Duality in exterior algebras

To fix ideas, let V be an n-dimensional real vector space. A k-vector is an element of ⋀kV (the exterior algebra) and a k-form is an element of ⋀kV∗. These spaces are dual to one another but not in a natural way.

Now, the top exterior power of a finite-dimensional vector space is one-dimensional, so both ⋀nV and ⋀nV∗ are of dimension one. We also see (this time by direct computation) that there are nondegenerate bilinear pairings

⋀kV×⋀n−kV→⋀nVand⋀kV∗×⋀n−kV∗→⋀nV∗

for all k.

The problem is that while the top exterior powers are of dimension one, we do not have a canonical isomorphism between them. To get one, we fix an inner product g on V. Its determinant (or volume form) is a nonzero element of ⋀nV∗ and thus gives an isomorphism of that space with the ground field R. We get a compatible isomorphism of ⋀nV with R by considering the dual metric. Composing these isomorphisms with the above nondegenerate pairings induces isomorphisms

⋀kV≅⋀n−kV∗and⋀kV∗≅⋀n−kV.

We finally get the correspondence between k-forms and (n−k)-vectors by following this last non-canonical isomorphism:

⋀kV∗→⋀n−kV.

Keep an eye: technically speaking we only need to fix a volume form on the space V to get these isomorphisms, because such a form will induce the necessary isomorphisms of the top exterior powers with R, which in turn will identify ⋀kV∗ and ⋀n−kV via the (now non-) canonical nondegenerate pairings. Sloppy, awful people sometimes do this silently by fixing a basis of V, which they then either define to be orthonormal or use to define a nonzero element of ⋀nV, which then brings us to this kind of confusion later on.

On the other hand, the inner product let us go further an define the Hodge star operator.