In his beautiful paper On the cohomology and K theory of the general linear group over a finite field, Quillen constructs (if I understand correctly) an isomorphism on connected components of K-theory $K(\mathbb{F}_q)\mid_{\ge 1} \cong \text{Fib}(1-\Psi^q:ku\to ku).$ His construction, while very cool, seems to be based on a miracle: that the Brauer lift produces complex representations from mod-$p$ ones. It also uses the +-construction for defining K-theory (though perhaps it's not that hard to update his construction to work with Segal formalism)?

I'm curious whether there is a more modern perspective on this equivalence of spectra, maybe using cyclic structure on topological Hochschild homology.

  • 2
    $\begingroup$ For what it's worth, many years ago (2006?) I posed this question (more or less) to an expert on trace methods, and the response seemed to be that no such proof of Quillen's result is known. It's a little uncler whether Quillen was aware of Segal's K-theory construction and the Q-construction before he wrote the paper in question. My impression is that invented the plus-construction specifically for this calculation. One could probably determine the dates by looking through Quillen's notebooks. $\endgroup$
    – Dan Ramras
    Aug 8, 2016 at 18:42
  • 1
    $\begingroup$ Quillen's famous Letter to Segal (where he talks about "freeing myself from the shackes of the simplicial way of thinking" and discovering the category Q) is from July 1972 - he says he discovered Q in Spring '72 - and the Annals paper in question was published in November 1972. There was an earlier preprint of the Annals paper, though, which is on the K-theory archive (math.uiuc.edu/K-theory/1006), but it's undated. $\endgroup$
    – Dan Ramras
    Aug 8, 2016 at 18:43
  • $\begingroup$ Didn't Dwyer say something about this at the 2012 WCATSS? $\endgroup$ Aug 10, 2016 at 12:17
  • $\begingroup$ @SeanTilson I don't recall anything about this from WCATSS. $\endgroup$
    – Dan Ramras
    Aug 13, 2016 at 16:48
  • 1
    $\begingroup$ The argument using Brauer lifting is a more complicated rearrangement to eliminate étale topology from his original very simple argument which appears in his 1970 ICM address, which proves the more general theorem that $BG(\mathbb F_q)^+[1/p]$ is the fiber of Frobenius on $BG[1/p]$. The cohomology of the connected algebraic group is the same as the compact group because the methods of computation (eg, Schubert cells) are all algebraic. $\endgroup$ Sep 22, 2020 at 21:53

1 Answer 1


Firstly, there is a proof using the motivic spectral sequence (the Atiyah-Hirzerbruch style spectral sequence from motivic cohomology to algebraic $K$-theory). This is written in the master's thesis of Gabe Angelini-Knoll.

Gabe and Andrew Salch are also working to answer this question and a paper is apparently due. From Andrew's website:

"My student Gabe Angelini-Knoll and I have been working on the problem of computing the Waldhausen algebraic K-groups of the algebraic K-theory spectra of certain finite fields. This is an example of "iterated K-theory" and Rognes' redshift conjecture is not known in these cases. Thus far, Gabe and I have (with the aid of a new "THH-May" spectral sequence for computing topological Hochschild homology) computed the homotopy groups of THH(K(F_q)) smashed with the p-primary Smith-Toda complex V(1), for p > 3 and for many (but not all) prime powers q. Gabe is working on the computations of the homotopy groups of the C_p fixed points of this spectrum (this will probably be Gabe's thesis), with the goal of using trace methods to recover the K-groups of K(F_q).

We expect to post and submit our first two papers on this topic before the end of summer 2016. "

  • 5
    $\begingroup$ In case anyone is curious about the papers of Andrew Salch and I that are referred to here, they are now on the arXiv: link and link I should also mention that Eva Honing, a current student of Ausoni, has been working on THH of more general finite fields whereas I have restricted my attention to fields of order q where q is prime power that topologically generates the p-adic units. $\endgroup$ Feb 15, 2017 at 16:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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