0
$\begingroup$

I'm currently reading about adjoint functor (page 22 of these notes: http://www.math.jussieu.fr/~schapira/lectnotes/AlTo.pdf). In particular, it says that two functors $F: C\rightarrow C'$ and $G: C' \rightarrow C$ are adjoint if there exists an isomorphism of bifunctors: $Hom_{C'}(F(\circ),\circ) \cong Hom_C(\circ,G(\circ))$

By plugging in $F(\circ)$, we have: $Hom_{C'}(F(\circ),F(\circ)) \cong Hom_C(\circ,G(F(\circ)))$ as two functors from $C \rightarrow Set$. It then says that we have morphisms $X \rightarrow GF(X)$, functorial in X. I don't understand why it's functorial. So for each $X \in C$, we have $id_{F(X)} \in Hom_{C'}(F(X),F(X))$, which gives us a morphism $\varphi_X: X \rightarrow GF(X).$ If we have $f: X\rightarrow X'$, why the diagram commutes, in particular, why $GF(f) \circ \varphi_X = \varphi_{X'}\circ f ?$ Can anyone help me please.

$\endgroup$
  • $\begingroup$ That's part of the definition of isomorphism of functors. $\endgroup$ – Zhen Lin Oct 20 '13 at 16:24
2
$\begingroup$

Let's name $\psi \colon \text{Hom}_{C'}(F(\bullet),F(\bullet))\cong\text{Hom}_C(\bullet,G\circ F(\bullet))$ functorial isomorphism. In you notation so $\varphi_X = \psi(1_{F(X)})$.

Now the following equation holds $$\varphi_{X'} \circ f = \hom(f,G \circ F(X'))(\varphi_{X'})=\hom(f,G \circ F(X'))\circ \psi(1_{F(X)})$$ and this is also equal to $$\psi \circ \hom(F(f),F(X'))(1_{F(X')})=\psi(F(f))$$ in similar way we have that $$G\circ F(f) \circ \varphi_X = \hom(X,G\circ F(f))\circ \psi(1_{F(X)})$$ which is also equal to $$\psi \circ \hom(F(X),F(f))(1_{F(X)})=\psi(F(f))$$

From this it follows the equation you were looking for.

| cite | improve this answer | |
$\endgroup$

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.