3
$\begingroup$

I'm studyng basic ring theory, in a "master's degree" in math. We've studied this objects in class: $\mathbb{Z}[\sqrt[3]{2}]$ and another one similar: $\mathbb{Z}[\sqrt[3]{2}, \sqrt{6}]$.

The definition this object, K[S], is: Given a ring R, let S $\subseteq$ R. Let K be a subring of R. Then K[S] is the smallest ring that contains both R and S.

Note: My rings all have unity by definition.

The question is this: Have you heard of a similar or equal definition, and could you provide me links (e.g. wikipedia or wolfram articles, or mathSE questions). I want to know if this concept appears somewhere else or is something only known to my professor and my university. So please point me out the "name" of this K[S] (I presume is something like "Ring generated by a set"), so I can search in the internet and in books to know more about this particular object.

Note: I'm not concerned with the special cases I mentioned at the beggining, only with the name and definition of K[S].

$\endgroup$
  • 2
    $\begingroup$ This is exactly what you expected: "ring generated by a set" (technically subring for technical reasons, but for all intents and purposes, we usually consider $K[S]$ independent of the ring $R$) en.wikipedia.org/wiki/Subring#Subring_generated_by_a_set $\endgroup$ – William Stagner Sep 28 '15 at 20:00
  • $\begingroup$ @WilliamStagner I know what a subring generated by some set of the ring is, I'm not asking that, I'm asking for clarification on what K[S] is. The defiition I gave for K[S] is equivalent to $<K \cup S>$ where <D> is notation for the ring generated by D. But then why is there an assymetry in the notation? Is K[S]=S[K]? (obviously by my definition is, but the notation seems to imply they are in some relation more assymetric than just union. $\endgroup$ – Santropedro Sep 28 '15 at 20:15
  • $\begingroup$ Then I'm not quite sure what you're asking. The reason we write $K[S]$ instead of $S[K]$ is because we are intuitively adjoining $S$ to $K$ in order to form a larger ring containing $K$ as a subring. It doesn't make sense to write $S[K]$ because $S$ is just a set. $\endgroup$ – William Stagner Sep 28 '15 at 20:21
  • $\begingroup$ Yes. My definition prohibited me of writing S[K] but relaxing the definition by not asking the first parameter to be a ring, I can. The important thing is : does this object have a name? $\endgroup$ – Santropedro Sep 28 '15 at 20:41
2
$\begingroup$

One interpretation that may clear things up is that $K[S]$ is not only a ring, but an algebra, specifically, a $K$-algebra, specifically, the $K$-algebra generated by $S$. That clarifies the asymmetry: $\mathbb{Q}[\sqrt 3]$ is an algebra over the rational numbers, not over the square root of 3! But as a ring, it's perfectly fine to say $\mathbb{Q}[\sqrt 3]=\langle \mathbb{Q},\sqrt 3\rangle$. On the other hand, $\sqrt{3}[\mathbb{Q}]$ doesn't make much sense. At best, you could interpret it as $\mathbb{Z}[\sqrt 3][\mathbb Q]$, which correctly exhibits $\mathbb{Q}[\sqrt 3]$ as a $\mathbb{Z}[\sqrt 3]$-algebra.

It might be helpful to observe that every algebra is a quotient of a polynomial algebra, for instance, $\mathbb{Q}[\sqrt 3]$ is isomorphic to $\mathbb Q[X]/(X^2-3)$, so that the notation you're asking about is just an extension of the ordinary notation for polynomials.

$\endgroup$
  • $\begingroup$ Though this is only true if the top ring is commutative, or in general if $K \subseteq R$ makes $R$ a $K$-algebra. (For example, $\mathbb{C}[\mathbb{H}] = \mathbb{H}$ is not a $\mathbb{C}$-algebra.) $\endgroup$ – Ben Sep 28 '15 at 20:43
  • $\begingroup$ @BenDyer thanks. I never think about noncommutative rings. $\endgroup$ – Kevin Carlson Sep 28 '15 at 20:51
  • $\begingroup$ @BenDyer surely you meant "K a subset of the center of R" to be consistent with your first statement: definitions of algebras over noncommutative rings are few and far between. Your last sentence could be looked at as a counterexample to the second half of the first sentence. Regards. $\endgroup$ – rschwieb Sep 29 '15 at 2:14
  • $\begingroup$ @rschwieb Thanks, I was meaning that $(ra)\cdot(sb) = rs(a\cdot b)$ for $r,s \in K, a,b\in R$, and wasn't bothering to make any further deductions. As you say, it is equivalent to $K$ being central. But perhaps there is some value in the first version as well, since the two are no longer equivalent when we allow $R$ to be a non-unital ring, I think. $\endgroup$ – Ben Sep 30 '15 at 18:13
  • $\begingroup$ @BenDyer Yep. As one who has thought about noncommutative definitions of algebras in the past, I can appreciate that. $\endgroup$ – rschwieb Sep 30 '15 at 18:27

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.