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

Let $E$ be a spectrum. Then $E \wedge E$ is a $\mathbb{Z}/2$-spectrum in the naivest possible sense, i.e., an object with $\mathbb{Z}/2$-action in the (∞,1)-category of spectra. Can I make it a $\mathbb{Z}/2$-spectrum in the less naive, but still not genuine, sense? (That is, a $\mathbb{Z}/2$-spectrum indexed on the trivial universe.)

I'm thinking of something like the following. I may represent $E$ as an (reduced & continuous) excisive functor from pointed spaces to pointed spaces. Then define

$$G(X) = \mathrm{colim}_{I \times I} \mathrm{Map}(S^{x_1} \wedge S^{x_2}, E(S^{x_1}) \wedge E(S^{x_2}) \wedge X)$$

where $I$ is the category of finite sets and inclusions. Hopefully $G$ is a functor from spaces to $\mathbb{Z}/2$-spaces. If I forget about the $\mathbb{Z}/2$-fixed point set, I can think of it as $E \wedge E$ with its $\mathbb{Z}/2$ action. What spectrum does $G(X)^{\mathbb{Z}/2}$ correspond to? Is there a more familiar name for it? Edit: I seem to be getting $E \vee (E \wedge E)^{h\mathbb{Z}/2}$, but without much confidence.

[Leftover part of the question: If so, by my question here I can think of the resulting object as a functor from the opposite of the orbit category of $\mathbb{Z}/2$ to spectra. Unpacking this amounts to giving some spectrum $F$ together with a map $F \to (E \wedge E)^{h\mathbb{Z}/2}$. What is $F$?]

share|cite|improve this question
By your other question, lifting E ^ E to a Z/2-spectrum is equivalent to picking F with such a map, and F (the fixed point object) can be anything. Do you want some relation of F to E? – Tyler Lawson Dec 2 '09 at 6:12
I sketched the construction I had in mind. Please tell me if I am saying something which is obviously wrong! – Reid Barton Dec 2 '09 at 6:36
up vote 3 down vote accepted

Given a spectrum $E$ there is a "standard" lift of $E \wedge E$ to a $\mathbb{Z}/2$-spectrum using the basic technique you describe. One way to describe it as follows.

You can construct the category of genuine $\mathbb{Z}/2$-spectra (indexed on the full universe) via collections of $\mathbb{Z}/2$-spaces $X\_n$ with equivariant structure maps $\sigma: S^V \wedge X\_n \to X\_{n+1}$, where $S^V = S^1 \wedge S^1$ is the 1-point compactification of the regular representation $\mathbb{R} \times \mathbb{R}$ with the "flip" action. Under this description, if $E$ is a spectrum made up of spaces $E\_n$ and structure maps $S^1 \wedge E\_n \to E\_{n+1}$, then you can construct $E \wedge E$ as a genuine $\mathbb{Z}/2$-spectrum with spaces $E\_n \wedge E\_n$ and structure maps $S^1 \wedge S^1 \wedge E\_n \wedge E\_n$ that simply twist and apply the structure map on each factor. (This, e.g., is one way to pass forward the equivariant structure on $TC$).

This fully genuine spectrum has an underlying spectrum indexed on the trivial subgroup. The $\mathbb{Z}/2$-fixed point object is the homotopy pullback of a diagram

$$ (E \wedge E)^{h\mathbb{Z}/2} \to (E \wedge E)^{t\mathbb{Z}/2} \leftarrow E $$

where $Z^{t\mathbb{Z}/2}$ is the so-called "Tate spectrum" of $Z$, which is to Tate cohomology as the homotopy fixed point spectrum is to group cohomology.

If $E = \Sigma^\infty W$ for a space $W$ then the map from $E$ to the Tate spectrum lifts (via the diagonal) to a map to the homotopy fixed point spectrum, and so the homotopy pullback will actually be homotopy equivalent to $E \vee (E \wedge E)_{h\mathbb{Z}/2}$. (This homotopy orbit is the fiber of the map from homotopy fixed points to Tate fixed points.)

For finite spectra the map from $E$ to its Tate spectrum is 2-adic completion; this is one way to state the content of the Segal conjecture that Carlsson proved (at least at the prime 2). Sverre Lunoe-Nielsen extended this result to a number of other spectra like the Brown-Peterson spectra. In these cases the fixed-point object is equivalent after 2-adic completion to the homotopy fixed point object.

All the above plays out the same way for a cyclic group of prime order.

share|cite|improve this answer
Is there a standard name for this object (the pullback of that diagram)? Do you know a place to read about it? – Reid Barton Dec 3 '09 at 2:38
I've only seen it referred to as the fixed point object $(E \wedge E)^{\mathbb{Z}/2}$ with the understanding that $E \wedge E$ has a standard genuine equivariant structure. Any $\mathbb{Z}/2$-spectrum has its fixed points living in such a pullback diagram; you can look in Greenlees-May's "Generalized Tate cohomology" book or some of the literature related to TC like Ausoni-Rognes, Hesselholt-Madsen, Madsen's survey article, or Tsalidis' paper. – Tyler Lawson Dec 3 '09 at 14:09

Here's a very simple way to obtain $(E \wedge E)^{\Bbb Z_2}$ without having to resort to representations (at least if $E$ is connective). Consider the functor from spectra to spectra given by $$E \mapsto \Sigma^\infty \Omega^\infty E .$$ That is, the suspension spectrum of the zeroth space of $E$.

Let $$E \mapsto P_2(E)$$ be its quadratic approximation in the sense of Goodwillie's calculus of homotopy functors (that is, the second stage of the Goodwillie tower).

Then $P_2(E)$ has the homotopy type of $(E \wedge E)^{\Bbb Z_2}$ whenever $E$ is connective.

Incidentally, one can also see that this comes equipped with a fiber sequence $$ D_2(E) \to P_2(E) \to E $$ which amounts the tom Dieck's splitting in the case when $E$ is a suspension spectrum. Here, $D_2(E) = (E\wedge E)_{h\Bbb Z_2}$ is the quadratic construction. In general, this sequence needn't split.

share|cite|improve this answer

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.