I couldn't recall the UPQ from my memory exactly, so I wrote the following:

Let $R, S$ be rings, $I$ be an ideal of $R$, $\pi:R\to R/I$ be the canonical quotient ring homomorphism and $\varphi:R\to S$ be a surjective ring homomorphism. Then there exists isomorphism $\overline{\varphi}:R/I\to S$, such that ${\varphi}=\overline{\varphi}\circ \pi$.

I'm looking at the original statement, but it is phrased differently - there is no word "surjective" and no word "isomorphism", and $I$ is said to be subset of $\ker\varphi$. But could there possibly be some truth to what I wrote?

  • 1
    $\begingroup$ I believe you mean $\varphi = \overline{\varphi} \circ \pi$, right? This equation alone tells you that $I \subseteq \ker(\varphi)$ since $I = \ker(\pi)$. You might also want to think about the case $S = R$ and $\varphi = \text{id}_R$. $\endgroup$ – Matthias Klupsch Apr 13 '17 at 10:08
  • $\begingroup$ @MatthiasKlupsch Yes, I made a typo here, will fix it now. $\endgroup$ – sequence Apr 13 '17 at 19:33

The general universal property of quotients states:

For a ring homomorphism $\varphi:R\to S$, and and ideal $I\subseteq\ker\varphi$, there exists a unique homomorphism $\overline{\varphi}:R/I\to S$ such that $\bar{\varphi}\circ\pi=\varphi$, where $\pi:R\to R/I$ is the usual projection.

In the case where $\varphi$ is surjective, $\overline{\varphi}$ must be surjective as well, since given any $s\in S$, there exists an element $r\in R$ such that $s = \varphi(r) = \overline{\varphi}(r+I)$.

The last bit should be a little more precise; you should have that $I=\ker\varphi$, not just containment $\subseteq$. When $I=\ker\varphi$, the first isomorphism theorem implies that the induced map $\overline{\varphi}:R/I\to S$ is an isomorphism. So, there is some truth to what you wrote, as long as you specify that $I=\ker\varphi$.

As an example, let $\varphi:\mathbb{Z}\to\mathbb{Z}/2\mathbb{Z}$ be the projection. Note that this is surjective. Then, we have $4\mathbb{Z}\subseteq\ker\varphi = 2\mathbb{Z}$, and thus there exists a unique map $$\overline{\varphi}:\mathbb{Z}/4\mathbb{Z}\to\mathbb{Z}/2\mathbb{Z}$$ such that $\overline{\varphi}\circ\pi=\varphi$. However, $\overline{\varphi}$ is clearly not an isomorphism, since one ring has $2$ elements while the other has $4$. So the condition $I=\ker\varphi$ is necessary to conclude that the induced map is an isomorphism.


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.