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

Which of the statements is wrong:

  1. a generalized cohomology theory (on well behaved topological spaces) is determined by its values on a point
  2. reduced complex $K$-theory $\tilde K$ and reduced real $K$-theory $\widetilde{KO}$ are generalized cohomology theories (on well behaved topological spaces)
  3. $\tilde K(*)= \widetilde{KO} (*)=0$

But certainly $\tilde K\neq \widetilde{KO}$.

share|cite|improve this question
up vote 12 down vote accepted

1 is doubly wrong. First, you need to distinguished generalized cohomology theories and reduced generalized cohomology theories. If you want to work with the latter, you should replace "a point" in 1 by "$S^0$", and then the corrected version of 3 no longer holds. But even this new version 1' is false; a generalized cohomology theory is not determined by its coefficients, unless they are concentrated in a single degree (example: complex K-theory vs. integer cohomology made even periodic).

share|cite|improve this answer
Thanks, Reid. Someone should change the "Generalized cohomology theories" section in Wikipedia's article on "Cohomology". – roger123 Mar 17 '10 at 19:14
Hmm, I see. Well, Wikipedia does use quotes around "determined by its values on a point" :) but I agree that this statement is misleading at best. – Reid Barton Mar 17 '10 at 19:24
It could easily be fixed by putting a link to the entry for the Atiyah-Hirzebruch spectral sequence as a way of making this precise. – Andy Putman Mar 17 '10 at 19:27
We're thinking of generalized cohomology theories as taking values in graded abelian groups. Thus $\tilde{K}^q(S^0)\not\approx \tilde{KO}^*(S^0)$, for instance if $q\equiv -1,-2,-6\mod 8$. – Charles Rezk Mar 17 '10 at 19:51
But the statement about cohomology theories being determined by their coefficients is not totally wrong -- if you have a natural transformation $K \to L$ of homology theories which induce an isomorphism on the point (or $\mathbf{S}^0$ in the case of reduced theories), then it's an isomorphism. But you do need to have a natural transformation in the first place. – Tilman Mar 17 '10 at 22:08

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.