0

Given a linear transformation $T : \mathbb{Q}^4 \rightarrow \mathbb{Q}^4$, there exists a nonzero proper subspace $V$ of $\mathbb{Q}^4$ such that $T(V ) \subseteq V$.

Is the statement true?

I think the answer is no. But why? I am not sure about it. Is there any theorem which contradicts the above statement?

I also know that if the above condition holds then there exists a $\lambda$ such that $Tv = \lambda v$ for each $v \in \mathbb{Q}^4$. Somehow this may contradicts.

Please help me.

  • When you say $\subset$, do you mean $\subseteq$ or $\subsetneq$? (This confusion is why people universally need to stop using $\subset$ when the sets are allowed to be equal. The subset notation clearly reflects $<$ and $\leq$, and using it any other way is just not right.) – Arthur Nov 05 '21 at 08:21
  • @Arthur Yes, that is $\subseteq$. Sorry for the mistake. –  Nov 05 '21 at 08:23

1 Answers1

2

The answer is indeed negative. Take$$T(a,b,c,d)=(-d,a,b,c).$$Then the minimal polynomial of $T$ is $x^4+1$. If there was such a space $V$, the minimal polynomial of $T|_V$ would divide $x^4+1$. But it would also belong to $\Bbb Q[x]$, and $x^4+1$ is irreducible in $\Bbb Q[x]$.