Category Theory
Zulip Server
Archive

You're reading the public-facing archive of the Category Theory Zulip server.
To join the server you need an invite. Anybody can get an invite by contacting Matteo Capucci at name dot surname at gmail dot com.
For all things related to this archive refer to the same person.


Stream: learning: questions

Topic: pullback-preserving non-PRA functor


view this post on Zulip James Deikun (May 09 2025 at 09:32):

Does anyone have an example of a pullback-preserving functor that is not a parametric right adjoint?

view this post on Zulip James Deikun (May 09 2025 at 13:21):

Actually the inclusion of FinSet\mathbf{FinSet} in Set\mathbf{Set} is one of these, but it doesn't really work as an example for me, so I'm going to amend and ask for an endofunctor.

view this post on Zulip Mike Shulman (May 09 2025 at 15:51):

Any functor that preserves finite limits but not arbitrary limits should work. Like sheafification for almost any nontrivial Grothendieck topology.

view this post on Zulip James Deikun (May 09 2025 at 18:13):

Ah, right, a lex reflector is an endofunctor ...

view this post on Zulip Chaitanya Leena Subramaniam (May 09 2025 at 23:30):

(I'm assuming you mean arbitrary wide pullbacks, not just binary/finite pullbacks --- in which case see Mike's comment)

Any accessible T:CCT:\mathcal C\to\mathcal C preserves wide pullbacks iff TT admits strict generic factorisations.
(Weber, "Generic morphisms, parametric representations and weakly cartesian monads", Thm 6.6)

If T:CCT:\mathcal C\to\mathcal C admits strict generic factorisations, then TA:C/AC/TAT_A : \mathcal C/A\to\mathcal C/TA is a right adjoint for every AA in C\mathcal C (i.e. TT is a local right adjoint).
(Weber, "Familial 2-functors and p.r.a.'s", Prop. 2.6)

So a wide pullback-preserving endofunctor TT can fail to be p.r.a. in two ways:

  1. TT is not accessible---I can't think of an example.
  2. TT is accessible, and is a local right adjoint, but C\mathcal C does not have a terminal object --- a nice example is the free-polynomial-monad monad on the category PolyEnd of polynomial endofunctors in Set (see J. Kock's "Polynomial functors and trees")

view this post on Zulip Mike Shulman (May 09 2025 at 23:38):

Inaccessible functors are generally rather tricky to get ahold of. Here is an example of an inaccessible functor between locally presentable categories that preserves all limits (hence a fortiori all wide pullbacks) but is not a right adjoint. It's not an endofunctor, but note that requiring the domain and codomain to be locally presentable also excludes examples like FinSetSet\rm FinSet \hookrightarrow Set.

view this post on Zulip Chaitanya Leena Subramaniam (May 09 2025 at 23:40):

Nice example!

view this post on Zulip Mike Shulman (May 09 2025 at 23:50):

An example of an inaccessible endofunctor (of Set, even!) that preserves finite limits (and also finite colimits!) is the ultrapower by the defining ultrafilter on a measurable cardinal (assuming such exists). In fact, Andreas Blass showed in "Exact functors and measurable cardinals" that the existence of a measurable cardinal is equivalent to the existence of a nonidentity endofunctor of Set that preserves finite limits and colimits.