This question concerns a previous question, Rotation number of inverse maps on the circle. in which all the terminology and notation used below is defined.

The question is given the rotation number $\rho (f)$, calculate the rotation number $\rho (f^{-1})$. All proofs I have found of this statement calculate $\rho(F \circ F^{-1})$ (for $F, F^{-1}$ lifts of $f, f^{-1}$) and end up with something like:

$\rho(F\circ G)=\lim_{n\to\infty}\dfrac{(F\circ G)^n(x)-x}{n}= \lim_{n\to\infty}\left(\dfrac{F^n(G^n(x))-G^n(x)}{n}+\dfrac{G^n(x)-x}{n}\right)=\rho(F)+\rho(G)$

For $F, G$ lifts of commuting circle homeomorphisms $f, g$, and the proof ends by letting $G = F^{-1}$. But in the definition of rotation number for $f$, i.e. $\rho(F) = \lim_{n\to\infty}\dfrac{F^n(x)-x}{n} = \lim_{n\to\infty}\dfrac{F^n(x)}{n}$, if we let $x = F^{-n}(y)$ for some $y$ as above we get that the rotation number is $\rho(F) = \lim_{n\to\infty}\dfrac{y}{n} = 0$ for any $f$ (note that the rotation number is always defined and has the same value for any $x$ in the domain of $F$).

So my question is are we allowed to use a function $G^n(x)$ in place of a variable $x$ in the definition of rotation number as the proofs commonly do, or are these proofs incorrect?


You are pretty confused. They are used as such in the definition. When you substitute $x = F^{n}(y)$, it changes to $y - F^{-n}y$. Since $f$ and $g$ commutes, their lifts too. Only that property is being exploited here.

  • $\begingroup$ I agree I'm confused, but what about when you use the equivalent definition of the rotation number without the $-x$ term in the numerator? The $F^n(x)$ and $x = F^{-n}(y)$ cancel each other out to get $y / n$ as $n \to \infty$ surely? The reason I ask is my course uses the definition $\rho(F) = \lim_{n\to\infty}\dfrac{F^n(x)}{n}$, without a $-x$ in the numerator. $\endgroup$ – D. P Apr 28 '16 at 15:36
  • $\begingroup$ @D.P Let $f$ be the anticlockwise rotation of $\Bbb{S}^1$ with angle $2\pi \theta$. Then what do you think the rotation number is ? $\endgroup$ – Mambo Apr 28 '16 at 19:25
  • $\begingroup$ $f(x) = x + 2\pi\theta \mod 1$, so lift $F(x) = x + 2\pi\theta$ and $F^n(x) = x+ 2n\pi\theta$ giving $\rho(F) = 2\pi\theta$, so $\rho(f) = 2\pi\theta \mod 1$. $\endgroup$ – D. P Apr 28 '16 at 19:35
  • $\begingroup$ If $f(z) = e^{2\pi i \theta }z$, then $F(x) = x + \theta$ will be lift and $F^n(x) = x + n\theta$. Then $\rho(f) = \theta \mod 1$. How is this possible by your definition? $\endgroup$ – Mambo Apr 28 '16 at 20:01
  • 1
    $\begingroup$ Ignore everything I have written on this comment section, I have finally realised that I am being dumb and the counterexample is wrong because $y = F^n(x)$ so $\dfrac{y}{n} \to \rho(F)$. $\endgroup$ – D. P May 1 '16 at 17:14

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.