MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let D be a division ring with center Z. Let R and K be two maximal subfields of D, both purely inseparable of exponent one ( means the p power of each of them in Z). Why are R and K isomorphic?

Or a better question, why there is only one maximal unique subfield which is purely inseparable of exponent one?

share|cite|improve this question

I think this question presupposes something that is incorrect. Here is an example. Fix a field $F$ of characteristic 2 and put $Z$ for the field $F(a,b)$ where $a$ and $b$ are indeterminates. Define $D$ to be the quaternion division algebra with center $Z$ generated by elements $i, j$ satisfying $$ i^2 = a,\quad j^2 + j = b,\quad ij = (j+1)i. $$ Let us denote it by $\lt a,b]$, as PK Draxl does in his book "Skew Fields". This is a division algebra because $a$ is not a norm from the separable quadratic extension $E$ obtained from $Z$ by adjoining a root of $x^2 + x + b$.

Obviously $D$ contains the purely inseparable quadratic extensions $Z(\sqrt{a})$. I claim it also contains the extension $Z(\sqrt{ab})$. To see this, we calculate in the Brauer group: $$ \lt a,b] = \lt ab^2,b] = \lt ab,b] + \lt b,b] = \lt ab, b] $$ where the last equality is because $\lt b,b]$ is split, i.e. isomorphic to 2-by-2 matrices, which follows from the fact that $b$ itself is a norm from $E$ (in fact, the norm of the element $x$).

Because $b$ is a nonsquare in $K$, we have found two non-isomorphic purely inseparable quadratic extensions in $D$ of exponent 1.

share|cite|improve this answer
Why is the field $Z(\sqrt{a})$ not isomorphic to $Z(\sqrt{ab})$? In fact, I believe those fields are isomorphic as $F$-algebras (though not as $Z$-algebras), since $Z = F(a,b) = F(ab,b)$. – George McNinch Aug 12 '10 at 14:56
I meant "not isomorphic as $Z$-algebras" (which is obvious because $b$ is not a square in $Z$); I agree they are isomorphic as $F$-algebras. I think $Z$-algebra isomorphism is what was intended in the original question. – Skip Aug 12 '10 at 18:16

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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