I'm reading Keith Conrad's note on the discriminant of a number field. So far I've made it about 1/3rd of a page :)

In example 1.2, he considers $K = \mathbb Q (\alpha)$ for $\alpha$ a root of $T^3 - 9T - 6$. He determines $N(\alpha) = N(\alpha+3) = N(\alpha-3) =6$ and then and then says "It follows that $(\alpha) = p_2 p_3$, $(\alpha - 3) = p_2' p_3$ and $(\alpha + 3) = p_2' p_3$". (Here I think $p_2, p_2'$ are primes above $2$ and $p_3$ above $3$). I can see why the norms imply $(\alpha)$ is the product of a prime above $2$ and $3$, but I don't immediately see why $(\alpha -3)$ must have a different prime above $2$ but the same prime above $3$, for example.

Using Dedekind Kummer's theorem I can show there's just one prime above $3$, and two above $2$, so that would get me part of the way. But the entire point of this example seems to be computing the factorization of $(2)$ and $(3)$, so I'm guessing there's a simpler way to see this intermediate step...


$(\alpha) = \prod \mathfrak{p}_j^{e_j}$

$N(\alpha)=6\implies (\alpha) = p_2p_3$ (where $N(p_2)=2,N(p_3)=3$)

$N(\alpha-3)=6\implies (\alpha-3) = p_2'p_3'$

$N(p_3)=3 \implies 3\in p_3 \implies (3)= p_3 I_3$ where $N(I_3)= 3^2$

$$(\alpha,\alpha-3) = (3,\alpha)= (p_3I_3,p_3p_2) = p_3 (I_3,p_2)=p_3$$ thus $\alpha-3\in p_3,\alpha-3\not \in p_2$ and $(\alpha-3) = p_3 p_2'$ where $p_2'\ne p_2$

  • $\begingroup$ Ah perfect, thank you very much! $\endgroup$ – vacant Oct 19 '19 at 4:12

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.