3
$\begingroup$

I'm trying to show that given a $1$-form $\omega$ and a $k$-form $\alpha$ such that $\alpha \wedge \omega = 0$ then there exists a $(k-1)$-form $\beta$ such that $\alpha = \omega \wedge \beta$.

I'm struggling to show this even in the case of multi-linear algebra and forgetting forms. I've been trying to use the fact that if $v_i$ are a basis for $V$ then $\{v_{i_1}\wedge...\wedge v_{i_k} : 1 \le i_1 < ... < i_k \le n\}$ is a basis for $\bigwedge ^k V$.

I thought perhaps we could extend $\omega$ (I'm talking just in the multilinear algebra case here and not thinking about manifolds - I'll worry about that after!) to a basis of $V$ and then $\alpha \wedge \omega = 0$ kills any terms containing $\omega$ in the expansion of $\alpha$ but I'm struggling to make any further progress,

Thanks for any help

$\endgroup$
3
  • $\begingroup$ What is $\theta$ here? Is that supposed to be $\omega$? $\endgroup$ Commented May 13, 2016 at 20:17
  • $\begingroup$ Corrected, thanks $\endgroup$
    – Wooster
    Commented May 13, 2016 at 20:18
  • $\begingroup$ You're close. You wrote, "$\alpha \wedge \omega = 0$ kills any terms containing $\omega$ in the expansion of $\alpha$." But actually it's the opposite: this equation only says something about the terms not containing $\omega$. $\endgroup$
    – Jack Lee
    Commented May 13, 2016 at 20:37

1 Answer 1

2
$\begingroup$

In terms of the multilinear algebra, it looks like you're on the right track. We extend $\omega$ to a basis $\omega, v_2, \dots, v_n$ and then write

$$\alpha = \sum a_I \omega \wedge v_{i_2} \wedge \dots \wedge v_{i_k} + $$ $$ \sum b_J v_{j_1} \wedge \dots \wedge v_{j_k} $$

Then as you've noted, wedging with $\omega$ kills everything in the first term. Now if you look at the second term of the above after wedging with $\omega$, what does this say about the constants $b_J$?

As for the case for forms, it's not literally true as stated, you probably need something like $\omega$ is never zero. For example take $\alpha = dx \wedge dy$ and $\omega = xdx$ on $\mathbb{R}^3$. Then $\alpha \wedge \omega = 0$, but $\alpha$ can never be written $\omega \wedge \beta$ for any $\beta$, since such a form will necessarily vanish anywhere $x=0$.

Edit: changed the example slightly

$\endgroup$
1
  • $\begingroup$ So all those second terms must have zero coefficients because the $w$ wedged with the $v_i$ form a basis for $\bigwedge^{k+1}V$? Then we have the result from looking at the first term. Great, thanks! $\endgroup$
    – Wooster
    Commented May 13, 2016 at 21:23

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .