If $K$ is a field, $A=K[X]$, take $m,n \in K$ such that $m \ne n$. Prove that the ideals $I=(X-m)$ and $J=(X-n)$ are coprime.

I know the regular definition of coprime. But here, should we prove $I + J = A$ or $K$? And what are the units in $K[X]$?

New contributor
Kevin is a new contributor to this site. Take care in asking for clarification, commenting, and answering. Check out our Code of Conduct.
  • 2
    $\begingroup$ You want to prove $I+J=K[X],$ and the unit of $K[X] \text{ is } 1.$ With these clarifications, let us know if you still need help. $\endgroup$ – Robert Shore Feb 10 at 4:36
  • $\begingroup$ still need help.... $\endgroup$ – Kevin Feb 10 at 5:03
  • $\begingroup$ Can you get any non-zero constant polynomial into $I+J$? If so, what consequences follow? $\endgroup$ – Robert Shore Feb 10 at 5:04
  • $\begingroup$ not sure what do you mean $\endgroup$ – Kevin Feb 10 at 5:32
  • 1
    $\begingroup$ @Kevin He means that $(X-m)+(n-X)=n-m\in I+J$. And an ideal which contain an invertible element equals the whole ring. $\endgroup$ – user26857 Feb 10 at 18:33

What is the unit in $K[X]$? A (multiplicative) unit in $A = K[X]$ is an element $1_A$ such that

$1_A a = a 1_A = a \tag 1$

for any

$a \in A = K[X]; \tag 2$

it is clear from the ordinary definition of multiplication in $K[X]$ that $1_K \in K$ fulfills this requirement; here of course I have invoked a slight abuse of notation insofar as I have tacitly identified $K$ with the set of $0$-degree polynomials or constants in $K[X]$; technically I suppose we should write $1_KX^0$ for the unit in $K[X]$ but as is conventional I will simply refer to $0$-degree elements in $K[X]$ as being elements of $K$; so the unit in $K[X]$ is $1_K$ or simply $1$.

We want to show that

$(X - m) + (X - n) = A = K[X], \tag 3$

since by definition, two ideals

$I, J \subset K[X] \tag 4$

are co-prime if and only if

$I + J = (1_K) = K[X]; \tag 5$

this is what we need to prove with

$I = (X - m), \; J = (X - n), \tag 6$


$m, n \in K, \; m \ne n. \tag 7$

Now, polynomials of degree $1$ such as $X - m$ and $X - n$ are clearly irreducible in $K[X]$, since if

$rX + s = a(x)b(x), \; a(x), b(x) \in K[X], \tag 8$


$\deg a(x) + \deg b(x) = \deg (rX + s) = 1; \tag 9$

this implies that at precisely one of $a(x)$, $b(x)$ is of degree $1$, and the other is of degree $0$; thus $rX + s$ cannot be factored into two polynomials of positive degree, so it is indeed irreducible. Thus we see that $X - m$, $X - n$ are irreducible over$K$; this in turn implies that

$d(X) = \gcd(X - m, X - n) = 1, \tag{10}$

for it is clear from the above that $\deg d(X) = 0$ or $\deg d(X) = 1$; in the latter case

$d(X) = pX + q \mid X - m$ $\Longrightarrow \exists b \in K, \; b(pX + q) = bpX + bq = X - m \Longrightarrow bp = 1, bq = -m; \tag{11}$

and likewise,

$d(X) = pX + q \mid X - n$ $\Longrightarrow \exists c \in K, \; c(pX + q) = cpX + cq = X - m \Longrightarrow cp = 1, cq = -n; \tag{12}$

now the equations

$bp = 1 = cp \tag{13}$


$p \ne 0, \; b = c; \tag{14}$

but then

$m = -bq = -cq = n, \tag{15}$

contradicting our assumption that $m \ne n$; therefore,

$p = 0 \tag{16}$

and any common divisor of $X - m$, $X - n$ must be an element $0 \ne q \in K$; since $d(X)$ is defined up to multiplication by a unit, we have established (10). Now since $K[X]$ is a principle ideal domain, there exist

$f(X), g(X) \in K[X] \tag{17}$


$f(X)(X - m) + g(X)(X - n) = d(X) = 1; \tag{18}$

thus for any $r(X) \in K[X]$

$r(X) = r(X)(1)$ $= r(X)f(X) (X - m) + r(X) g(X)(X - n) \in (X - m) + (X - n), \tag{19}$

and thus we see that

$K[X] \subset (X - m) + (X - n) \Longrightarrow K[X] = (X - m) + (X - n) \tag{20}$

proved as per request.

  • 1
    $\begingroup$ Isn't pretty clear that $I+J$ contains a non-zero element from $K$, that is, a unit, and therefore equals the whole ring? $\endgroup$ – user26857 Feb 10 at 18:32
  • $\begingroup$ @user26857: you mean as in $(X - m) - (X - n) = n - m \ne 0$? Nice catch, maybe you ought to write it up as an answer! $\endgroup$ – Robert Lewis Feb 10 at 18:39
  • 1
    $\begingroup$ Yes, this is what I meant. (I let it as a hint for the OP.) $\endgroup$ – user26857 Feb 10 at 18:40

Note that, if $I + J = k[X]$, then there exist $p,q$ polynomials with $p(X-n) + q(X-m) = s$ for some $s \in k$ (actually, we can do this for any $s$ in the field).

Reciprocally, if $s \in I+J$, for some $s$ in $k$, then $1 \in s^{-1}(I+J) = I+J$ and so $k[X] = I+J$. Thus our problem reduces to showing that there exist $p,q \in k[X]$ with $p(X-n) + q(X-m) = s$ for some nonzero $s$. But this is rather direct, as

$$ (X-n) - (X-m) = m-n \neq 0. $$


Your Answer

Kevin is a new contributor. Be nice, and check out our Code of Conduct.

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

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