I remember being told that this was true by a professor, but I haven't been able to find a source for it yet.

In the theorem as stated, $\mathbb{F}$ is any field and $T_n(\mathbb{F})$ denotes the algebra of upper triangular $n\times n$ matrices over $\mathbb{F}$.

Theorem: Let $A,B\in T_n(\mathbb{F})$ be such that for all $X\in T_n(\mathbb{F})$, $$AX=XA\implies BX=XB$$ Then $B=p(A)$ for some $p\in \mathbb{F}[t]$.

If we replace $T_n(\mathbb{F})$ by $M_n(\mathbb{F})$, the question is answered in this paper. Unfortunately, the argument doesn't seem to translate directly, as I can't find a way to force the $M_i$ maps to be upper-triangular.

Update: I have re-asked the question here on MO. Thanks to David E Speyer, we now know that this theorem is false. In particular, if $$A=\left[\begin{array}{cccc}0&0&0&1\\ &0&1&0\\ & &0&0\\ & & & 0\end{array}\right]$$ then the matrices commuting with $A$ are those of the form $$\left[\begin{array}{cccc}a&0&*&*\\ &b&*&*\\ & &b&0\\ & & & a\end{array}\right]$$ The matrix $$B=\left[\begin{array}{cccc}0&0&0&1\\ &0&0&0\\ & &0&0\\ & & & 0\end{array}\right]$$ commutes with all matrices of this form, but is clearly not a polynomial in $A$.

  • 1
    It is not considered good form to cross post, mathoverflow.net/q/304568/31729 – copper.hat Jul 10 at 22:49
  • And what's your question then? Sounds like you have answered your own question...... – Mostafa Ayaz Jul 14 at 18:19
  • @copper.hat I am sorry. I did not know that cross-posting was poor form. I will delete this question tomorrow when the bounty ends. (Apparently I can't delete questions with bounties on them.) Thank you for correcting me. – Spot Jul 14 at 19:51
  • @Spot: Just informing of the site etiquette, not a criticism. I have broken most rules inadvertently (and some knowingly!). – copper.hat Jul 14 at 19:54
  • but .. what is the exact meaning of ".. for all X" then ? because the counterexamples you are citing do not involve all the UT matrices , or did I loose something in understanding the statement? – G Cab Jul 15 at 21:38

Your Answer


By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.