3 ms·
I wish I had an elementary conceptual explanation for why this association exists. It's a simple calculation, but hard to describe without resorting to equatio
by auntienomen 9y ago
I wish I had an elementary conceptual explanation for why this association exists. It's a simple calculation, but hard to describe without resorting to equations. Maybe the best I can say is that when one writes down the energy function for a quantum mechanical system of many particles, it has a form which is _strongly_ constrained by special relativity. The variables which describe the creation of particles are perfectly paired with the variables which describe the annihilation of particles. For bosons, these variables commute, just like ordinary functions. For fermions, these variables anti-commute, picking up a sign when you exchange them. If you repeatedly exchange these variables to rewrite energy function as vacuum energy + energy of 1 particle + energy of 2 particles + ...., you discover that the fermions contribute negatively to the vacuum energy while bosons contribute positively.
- klank 9y agoThanks. This is what I was looking for when I originally posed the question. So, if I understood you correctly, the restriction of Lorentz symmetry on the Hamiltonian in turn requires QFT creation and annihilation operators to commute/anticommute (the anticommute part what then ultimately drives a negative energy contribution from fermionic fields when summing over all possible energy states). What causes the anticommute properties? Is there a non-abelian symmetry group backing fermion fields somewhere?
- auntienomen 9y agoThat's about right. There's a tiny bit of confusion The anticommutation of fermions is a consequence (the content, really) of the spin-statistics theorem, which is a consequence of Lorentz invariance & unitarity.