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

In his ICM 2002 talk (Topology of singular algebraic varieties, available also on arXiv) B. Totaro says on p. 3 (of the arXiv version): "Using the method of Guillen and Navarro Aznar I was able to define the weight filtration for complex and real analytic spaces." However, no reference to this statement is given there. The definition is not explicitly given either, but from the preceding discussion a reasonable guess seems to be that one considers simply the Leray filtration of the open embedding into some compactification. The result will depend on the compactification, but if one compactification dominates another one, then the weight filtrations are the same (ibid, theorem 2.2).

I wasn't able to find a proof of the latter statement in the literature and I would like to ask if anyone knows a reference.

share|cite|improve this question
Here's a link to the arxiv version – j.c. Apr 14 '10 at 15:44

For a smooth complex algebraic variety X, the weight filtration is as you say the Leray filtration, up to reindexing, for a good compactification (the complement of should be a normal crossing divisor). For singular X, one needs to choose a diagram $X\leftarrow X_\bullet \to \bar X_\bullet$ such that the first arrow satisfies cohomological descent and the second arrow is a good compactification of a smooth simplicial (or cubic) scheme. See Deligne's Theorie de Hodge II, III, or Navarro Aznar et. al [Springer LNM 1335]. Although this approach hinges on certain Hodge theoretic facts that won't generalize (see comments), there is a refinement which I suspect will. Gillet-Soul\'e [Crelles 1996] showed that if one uses a slightly stronger descent property, then it determines a certain well defined complex $W(X)$ in the homotopy category of pure Chow motives. It is essentially the mapping cone $$Cone ([\bar X_\bullet]\to [\bar X_\bullet- X_\bullet])$$ where I write [...] for the complex of motives associated to a smooth (bi)simplicial scheme. $W(X)$ can be used to define an integral weight filtration on compactly supported cohomology. I think that Guillen-Navarro Aznar [IHES 2002] get a similar result.

I suspect that Totaro has some sort of refinement of these ideas, but you should simply ask him.

share|cite|improve this answer
Donu, thanks! I also imagine that Totaro is following Deligne's strategy. What I don't see how to prove is the independence of the compactification in the equivalence class of compactifications. Deligne's strategy is this (Theorie de Hodge 2, p. 37-38): something is an isomorphism and a morphism of mixed Hodge structures, hence it is an isomorphism of mixed Hodge structures. This doesn't work if we have the weight filtration alone, without the Hodge filtration and the compatibility between the two. – algori Apr 13 '10 at 19:47
I suppose he couldn't be using Deligne's method directly. Also I just noticed he's always working with arbitrary coefficients. I do have a guess as to what he's doing, but I won't be able to write anything coherent right now. – Donu Arapura Apr 13 '10 at 20:07

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.