The neurochemical released in our brains from aggression is dopamine. This is why when you finally tell someone off that you've been steaming about, it feels so damn good. But the real reinforcing thing for the use of corrections is that, applied forcefully enough and with the RIGHT timing, they do work...it suppresses or changes the behaviour, we are rewarded by THAT and so it is used again because it worked before. It's learning theory applied to the handler.
This does not of course take into account the long term effects on the dog.
Punishment is part of the learning quadrants because it works. Whether it is appropriate? That is a different ball of wax.
This is also a great indicator of why it is so important through management and prevention to keep the dog (or human) under threshold while teaching a new behaviour or new response to a trigger. Since it is neurochemically self rewarding, keeping a organism under threshold prevents them from practicing (ie learning) the wiring in the brain that cements the process. Neurons that fire together, wire together.