I'm trying to prove the following, and I've made some progress on (i) and am having a bit of trouble with (ii)

A polynomial $p(x)$ over a field $k$ is $monic$ if the highest power of $x$ has coefficient $1$. Let $p(x)$ be monic and let $r\in k$.

(i) Show that if $p(x)$ is divided by $x-r$, then the remainder is $p(r)$.

(ii) Show that this remains true if $p(x)$ has coefficients in $\mathbb R$ and $r \in \mathbb C$.

For (i), when I set up polynomial long division, I start to see the pattern of $a_{n-1}+r$, then $r(a_{n-1}+r)+a_{n-2}$, then $r(r(a_{n-1}+r)+a_{n-2})$, etc as the coefficient (and I'm assuming eventually the remainder) from each additional subtraction of two terms at the bottom when we add the next term at the top. I think that this is the key element to the proof, as the sequences goes on to be p(r), but not sure how to word this rigorously into a proof.

For (ii) I'm really not sure how to go about this. It almost seems like the polynomial division algorithm works just as well if r is complex as if it isn't, but not sure how to prove it or if it's true.

  • $\begingroup$ (ii) is just a special case of (i), where $k = \Bbb C$ and $p$ has real coefficients. It is strange that you are requested to show that it "remains true" for a special case. Are you sure that the problem is not misstated? Also strange is that $p$ is assumed monic, since that plays no role. Nor does the hypothesis that the ring $k$ is a field, since the standard simple proof shows that the result is true over any ring $k$. Perhaps you know fields but not rings? $\endgroup$ – Math Gems Jan 18 '13 at 18:09
  • $\begingroup$ Yeah, I find that strange as well, but the problem isn't misstated. Regarding rings, this is from a linear algebra book, so I'm guessing they give us just enough abstract algebra to do what we need (ie no rings). $\endgroup$ – mlstudent Jan 19 '13 at 10:07

(i) In general, if you divide a polynomial $f(x)$ by a polynomial $g(x)$, you obtain a quotient $q(x)$ and a remainder $u(x)$ with $ \deg u<\deg g$ such that $$\tag1f(x)=q(x)g(x)+u(x).$$ Here, $u(x)$ must be some constant $c$ because $\deg g=1$. Now set $x=r$ in $(1)$ ...

(ii) Nothing special was used.

| cite | improve this answer | |
  • $\begingroup$ Your $(1)$ is not true in an arbitrary ring. For example consider $x^2 / 2x$ in $(\mathbb Z / 4 \mathbb Z)[x]$. It holds however when dividing by the monic polynomials, which is the case here. $\endgroup$ – Marek Jan 18 '13 at 16:06
  • $\begingroup$ @Marek The problem is about polynomials over a field. Fields are not arbitrary rings. Of course my "in general" was intended to mean "also for other divisors" not "also for other rings". $\endgroup$ – Hagen von Eitzen Jan 18 '13 at 16:13
  • $\begingroup$ I didn't notice that assumption, sorry. Nevertheless, I think it's worth pointing out that the assumption that $R$ be a field is not necessary here and that recalling general division algorithm is quite an overkill when dividing by linear monic. $\endgroup$ – Marek Jan 18 '13 at 16:27

Express $p(x) = q(x)(x-r) + a$. Try to prove that this holds in $R[x]$ for any ring $R$ whatsoever. This takes care of $(i)$.

Now, $(ii)$ is just a special case of the above when $R = \mathbb C$ and we restrict coefficients to $\mathbb R$.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.