2
$\begingroup$

In theorem A3.5 of Ash's book Abstract Algebra: The Basic Graduate Year (page 20 in this pdf), the author set out to prove the following.

Let $\sigma: F \rightarrow L$ be a field monomorphism where $L$ is algebraically closed. If $E/F$ is algebraic, then $\sigma$ extends to a monomorphism $\tau: E \rightarrow L$.

Here is the full statement and proof:

enter image description here

The proof is essentially an application of Zorn's lemma. However, the author never explicitly explained why it is necessary for $L$ to be algebraically closed.

Here is a counter-example with the condition removed: Stack $E/L/F$ into a chain of extensions where $E$ is algebraically closed and a proper extension of $L$. As a result, we cannot have any monomorphism from $E$ into $L$.

Now, how do I weave this observation into the proof?

$\endgroup$
  • 3
    $\begingroup$ Well, I would look into (A3.4). $\endgroup$ – Kevin Carlson Aug 2 '15 at 23:08
1
$\begingroup$

Take a look at A3.4:

enter image description here

The fact that $L$ is algebraically closed implies that the number of distinct roots of the polynomial $f\in F[x]$ in an algebraic closure of $F$ is equal to the number of distinct roots of $\sigma f\in L[x]$ in $L$.

$\endgroup$
  • $\begingroup$ I did not entirely understood A3.4 and skipped it. On second look, it seems that we want $L$ to be algebraically closed so that the image of the extension is in $L$. If $\beta_1, \beta_2 \ldots \beta_m$ are the distinct roots of $\sigma f$, then $F(\alpha) \cong (\sigma F)(\beta_i)$ where $i \in \{ \, 1, 2 \ldots m \, \}$. This gives us the $m$ extensions associated with $\alpha \in E$. If $L$ is not algebraically closed, then some of the extensions may not land on $L$. I still need to stew over the details, but the big picture looks right, TY .... $\endgroup$ – Andy Tam Aug 3 '15 at 14:37

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.