9

I was just wondering whether anyone knows an answer to the following:

Suppose that ${\mathbb F}$ is a subfield of a field ${\mathbb G}$ and that $v_1,\ldots ,v_k$ are linearly independent vectors in ${\mathbb F}^n$ (over $\mathbb F$). Is it necessarily true that $v_1,\ldots ,v_k$ are also linearly independent when considered as vectors in ${\mathbb G}^n$ (over $\mathbb G$})?

Any help would be much appreciated. Thanks!

Sean
  • 93
  • Related: https://math.stackexchange.com/questions/611630/how-can-two-vectors-be-dependent-in-one-field-and-independent-in-another-field – A.P. Apr 30 '20 at 07:10

1 Answers1

6

Yes. Extend the set $\{v_1,\ldots,v_k\}$ to a basis of $\mathbb{F}^n$ and put the basis vectors together to form a square matrix $A$. Then $A$ is invertible over $\mathbb{F}$. That is, $A^{-1}\in M_n(\mathbb{F})\subseteq M_n(\mathbb{G})$. So, $A$ is invertible over $\mathbb{G}$ and its first $k$ columns are linearly independent over $\mathbb{G}$.

However, do not confuse your question with the following one:

Suppose $\mathbb{F}$ is a subfield of $\mathbb{G}$ and the set $V$ is a vector space over each of $\mathbb{F}$ and $\mathbb{G}$. If $v_1,v_2,\ldots,v_k\in V$ are linearly independent over $\mathbb{F}$, are they necessarily linearly independent over $\mathbb{G}$?

The answer to this seemingly similar question is negative, as illustrated by the following counterexample: $V=\mathbb{C},\,\mathbb{F}=\mathbb{R},\,\mathbb{G}=\mathbb{C}$ and $\{v_1,v_2\}=\{1,\,i\}$. It is easy to see that $v_1$ and $v_2$ are linearly independent over $\mathbb{R}$ but linearly dependent over $\mathbb{C}$.

user1551
  • 139,064
  • Thanks very much. That's great. – Sean Oct 31 '13 at 11:52
  • Well, that means if V is a finite-dimensional vector space over some field $k$, and a subset of $V$ is linearly independent over $k$, then it is also linearly independent over any extension $L$ of $k$[provided V is a V.Sp. over L]. Am I correct? – Saikat Oct 06 '20 at 03:13
  • @Saikat No. This has little to do with the dimensions of the vector spaces. I have updated the counterexample in my answer to refute your claim. – user1551 Oct 10 '20 at 12:45
  • @user1551 can u explain why it will not hold for arbitary vector space over F and G.while F^n and G^n are also a vector space! What special property they have? (Sorry for symbols) – MEET PATEL Jul 24 '22 at 03:38