0

I am looking at the first answer to this question: How to show $\mathbb Z_2[x]/(x^3+x+1)$ is isomorphic to $\mathbb Z_2[x]/(x^3+x^2+1)$

Why does the existence of an automorphism between $x$ and $x+1$ imply that the quotient rings are isomorphic? I understand that when you substitute $x$ with $x+1$ in the original expression you get $$(x+1)^3 + x+1 +1 = x^3 + 3 x^2 + 4 x + 3 = x^3 + x^2 +1$$ in $\mathbb{Z}_2[x]$, but I don't understand the connection to the quotient ring.

I am trying to apply the logic to a similar question, to show that $$\mathbb{Z}_2[x] / \langle x^2 \rangle \cong \mathbb{Z}_2[x]/\langle x^2-1\rangle$$ But I cannot convince myself why this automorphism leads to the conclusion that the quotient rings are isomorphic.

1 Answers1

2

For the first part:

The polynomial ring $\mathbb{Z}_2[x]$ has a universal property:

Given any commutative ring with unity $S$, any ring homomorphism $g\colon\mathbb{Z}_2\to S$, and any $s\in S$, there exists a unique homomorphism $f\colon \mathbb{Z}_2[x]\to S$ such that $f(x)=s$ and $f(a)=g(a)$ for all $a\in\mathbb{Z}_2$.

Here, we start with the ring $S=\mathbb{Z}_2[x]$, $g$ the embedding of $\mathbb{Z}_2\hookrightarrow S$, and take the element $x+1$. So we have a unique morphism $f\colon\mathbb{Z}_2[x]\to\mathbb{Z}_2[x]$ that sends $x$ to $x+1$.

Now compose $f$ with the quotient map $\mathbb{Z}_2[x]\to \mathbb{Z}_2[x]/\langle x^3+x^2+1\rangle$. Call this composite map $$g\colon \mathbb{Z}_2[x]\to \frac{\mathbb{Z}_2[x]}{\langle x^3+x^2+1\rangle}.$$ The map is surjective, because both $f$ and the projection are surjective. What is its kernel? The kernel is generated by the pre-image of $x^3+x^2+1$: but this is precisely $x^3+x+1$, because as you note $(x+1)^3 +(x+1) + 1 = x^3+x^2+1$. Thus, by the Isomorphism Theorems, $$\frac{\mathbb{Z}_2[x]}{\langle x^3+x^2+1\rangle} \cong \frac{\mathbb{Z}_2[x]}{\mathrm{ker}(g)} = \frac{\mathbb{Z}_[x]}{\langle x^3+x+1\rangle},$$ proving the desired isomorphism.


For the second part: we use the fact that $(x^2-1) = (x^2+1) = (x+1)^2$. Once again, take the map that sends $x\to x+1$ to get a morphism $f\colon\mathbb{Z}_2[x]\to \mathbb{Z}_2[x]/\langle (x+1)^2\rangle$. Prove the map is surjective, with kernel $\langle x^2\rangle$.

Note this is not going to work over a field that is not of characteristic two, since then $\langle x^2\rangle$ is the square of an ideal, while $\langle x^2-1\rangle = \langle x+1\rangle\langle x-1\rangle$ is the product of two coprime ideals. So the left side would have nilpotent elements, whereas by the Chinese Remainder Theorem the right hand side would be isomorphic to $F\times F$, which does not.

Arturo Magidin
  • 398,050
  • @ViktorVaughn I don't think the "not"s are the issue. The universal property should specify $S$ has characteristic two, and then the last paragraph starts with "Note this is not going to work over a field that is not of characteristic two," – Brian Moehring Mar 06 '21 at 20:40
  • @BrianMoehring "Note this is not going to work over a field of characteristic zero..." would be fine, too, but I agree that changing zero to two is better. I still think having two "not"s makes the sentence harder to understand. "Note that this only works over a field of characteristic two..." is clearer to me. – Viktor Vaughn Mar 06 '21 at 20:45
  • @ViktorVaughn Your last rephrasing doesn't quite work (it actually changes the meaning from my proposed correction), but I think your broader point is that double-negatives can be confusing, and on that I agree. – Brian Moehring Mar 06 '21 at 20:51
  • @Brian: You are right about the universal property; what is missing is a ring homomorphism from $\mathbb{Z}_2$ to $S$. – Arturo Magidin Mar 06 '21 at 21:38
  • @ViktorVaughn: A think-o of “zero” instead of “two”. Thank you. – Arturo Magidin Mar 06 '21 at 21:39
  • @ArturoMagidin Curious on your correction to the universal property... Is that how the property of $\text{char}(S)=2$ in the category of rings with unity generalizes to the category of rngs? (I have a lot less experience with rngs) – Brian Moehring Mar 06 '21 at 21:49
  • @BrianMoehring: Is how the universal property of the polynomial ring generalizes to the category of commutative rings: the polynomial ring $R[X]$ has the property that for every ring homomorphism $g\colon R\to S$ of commutative rings, and every set map $f\colon X\to S$, there exists a unique ring homomorphism $F\colon R[X]\to S$ such that $F|{R}=g and $F|{X}=f$. You can do it with commutative unital rings, or commutative not-necessarily-with-unity rings. – Arturo Magidin Mar 06 '21 at 21:51