1

Assume $\beta$ can be expressed in terms of polynomial relation in $\mathbb{Z}[\alpha]$, where $\alpha$ is an algebraic integer (i.e. $\alpha$ is the root of a polynomial in $\mathbb{Z}[X]$).

How can I argue that $\beta$ is itself an algebraic integer?

Attempt:

  • assume: $r_0 + r_1 \beta+ r_2 \beta^2 + \ldots + r_t \beta^t=0$ with $r_i \in \mathbb{Z}[\alpha]$
  • assume wlog, that $r_1$ contains some exponent $\alpha^l$
  • how can I substitute that exponent, since I only have the condition $q_0 + q_1 \alpha + q_2 \alpha^2 + \ldots + q_r \alpha^r=0$.
J. W. Tanner
  • 64,851
  • 5
  • 45
  • 89
  • 1
    Note that you need $r_t = 1$ for $\beta$ to be an algebraic integer (and you have $q_r = 1$ in the condition on $\alpha$). – Magdiragdag Mar 15 '14 at 18:23

2 Answers2

9

Worth emphasis: we can give a simple more concrete (high-school level) proof of the key Theorem in the accepted answer as follows. Recall we can represent algebraic integers as eigenvalues of square-integer matrices. Such matrices have integral characteristic polynomials $\in\Bbb Z[x]$ that are monic (lead coef $ = 1)$ so their eigenvalues are algebraic integers. Conversely, $\alpha\,$ is an eigenvalue of its companion matrix, i.e. the matrix of the linear map $\,x\mapsto \alpha \:\!x\,$ on $\,\Bbb Z[1,\alpha,\ldots\,\alpha^{n-1}].\,$

Now in $\,\Bbb Z\:\![1,\alpha,\ldots,\alpha^{n-1}]\!\times\! [1,\beta,\ldots,\beta^{k-1}]=\Bbb Z[\alpha^i\beta^j],\,$ $\, i<n,\,j<k,\,$ with basis $\,{\bf v} = [\alpha^i\beta^j],\,$ we can construct $\,nk \times nk\,$ matrices $\,L,M\,$ such that $\, \alpha\:\! {\bf v} = L\:\!{\bf v},\ \beta\:\! {\bf v} = M\:\!{\bf v},\,$ since in products $\,\alpha^i\beta^j\,$ that occur as entries in $\,\alpha\:\!{\bf v},\:\! \beta\:\!{\bf v}\,$ we can use the (minimal) polynomials of $\,\alpha\,$ and $\,\beta\,$ to rewrite $\,\alpha^n\,$ and $\,\beta^k\,$ into polynomials in lower powers of $\,\alpha\,$ and $\,\beta.\,$ Thus we get

$$\begin{eqnarray} (L+M)\:\!{\bf v} &=& L{\bf v} + M{\bf v} = \alpha\:\!{\bf v} + \beta\:\!{\bf v} = (\alpha+\beta){\bf v}\\[.5em] (LM)\:\!{\bf v} &=& L(M{\bf v}) = L(\beta{\bf v}) = \beta(L{\bf v}) = \beta\alpha\:\! {\bf v}\\ \end{eqnarray}\qquad$$

Thus, being eigenvalues of square integer matrices, $\,\alpha+\beta\,$ and $\,\alpha\beta\,$ are algebraic integers.

Algebraic integers are closed under sums and products, so the sought result follows immediately, since polynomials $\in\Bbb Z[\alpha]\,$ are compositions of sums and products of algebraic integers.

Example $ $ For a concrete calculation using this method see Robjohn's $2020$ answer here.

Readers familiar with linear algebra may find it instructive to express the above in the language of Kronecker products, e.g. see this answer, or The Arithmetic of Algebraic Numbers: An Elementary Approach, College Math Jnl, $2004$ (non-paywalled here).

Bill Dubuque
  • 285,713
  • 4
    Nice. I've seen a few questions on algebraic integers like "why is the sum of algebraic integers an algebraic integer?"; when I answered I always referred to the generic proof. This is a much more concrete approach; maybe much more appropriate if you haven't seen modules in general yet. Nevertheless, I think this only shows that poly's in algebraic int's are algebraic int's. The questions asks about transitivity of integrality. (Presumably, the question is not totally consistent about that). Can this argument be extended to show that a $\beta$ integral over $Z[\alpha]$ is integral itself? – Magdiragdag Mar 15 '14 at 18:28
  • For a concrete example see thsi answer. $\ \ $ – Bill Dubuque Mar 14 '24 at 23:06
  • @Magdiragdag This post was intended as a supplement to the proof in your answer. – Bill Dubuque Mar 15 '24 at 06:50
  • @BillDubuque: (+1) Nice; I had not seen this answer before. It is nicely related to the answer that you cited. – robjohn Mar 15 '24 at 07:08
8

The standard argument for this requires a bit of machinery, namely the following theorem (the formulation below is from Atiyah-MacDonald, Proposition 5.1).

Theorem. Let $B$ be a (commutative) ring and $A$ a subring and $x \in B$. Then the following four statements are equivalent.

  1. $x$ is integral over $B$;
  2. $A[x]$ is a finitely generated $A$-module;
  3. $A[x]$ is contained in a subring $C$ of $B$ such that $C$ is a finitely generated $A$-module;
  4. There exists a faithful $A[x]$-module $M$ which is finitely generated as an $A$-module.

Then, as a corollary (Corollary 5.4. in Atiyah-MacDonald), you get the following.

Corollary. Let $A \subseteq B \subseteq C$ be (commutative) rings. Assume that $B$ is integral over $A$ and that $C$ is integral over $B$. Then $C$ is integral over $A$.

Proof. Take $x \in C$. Because $C$ is integral over $B$, $$x^n + b_{n-1} x^{n-1} + \dots + b_0 = 0$$ for certain $b_0, \dots, b_{n-1} \in B$. The ring $B' = A[b_1,\dots,b_n]$ is a finitely generated $A$-module (repeatedly the theorem above) and $B'[x]$ is a finitely generated $B'$-module (also by the theorem above). Hence $B'[x]$ is a finitely generated $A$-module and therefore (again by the theorem above) $x$ is integral over $A$.

J. W. Tanner
  • 64,851
  • 5
  • 45
  • 89
Magdiragdag
  • 15,545
  • @Readers I added an answer giving a simpler more concrete view accessible at high-school level, and linked to an example. These may provide motivation for the more abstract approach above. – Bill Dubuque Aug 04 '25 at 23:42
  • 2
    @BillDubuque Don't you think your answers stand on their own merits? Attempts to steer attention away from good answers by others are bad form reeking of arrogance. If you think you have an answer, just post it, and let the voters judge for themselves. Note, I'm not saying that the points raised in your answer would not help the readers. – Jyrki Lahtonen Aug 05 '25 at 02:34