthanBolker I suppose that is part of my question. I know in the finite-dimensional case, we consider a dimension $n - 1$ subspace $H$ of a dimension $n$ vector space $V$ to be a hyperplane. But how do we modify this in the infinite-dimensional case? Is it precisely that the quotient space is dimension 1? – kodiak Feb 11 '24 at 15:24
  • That's the only completely general definition I can imagine. For Banach spaces where you have a norm and can think about infinionite as well as finite sums you can say a little more. – Ethan Bolker Feb 11 '24 at 16:00

0 Answers0