I know clicker training have their horses "pose" to prevent nippiness and it's great but it doesn't work on all horses, this one is just too eager. Without natural hrosemanship to go with it I wouldn't be able to use him as a school horse.
I too agree, as I've said, that for me at least, using negative reinforcement as well is best, although there are those who do claim to use solely positive reinforcement successfully(any here, care to share?)
I don't personally go along with the strict behaviourist model - I think there's far more to it than straight input=output - but I do think the 'laws of learning' hold true consistently. I almost never say never, but I think the principles
behind c/t do work all the time.
Eg. If the horse NEVER got reinforced for 'muggin' behaviour, he'd eventually stop trying, because the behaviour doesn't work. Especially if he also got reinforced for a conflicting behaviour, such as tucking his nose in & down or taking a step backwards.
If the horse occasionally gets reinforced for the 'mugging' though, this will actually strengthen the behaviour (another reason it's good to drop back on reinforcements once the desireable behaviour is learned). The stronger a behaviour has been, the longer it takes to eliminate it - eg. If your horse knows it doesn't work all the time but it's worth a try, if you decide to become consistent & NEVER allow that behaviour to work for him, he'll likely keep trying for some time before ultimately giving up. It reminds me of the difference of experiences driving cars; if you jump in a shiny new car and turn the key, if it won't turn over, you're only likely to have a couple of goes before giving up, but if you jump into your old clunker, with a history of reliability... eventually, you might sit there trying to start it for ages before giving up.