Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Let $M$ be a compact, smooth Riemannian manifold with tangent bundle $TM$. I will not distinct between $TM$ and the associated $O(n)$-frame bundle. I believe the following statements are true, but if they are not, please correct me.

1) An orientation of $M$ is an equivariant lift of $TM$ to a $SO(n)$-bundle under the natural inclusion $SO(n) \hookrightarrow O(n)$. The only obstruction to the existence of orientation of $M$ is the first Stiefel-Whitney class $w_{1}(TM) \in H^{1}(M, \mathbb{Z} _{2})$. Moreover, if one exists, then the set of orientations is a $H^{0}(M, \mathbb{Z} _{2})$-torsor.

2) A spin structure on an oriented manifold is an equivariant lift of the $SO(n)$-bundle of oriented frames in $TM$ under the natural projection $Spin(n) \rightarrow SO(n)$. The only obstruction to the existence of a spin structure on a manifold is the class $w_{2}(TM) \in H^{2}(M, \mathbb{Z}_{2})$. Moreover, if one exists, then the set of spin structures is a $H^{1}(M, \mathbb{Z}_{2})$-torsor.

Are these phenomena somehow naturally related? How can one see this? Do there exist some "higher structures" on the tangent bundle such that higher Stiefel-Whitney classes are obstructions to their existence?

I have been trying to push this analogy even further and realized, that $SO(n) \rightarrow O(n)$ is the universal recipient of group homomorphisms from connected topological groups $G$ into $O(n)$. Moreover, by covering theory, $Spin(n) \rightarrow SO(n)$ is the universal recipient of group homomorphisms from 1-connected topological groups G into $SO(n)$. I have no idea how to extend this viewpoint further, since $\pi_{2}$ always vanishes for Lie groups, in particular it vanishes for $Spin(n)$.

share|improve this question
    
This is discussed on p. 82 of Michelsohn and Lawson's spin geometry. As you point out, $\pi_2(G)$ always vanishes for Lie groups and if $\pi_3(G) = 0$ then $G$ is contractible (so any $G$-bundle is trivial). –  Eric O. Korman Feb 7 '13 at 17:12
add comment

1 Answer 1

up vote 2 down vote accepted

There are higher structures on vector bundles, but the pattern of obstructions being Stiefel-Whitney classes ends with spin structures. I posted a relevant answer on MathOverflow.

You're pushing the analogy in the correct way. The relevant construction here is the Whitehead tower of the orthogonal group. Going up a step in the Whitehead tower kills the lowest homotopy group. Now we can't continue getting finite-dimensional covering groups once we get to $\mathrm{Spin}(n)$ in the Whitehead tower for $\mathrm{O}(n)$, but we can get an infinite-dimensional $3$-connected Lie group $\mathrm{String}(n)$. Then a string structure on a spin vector bundle is a lift of the structure group through $B\mathrm{String}(n) \longrightarrow B\mathrm{Spin}(n)$. The obstruction to the existence of a string structure is not a Stiefel-Whitney class as the analogy may suggest, but it is instead the fractional Pontrjagin class $\tfrac{1}{2}p_1$.

You can go a step further and form the $7$-connected cover $\mathrm{Fivebrane}(n)$ of $\mathrm{String}(n)$. Then a fivebrane structure on a string vector bundle is a lift of the structure group through $B\mathrm{Fivebrane}(n) \longrightarrow B\mathrm{String}(n)$. The obstruction to the existence of a fivebrane structure on a string vector bundle is the fractional Pontrjagin class $\tfrac{1}{6}p_2$.

share|improve this answer
    
I am very thankful for this answer. The nLab article you mention (along with "String structure") clarifies everything. –  Piotr Pstrągowski Feb 7 '13 at 17:57
add comment

Your Answer

 
discard

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.