1
$\begingroup$

So we know that Hodge duality works this way

$$⋆(dx^{i_1} \wedge ... \wedge dx^{i_p})= \frac{1}{(n-p)!} \epsilon^{i_1..i_p}_{i_{p+1}..i_n} dx^{i_{p+1} } \wedge dx^{i_n}$$

where $p$ represents the $p$ in $p$-form and $n$ is the dimensional number.

My question is: How does hodge duality work on the imaginary number $i$ on one hand and on partial derivative like $\partial_x\alpha$ let us say on the other where $\alpha$ is a complex function.

$\endgroup$

2 Answers 2

4
$\begingroup$

If you are working on a complex manifold with a Hermitian metric, then the Hodge star operator should be taken antilinear: $\star(\alpha + i\beta) = \star\alpha - i\star\beta$. If you work on a real manifold without a complex structure by itself, and you extend the scalars to the complex numbers, it may be that there is no harm in taking it to be linear.

$\endgroup$
3
$\begingroup$

The Hodge star operation acts on differential forms. Numbers, real or complex, transform as 0-forms. The Hodge dual of a 0-form will result in something proportional to the volume form of the manifold. In detail, for a $d$-dimensional manifold,

$\star 1 = \text{vol}_d = \sqrt{|g|}dx^1 \wedge ... \wedge dx^d$,

and the Hodge operation commutes with multiplication by complex numbers.

A partial derivative like $\partial_x \alpha$ is also a complex number (in general), so the above applies. However, given a scalar function such as $\alpha$, a 1-form can be obtained as

$d\alpha = (\partial_i \alpha) dx^i$.

Then the Hodge star acts on $d\alpha$ as it would on any 1-form.

$\endgroup$
6
  • $\begingroup$ Thanks for your reply. I have some questions on what you said. Do you mean that $\star i = \text{vol}_d = i\sqrt{|g|}dx^1 \wedge ... \wedge dx^d$, or just the same as the one you wrote, $\star 1 = \text{vol}_d = \sqrt{|g|}dx^1 \wedge ... \wedge dx^d$? Second I can't see how the partial derivative is also a complex number? May you please elaborate on this one? Third if it happens that $\star{(i\partial _x \alpha)}$ what will happen in this case? $\endgroup$ Commented Feb 1, 2015 at 15:31
  • 1
    $\begingroup$ So $\star c$, where $c$ is ANY complex number, is simply $c \star 1$, where $\star 1$ is defined above. The complex derivative is a number, either real or complex valued. Perhaps I should have emphasized this, but real vs. complex is a completely irrelevant issue for Hodge duality--the only thing that matters is the tensor structure of the object you are dualising. So $\partial_x \alpha$ is just some number, call it $c$, and everything I said before goes through. $\endgroup$ Commented Feb 1, 2015 at 18:13
  • 1
    $\begingroup$ By the way, there are many good references on Hodge duality. I learned from this book: amazon.com/Geometry-Topology-Physics-Edition-Graduate/dp/… $\endgroup$ Commented Feb 1, 2015 at 18:14
  • 1
    $\begingroup$ Well the determinant is secretly present in your expression above, it's hidden in the epsilon tensor. In curved space there is a difference between the usual Levi-Civita symbol which is a tensor density, and the symbol which appears in your formula which is a tensor. Any book which discusses Hodge duality should also cover the differences between the two. $\endgroup$ Commented Feb 1, 2015 at 18:54
  • 1
    $\begingroup$ Happy to help!! $\endgroup$ Commented Feb 1, 2015 at 18:57

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.