MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

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

The following arose when I was thinking about a talk at the Midwest Topology Seminar:


I want to consider a generalization of a Waldhausen-like structure on a category $C$ with 0-object $\ast$. Namely, I will have a category $\text{co}C$ consisting of "cofibrations," but I will relax the usual axioms to be:

Cof1: The isomorphisms of $C$ are cofibrations;

Cof2: The arrow $\ast \to A$ is a cofibration for any object $A$ of $C$.

Cof3': Pushouts of cofibrations along cofibrations are always defined, and all arrows in the the pushout will be cofibrations.

(Note that Cof3' is different from Waldhausen's usual axiom. In particular, quotient objects are not necessarily defined.)

For my purposes, I am happy to work with the isomorphisms as the weak equivalences.

Using the above structure, one can apply a construction due to Thomason to form a simplicial category $iT.C$. An object in simplicial degree $n$ is a string of cofibrations $$ C_\bullet \quad \equiv \quad C_0 \rightarrowtail C_1 \rightarrowtail \dots \rightarrowtail C_n $$

A morphism is filtration preserving map $C_\bullet \to D_\bullet $, such that each $C_k \to D_k$ is a cofibration and each square

$$ C_{k-1} \to C_k $$ $$ \downarrow \qquad \quad \downarrow $$ $$ D_{k-1} \to D_k $$ is a pushout (Note: I do not impose the condition that the vertical maps be weak equivalences in this set up.)

The simplicial structure is defined by dropping the $k$-th term of a filtration or inserting the identity in the appropriate place.

Waldhausen showed that Thomason's construction is homotopy equivalent to the $wS.$-construction whenever the above structure arises from a bona fide Waldhausen category structure on $C$. The idea, roughly is given by mapping the filtration above to the filtration $C_1/C_0 \rightarrowtail \cdots \rightarrowtail C_n/C_0$, provided we incorporate quotient data into the definition. (See Waldhausen LNM1126, p.334.)

My Question

My question is basically whether a version of the additivity theorem holds in this context. To formulate this, consider the category whose objects are pushout squares of objects of $C$: $$ A \to B $$ $$ \downarrow \quad \quad \downarrow $$ $$ C \to D $$ in which all displayed maps are cofibrations. Morphisms are defined in the obvious way (natural transformations of such diagrams) It seems to me that this diagram category is also equipped a cofibration structure as above.

Then we may consider the functor which sends this diagram to $$ D \vee A $$ as well as the functor which sends the diagram to $$ B \vee C $$ These are both exact functors, and one can ask:

Question: Are these functors homotopic after passing to $iT_\bullet$ constructions?

share|cite|improve this question
Could you give an example of a category which satisfies the definition above but not Waldhausen's original definition? And secondly, do you really want to drop stability of acyclic cofibrations under pushouts? (Or did you just drop it because you assumed that weak equivalences are isomorphisms?) – Karol Szumiło Oct 30 '12 at 6:47
The following paper may be interesting for you, maybe you already know it, look also for other papers by the same authors: MR2764905 Blumberg, Andrew J.(1-TX); Mandell, Michael A.(1-IN) Algebraic K-theory and abstract homotopy theory. (English summary) Adv. Math. 226 (2011), no. 4, 3760–3812. – Fernando Muro Oct 30 '12 at 8:41
@Karol: An example is provided in Waldhausen's "Manifold Approach" paper. If $X$ is a fixed manifold, call a partition a compact codimenson zero submanifold of $X \times I$ which contains $X\times 0$. Partitions form a poset with respect to inclusion. This gives a category with cofibrations in the above sense. – John Klein Nov 1 '12 at 1:14
@Fernando: I looked at the paper you mention above. I could not see how that would be of help to me. – John Klein Nov 1 '12 at 1:15
@Karol: Here's a link to Waldhausen's paper:… – John Klein Nov 1 '12 at 1:16

I tried thinking about this directly, via a direct homotopy as in the paper, but didn't get anywhere. Suppose we pin down the definition of iT. a bit so that in this case the vertical maps are cofibrations, too; otherwise, the use of squares that "happen" to be pushouts seems unnatural. Now try to construct the homotopy bit by bit, starting with the 0,0-cells (vertices) of the source and progressing to the 1,0-cells and the 0,1-cells. Nothing reasonable seemed to pop up.

share|cite|improve this answer
Dan, thanks for trying anyway. – John Klein Jul 3 '13 at 16:09

Your Answer


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

Not the answer you're looking for? Browse other questions tagged or ask your own question.