Suppose that we have $$ L :C\leftrightarrow D: R$$ an adjoint Quillen pair. We assume that both model categories are combinatorial model categories.

Suppose that the functor $L$ (left adjoint) takes fibrant-cofibrant objects to fibrant-cofibrant objects. I was wondering if it follows that $L$ sends fibrant objects to fibrant objects ?


To find a counterexample, we should choose $C$ to have a lot of fibrant objects, but few cofibrant objects. So let $D$ be a combinatorial model category and let $C$ be the model category structure on the underlying category of $D$ in which every morphism is an acyclic fibration. This is again a combinatorial model category (with empty sets of generating (acyclic) cofibrations), and the identity functor $C \to D$ is a left Quillen functor. In $C$ every object is fibrant, but only the initial object is cofibrant.

In this case the question becomes: if the initial object of $D$ is fibrant, then is every object of $D$ fibrant? And of course there are many counterexamples.

  • $\begingroup$ To make the last sentence concrete, I think a counterexample $D$ is the injective model structure on $Ch(R)$ where $R$ is a self-injective ring. Or, the category of pointed simplicial sets. $\endgroup$ – David White Nov 26 '18 at 13:21
  • $\begingroup$ The counterexample you gave is extreme :) Is there still a "reasonable" additional condition that we could impose on the functor $L$ such that under my assumptions, the functor L will takes fibrant objects to fibrant objects? $\endgroup$ – ABC Nov 26 '18 at 14:57

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.