Consider the page from the book by Adamek & Rosicky: Locally presentable and accessible categories. given below. I need to derive this: Here is the statement

(2) to prove the universal property of $f$, it is sufficient to show that for each $\lambda-$presentable object $H$, each morphism $h:H\to B$ merging $p$ and $q$ factorizes uniquely through $f$.We know that $f$ factorizes through some $b_i$, say $h=b_i\cdot h'$. Then $$d_i^*\cdot (p_i\cdot h')=p\cdot b_i\cdot h'.$$

My question is why this equation holds? All the relevant letters are given here: PAGE FROM THE BOOK

And here is the ERRATA to that page from the book.


The equation follows from

$$ d^*_i \cdot p_i = p \cdot b_i $$

and this equation holds because of the definition of $p\colon B\to B^*$.

Recall that two diagrams of shape $I$ are introduced on page 88:

  • $D\colon I\to {\cal K}$ with $D_i=B_i$ and $D(i,j)=b_{i,j}$

  • $D^*\colon I\to {\cal K}$

These have colimits $B=\mathrm{colim} B_i$ and $B^*=\mathrm{colim} D^*_i$, together with maps $b_i\colon B_i\to B$ and $d^*_i\colon D^*_i \to B^*$.

The family of maps $p_i\colon B_i\to D^*_i$ forms a natural transformation from $D$ to $D^*$. Therefore the maps $d^*_i \cdot p_i\colon B_i \to B^*$ form a cocone which factors through the colimit cocone via a unique map $p: B\to B^*$. Similar for the $q_i\colon B_i\to D^*_i$ and $q$.

  • $\begingroup$ Thank you for your detailed and clear answer. I just do not follow where (and if) we have used the uniqueness of such a $p$? $\endgroup$ – user122424 Feb 14 '19 at 17:09
  • 1
    $\begingroup$ As far as I can see, uniqueness of such a $p$ was not used so far. $\endgroup$ – Marc Olschok Feb 14 '19 at 18:35

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.