# The Unapologetic Mathematician

## The Determinant

Let’s look at the dimensions of antisymmetric tensor spaces. We worked out that if $V$ has dimension $d$, then the space of antisymmetric tensors with $n$ tensorands has dimension

$\displaystyle\dim\left(A^n(V)\right)=\binom{d}{n}=\frac{d!}{n!(d-n)!}$

One thing should leap out about this: if $n$ is greater than $d$, then the dimension formula breaks down. This is connected with the fact that at that point we can’t pick any $n$-tuples without repetition from $d$ basis vectors.

So what happens right before everything breaks down? If $n=d$, then we find

$\displaystyle\dim\left(A^d(V)\right)=\binom{d}{d}=\frac{d!}{d!(d-d)!}=\frac{d!}{d!}=1$

There’s only one independent antisymmetric tensor of this type, and so we have a one-dimensional vector space. But remember that this isn’t just a vector space. The tensor power $V^{\otimes d}$ is both a representation of $\mathrm{GL}(V)$ and a representation of $S_d$, which actions commute with each other. Our antisymmetric tensors are the image of a certain action from the symmetric group, which is an intertwiner of the $\mathrm{GL}(V)$ action. Thus we have a one-dimensional representation of $\mathrm{GL}(V)$, which we call the determinant representation.

I want to pause here and point out something that’s extremely important. We’ve mentioned a basis for $V$ in the process of calculating the dimension of this space, but the space itself was defined without reference to such a basis. Similarly, the representation of any element of $\mathrm{GL}(V)$ is defined completely without reference to any basis of $V$. It needs only the abstract vector space itself to be defined. Calculating the determinant of a linear transformation, though, is a different story. We’ll use a basis to calculate it, but as we’ve just said the particular choice of a basis won’t matter in the slightest to the answer we get. We’d get the same answer no matter what basis we chose.

December 31, 2008