> Or another constraint on how the space operates which results in the same result.
I see that would be the orthonormality of i and j.