Let $k$ be a finite field, $THH(k)$ its topological Hochschild homology spectrum. For essentially formal reasons, we know that it's an $E_\infty$-algebra over the Eilenberg-Mac Lane spectrum $Hk$, and so can be modeled by an $E_\infty$-dga over $k$.

Question: Is this $E_\infty$-structure equivalent to a strictly commutative multiplication?

We know that the homotopy algebra $\pi_* THH(k)$ is a polynomial ring on a generator in degree $2$, so in particular it cannot be equivalent to a simplicial commutative $k$-algebra (for then it would support a divided-power structure relative to its ideal of positive-degree elements.) So in a sense a positive answer to the question above would say that $THH(k)$ is "as commutative as it could possibly be" given its homotopy algebra.


This is not the case, and you can use Dyer-Lashof operations to see so. In the following I'll show this for $k = \Bbb F_2$ because that's the easiest case to compute. Bokstedt proved that $THH_*(\Bbb F_2) = \Bbb F_2[\sigma]$ where $|\sigma| = 2$. If it came from a commutative DGA, then the only nonzero Dyer-Lashof operations would be the squaring operations (because the operad action would factor through the projection down to the bottom class). We'll show that $Q^6 \sigma = \sigma^4$.

The description of $THH(k)$ by the cyclic bar construction makes it the derived smash product $k \wedge_{k \wedge^{\Bbb L} k^{op}} k$. The associated Kunneth spectral sequence comes from a simplicial commutative $k$-algebra, which gives the associated spectral sequence compatible Dyer-Lashof operations. It sits inside the Bokstedt spectral sequence (this is how Bokstedt computed the hidden extensions to show $THH_*(\Bbb F_2) = \Bbb F_2[\sigma]$). In particular, this spectral sequence takes the form of a Tor-spectral sequence $$ Tor_{A_*} (\Bbb F_2, \Bbb F_2) \Rightarrow THH_*(\Bbb F_2) $$ where $A_*$ is the dual Steenrod algebra $\Bbb F_2[\xi_i]$. The Tor-groups form an exterior algebra generated by $I/I^2$, where $I$ is the augmentation ideal of the dual Steenrod algebra. The exterior algebra is therefore generated by $[\xi_i]$.

However, Steinberger's calculations from the $H_\infty$ book tell us that $$Q^6 \xi_1 =\xi_1^7 + \xi_1^4 \xi_2 + \xi_1 \xi_2^2 + \xi_3,$$ and therefore in $THH(k)$ we have an identity $$Q^6 [\xi_1] = [\xi_3]$$ because everything else is decomposable. The spectral sequence degenerates and so $Q^6$ on the generator in degree two hits the only nonzero thing in degree 8.

  • 1
    $\begingroup$ Thanks, Tyler, this is very useful. I'm a little confused by your statement that "The associated Kunneth spectral sequence comes from a simplicial commutative $k$-algebra," though. I believe that it comes from a simplicial $E_\infty$-algebra over $k \wedge^\mathbb{L} k^{op}$, but if I'm not mistaken that thing is not a simplicial commutative $k$-algebra. Perhaps I'm misunderstanding what you mean there? $\endgroup$ – Aaron Royer May 22 '15 at 20:08
  • 2
    $\begingroup$ @Aaron Ah, that's because I meant an $Hk$-algebra, meaning over the Eilenberg-Mac Lane spectrum. Sorry. $\endgroup$ – Tyler Lawson May 22 '15 at 20:32
  • $\begingroup$ Awesome, that makes me much happier. Thanks again! $\endgroup$ – Aaron Royer May 22 '15 at 20:36
  • 1
    $\begingroup$ It might be worth noting that "compatible Dyer-Lashof operations" is a phrase that should be interpreted with care. For example, you do not seem to mean that there are operations on the $E_2$-page that behave like the DL operations but rather that there is a structure on the filtration that gives rise to a theory of operations whose algebraic behavior may be unpredictable. $\endgroup$ – Sean Tilson May 25 '15 at 14:07
  • 3
    $\begingroup$ The compatibility can be expressed by saying that the suspension operator from the (mod p) homology of any E-infinity ring spectrum R to the homology of THH(R) commutes with the Dyer-Lashof operations. See Bökstedt's original preprints, or my paper with Angeltveit. $\endgroup$ – John Rognes May 28 '15 at 22:55

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.