MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I'm trying to understand the relationship between the notions of looping and delooping in category theory and topology.

The morphisms in a category with one object have the structure of a monoid. Similarly, a 2-category with one object is a monoidal (1-)category, and so every monoidal category can be thusly viewed as a 2-category with one object by promoting the 1-morphisms to 2-morphisms and the objects to 1-morphisms, just as every monoid $M$ can be viewed as a 1-category $\underline{M}$ with one object by promoting the elements to morphisms.

I have inferred from some context that this process may be called looping/delooping, though I haven't found a reference for this, and the nLab article says stuff which I don't understand. But it's easy enough to see that if $C$ is a category, then the morphisms of $C$ are the objects of $C^I$, where $I=\[0\to1\]$ is the interval category, and this is analogous to the loop space construction in topology $\Omega X=[S^1,X]$. So by demoting morphisms to objects, you get the loop space. In topology for any group we also have $BG$ which satisfies $\Omega BG=G$, hence is called the delooping of $G$; the space whose loop space is $G$. Continuing the analogy, taking a category and promoting its objects to 1-morphisms and 1-morphisms to 2-morphisms, this process may be called "delooping".

For any group $G$, in addition to the delooping $\underline{G}$ category with one object, we also have the "translation category" $TG$ with one object for each group element and exactly one morphism for each pair. $TG$ is a monoidal category, so call its delooping $\overline{G}.$ Like $\underline{G},$ it has one object, and one morphism for every group element. But $\overline{G}$ also has 2-morphisms, exactly one for every pair of 1-morphisms, since it is the delooping of a 1-category, whereas $\underline{G}$ was the delooping of a discrete category (0-category).

So we might consider $G$ the delooping of $TG$ in the category theoretic sense of the word, then we apply the geometric realization of the nerve to both categories, and we get $BG$ and $EG$, the former being the topological delooping.

Is there some sense in which $EG$ is the looping of $BG$, to finish the analogy? It seems not, since $\Omega BG=G$, whereas $EG$ is contractible. Is my definition of looping and delooping of categories correct? What is the relationship between the delooping of a category and the delooping of a space? What is a good reference for understanding this stuff?

share|cite|improve this question
EG is not the looping of BG. Your TG is a 2-group, and so naturally a 2-groupoid with one object. This 2-groupoid is the delooping of TG, and TG is the looping of this 2-groupoid. Just as G sits inside TG, the delooping of G (the one-object groupoid) sits inside the delooping of TG. The quotient of TG by G is the delooping of G, and the same story for the other inclusion. If you suspect there is some iteration going on, you are not wrong. May I suggest and for some related results. – David Roberts Apr 24 '13 at 1:08

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.