Let me give some background to the question. For a category $\mathscr{C}$, define the tautological correspondence $$\tau : Ob\space Arr \space\mathscr{C} \longrightarrow Mor\space\mathscr{C}$$ to be the identity correspondance between objects in $Arr\space\mathscr{C}$ and morphisms of $\mathscr{C}$. It can be checked that $\tau$ is a natural transformation between the source and the target functors of $\mathscr{C}$.

The statement to prove is:

If $F$ and $F'$ are functors from $\mathscr{B}$ to $\mathscr{C}$ and $\phi$ is a natural transformation from $F$ to $F'$, then $\phi$ is pulled uniquely from the the tautological correspondance $\tau$ on $Arr\space\mathscr{C}$ defined above. In other words, there exists a unique functor $\Phi:\mathscr{B} \longrightarrow Arr\space\mathscr{C}$ such that $\phi=\tau\circ\Phi$.

I am able to show the existence by constructing such a $\Phi$: assign $\phi_b$ to each object b of $\mathscr{B}$. Assign the pair of morphisms ($F\alpha,F'\alpha$) to each morphism $\alpha:c\rightarrow c'$ of $\mathscr{B}$. Since $\phi$ is a natural transformation the following diagram commute : $\require{AMScd}$ \begin{CD} Fc @>{\phi_c}>> F'c\\ @V{F\alpha}VV @VV{F'\alpha}V\\ Fc' @>>{\phi_{c'}}> F'c' \end{CD}

and it is easy to check that $\Phi$ is a functor from $\mathscr{B}$ to $Arr\space \mathscr{C}$.

Now the part I'm stuck on is uniqueness. Perhaps, I can use the fact that $\tau$ is a natural transformation between the source and the target functors of $\mathscr{C}$. But I don't know how to carry this through. Any help would be appreciated.

As a beginner of Category Theory, I apologize for any sloppiness that would baffle those fluent in the language.


Uniqueness on objects is easy:
For any $b\in Ob\mathcal B$, we must have $\phi(b)=\tau(\Phi(b))$ but $\tau$ is the identical map of arrows of $\mathcal C$, so $\phi(b)=\Phi(b)$.

For morphisms, we indeed have to interpret $\tau$ as a natural transformation $\mathrm{dom}\to\mathrm{cod}$.
Then $\tau\circ\Phi$ is again a natural transformation, $\mathrm{dom}\circ \Phi\to\mathrm{cod}\circ\Phi$.
Using again the condition $\phi=\tau\circ\Phi$, we must have
(*) $\quad\mathrm{dom}\circ\Phi=F\ $ and $\ \mathrm{cod}\circ\Phi=F'$, $\ \ $ because $\phi:F\to F'$.

Now, if $\beta:b\to b'$ is given, $\Phi(\beta)$ is a commutative square in $\mathcal C$, its top and bottom are $\phi_b$ and $\phi_{b'}$, respectively, and (*) says that the left and right sides are $F(\beta)$ and $F'(\beta)$.

| cite | improve this answer | |

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.