2
$\begingroup$

This is a mostly philosophical question. Is it fair to think of usual binomial coefficients and their identities as an $F_1$ case of $q$-binomial coefficients and identities? Here $F_1$ is the field of one element.

$\endgroup$
3
  • 2
    $\begingroup$ Yes, I would say you can. $\endgroup$
    – Ben Webster
    Jan 13, 2015 at 21:38
  • $\begingroup$ This is indeed a very philosophical question. They do count something that may be interpreted as something over $\Bbb F_1$: Subsets are subspaces $\endgroup$ Jan 13, 2015 at 21:55
  • 1
    $\begingroup$ See, for example, arxiv.org/abs/math/0407093. $\endgroup$ Jan 13, 2015 at 22:34

1 Answer 1

4
$\begingroup$

Yes, it is not only fair, but also often very useful. The analogy is that sets are like vector spaces over $\mathbb F_1$, so for example the analog of $\binom{n}{k}$ counting $k$-element subsets of an $n$-element set, is that $\binom{n}{k}_q$ counts the number of $k$-dimensional subspaces of $\mathbb F_q^n$.

In lack of a rigorous theory of $\mathbb F_1$, this usefulness can be illustrated through numerous examples. Take for example the Erdos-Ko-Rado theorem, which is an important starting point of extremal combinatorics. Even though writing down the analogous statement over $\mathbb F_q$ is very easy, by just interchanging subsets with subspaces, the proof turned out to be much harder (the first proof was due to Frankl and Wilson).

The only uniform proofs I know Erdos-Ko-Rado over all characteristics (including characteristic 1) are "geometric" in nature and were discovered much later. However I'm afraid this isn't particularly emphasized in the literature. For example I haven't seen anywhere explicitly mention the analogy between the method of "shadows" and the study of the relevant partial flag variety over $\mathbb F_q$. Still, I think this is a very fun example of interplay between combinatorics of sets and geometry over $\mathbb F_q$ that has produced non-trivial theorems.

$\endgroup$
3
  • $\begingroup$ Where would one look to see a uniform-in-characteristic-including-1 proof (of anything)? I thought that the theory was still so much in flux that such uniformity would be very hard to come by. $\endgroup$
    – LSpice
    Jan 13, 2015 at 22:56
  • 1
    $\begingroup$ @LSpice, I just meant a uniform proof for finite fields, that works when you "translate" it to $\mathbb F_1$. $\endgroup$ Jan 14, 2015 at 1:14
  • $\begingroup$ Henry Cohn gives such a proof in the paper I linked to above (arxiv.org/abs/math/0407093). $\endgroup$ Jan 14, 2015 at 5:39

Your Answer

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

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