Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $f \in K[X] $ be an irreducible, separable polynomial, and let $M/K$ be a splitting field for $f$. Let $\alpha \in M$ be a root of $f$. Then $D(f) = (-1)^{d(d-1)/2} N_{K/k} (f'(\alpha))$, where $D(f)$ is the discriminant of $f$.

I have no idea what $N_{K/k}$ means. It hasn't been used anywhere else in the notes, and I can't find it by searching online.

share|cite|improve this question
Some playing around has led me to believe $ N_{K/k}(f'(\alpha)) = \Pi \sigma_i (f'(\alpha))$, where the $\sigma_i$ are the ($d$) distinct $K$-embeddings of $K(\alpha)$ into the splitting field $M$ of $f$. Still don't see what it means though, or what $k$ is... – Jonathan Dec 12 '11 at 18:00
Are all your letters accurate? You have an $M$ but no $k$, and your formula has a $k$ but no $M$. But, yes, if $K$ is an algebraic extension of $k$, then $N_{K/k}$ is the "norm map of $K$ over $k$": $N_{K/k}(a)$ is the product of the images of $a$ under all $k$-embeddings of $K$ into an algebraic closure of. – Arturo Magidin Dec 12 '11 at 18:05
I suspect it's a typo and instead means $N_{L/K}$. Thanks Arturo, that makes sense (never heard of it though) – Jonathan Dec 12 '11 at 18:07
I think you mean $M/K$, not $L/K$. The norm is very standard. – Arturo Magidin Dec 12 '11 at 18:11
Yes, you're right. Now I know what to wiki, it all makes sense. Thanks – Jonathan Dec 12 '11 at 18:20
up vote 1 down vote accepted

From Lang's Algebra, revised 3rd Edition, Chapter VI, Section 5.

Let $E$ be a finite extension of $k$, and let $[E:k]_s=r$ (the separable degree; given an embedding $\sigma\colon k\to L$, where $L$ is algebraically closed, the number of distinct extensions of $\sigma$ to $E$), and let $p^{\mu}=[E:k]_i$ if the characteristic is $p\gt 0$ (the inseparable degree; in a finite extension, $[E:k]_s[E:k]_i = [E:k]$).

Let $\sigma_1,\ldots,\sigma_r$ be the distinct embeddings of $E$ into a fixed algebraic closure $\overline{k}$ of $k$. If $\alpha\in E$, then the norm of $\alpha$ from $E$ to $k$ is defined to be: $$N_{E/k}(\alpha) = N^{E}_k = \prod_{\nu=1}^{r}\sigma_{\nu}(\alpha^{p^{\mu}}) = \left(\prod_{\nu=1}^r \sigma_{\nu}(\alpha)\right)^{[E:k]_i}.$$ The trace of $\alpha$ from $E$ to $k$ is defined to be: $$\mathrm{Tr}_{E/k} = \mathrm{Tr}^E_k(\alpha) = [E:k]_i\sum_{\nu=1}^{r}\sigma_{\nu}(\alpha).$$

The norm is the subject of the famous "Hilbert's Theorem 90" (which has an additive version involving the trace).

One reason for the names is a different way to compute them (Exercise 17 in Chapter 2 of Daniel Marcus's Number Fields): Let $E/k$ be a finite extension, and fix $\alpha\in E$. Multiplication by $\alpha$ gives a linear mapping of $E$ to itself, considering $E$ as a vector space over $k$. Pick a basis for $E$ over $k$ (as a vector space), and let $A$ be the matrix of this linear mapping. Then $$\mathrm{Tr}_{E/k}(\alpha) = \mathrm{trace}(A)\qquad\text{and}\qquad {N}_{E/k}(\alpha) = \det(A).$$

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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