2
$\begingroup$

I'm working through this problem in Axler's Linear Algebra Done Right (4th edition).

It says:

Suppose that V is finite-dimensional and $k \in \{1,...,\dim(V)-1\}$. Suppose $T \in L(V) $ is such that every subspace of $V$ of dimension $k$ is invariant under $T$. Prove that $T$ is a scalar multiple of the identity operator.

The first question I have is whether the $k$ is fixed or if it is for all $k$ in the set. It seems like the $k$ would be fixed based on the wording.

Aside from that, I'm pretty lost on how to even start and would appreciate a hint.

$\endgroup$

2 Answers 2

3
$\begingroup$

Assume $\dim(V) = n$, and if possible, let there exist a $v\in V$ such that $Tv$ is linearly independent of $v$. Then $\{v, Tv\}$ is an independent set, then we can extend this basis to a basis of $V$. Let $S =\{v,Tv,v_1, \cdots, v_{n-2}\}$ is a basis of $V$.

Now let $W = span\{v,v_1, \cdots, v_{k-1}\}$, which is a subspace of dimension $k$ for each $k \in \{1,2, \cdots, n-1\}$, which contains $v$ but not $Tv$. Which is a contradiction, as each subspace of dimension $k$ has to be $T$ invariant. So $v$ is dependent on $Tv \implies Tv = \lambda_v v$. Now, as we take this $v \in V$ is arbitrary, so this is true for all $v \in \mathbb{F}$.

Now let $v,w$ be two independent vectors, then $T(v+w) = \lambda_{v+w}(v+w) = \lambda_vv + \lambda_ww \iff \lambda_v = \lambda_{v+w} = \lambda_{w}$. So there must be a unique scalar for all vectors in $V$.

Which finally shows $T = \lambda I$.

$\endgroup$
1
$\begingroup$

Based on the wording, indeed $k$ is fixed.

Hint 1: Show that if it holds for some $k\geq 1$, then it holds for $k=1$.

If $U_1,U_2\subset V$ are $k$-dimensional subspaces with $T(U_i)\subseteq U_i$ then what can you say about $T(U_1\cap U_2)$?

Hint 2: Show that all $1$-dimensional subspaces are scaled by the same factor.

If $u_1,u_2\in V$ and $T(u_i)=\lambda_i u_i$, then what can you say about $T(u_1+u_2)$?

$\endgroup$
4
  • $\begingroup$ that the action of T on the sum of is also invariant. I'm tried making an argument like this by considering the action of T on the basis vectors of the subspaces but I'm getting nowhere. $\endgroup$ Commented Oct 18 at 17:19
  • $\begingroup$ It is really a matter of writing out the definitions, once you know what you want to prove. $\endgroup$ Commented Oct 18 at 17:20
  • $\begingroup$ ok so the union of the subspaces gives you the entire space V, so you can write any vector in V as a sum of vector from each of the subspaces and consider the action of T on any arbitrary vector v in V? $\endgroup$ Commented Oct 18 at 17:26
  • $\begingroup$ @user8083 I've expanded the hint(s) a bit. $\endgroup$ Commented Oct 18 at 17:34

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.