I've enjoyed reading this thread after giving this forum a miss for several months.
Using a treat (+R) to train a correct canter lead is easy to do (have used it successfully in the past) provided your timing is correct and you can sit a sudden stop. The reason the horse doesn't get confused is because you make the bridge sound (kiss, whistle, click whatever) at the exact moment the horse gets the lead right. Because the horse has already made an association between that bridge cue and the arrival of the food it knows that the sound means it performed a behaviour that will result in food. All the bridge does is buy you some extra time to give the food, which when you are riding is handy because you can't give the food straight away. Despite the skepticism, the horses get it very quickly. The downside is that many will come to a sudden stop when they hear the sound because they want the food. If you aren't prepared it can be a bit of a shock though could make teaching reining stops quite easy I would think.
The hostility to positive reinforcement from horse trainers who frequently remind us of the length of their experience or the number of horses they have trained never ceases to amaze me. Trainers of dolphins, dogs, seals, horses, cats, sheep, cattle, chickens, zebra and now even elephants have been using these techniques, in some cases for close to 50 years. The vast majority of animals used in movies are trained with PR
these days. Yet us horse people are still out there arguing that it dosen't work (despite 1000's of studies in a wide variety of species incl horses which show that it does) or they shouldn't be used because- pick whatever reason you like. Or they say that food rewards spoil horses on the basis that SOME horses trained with PR
ARE pushy so don't use PR
ever, for anything. No matter that horses trained with negative reinforcement buck, shy, rear, bolt and sometimes kill or main their riders because the negative reinforcement hasn't worked. Unless you are very bad with your timing and technique the worst that will happen with a badly trained PR
horse is that they will get pushy and maybe nippy. That is not due to the method but due to the incorrect application of it. It is very easy to train horses with PR
without them getting pushy or nippy.
Because some horses don't stop when the bit is pulled or buck when they feel leg pressure do we think that we should never use NR. No, we work out what's gone wrong and try to change what we are doing or if we think its the horse's fault we buy a bigger bit or spurs- to apply more pressure in case the horse didn't get it the first time. We rarely seem to stop and think what we might have done (or not done) to cause the horse to buck or bolt, stall or drift. Like good NR training, PR
simply requires good timing, consistency and shaping of responses from simple tries to the final polished behaviour.
That said I am not a believer in the notion that using only PR
is good for horses. Some horses can get very frustrated, even with good timing and I don't think a frustrated horse is a happy one. I also believe that there is and always will be a place for negative reinforcement in horse training. In fact, unless we never use any kind of tack we will always be using NR to some degree. I usually use PR
when training trailer loading but I always make sure that the horse will load and back off from pressure cues alone. This is a deal breaker for me as far as safety is concerned- a frightened horse won't be tempted by food so its imperative they understand and respond to pressure cues. The same is true for ridden work- the food often makes it happen more quickly with less "mistakes" but in the end the horse must be trained to stop, turn and go from those pressure cues and their release. Whether I achieve this using PR
, NR or a combo (usually), the fact remains that the only way the horse will learn how to give me the response I want to the cue I apply is if I train it and make the associations clear and consistent. If the horse doesn't comply its because I have stuffed up- whether going too fast, not taking into account environmental stimuli, its sore etc.
As others have posted, the simple rules of reinforcement and punishment developed by Watson, Skinner, Hull and others (all Americans by the way) are not the whole story with all animal training. There is now some very interesting research being conducted into areas such as arousal, affective state and learning (basically emotionality), attachment (whether there is evidence that horses get attached to and respond better to some individuals than others and why), the attributes that make some people successful trainers compared to people who aren't and many more. If vets and the medical profession had the same attitude towards opening their minds to ideas that challenge their beliefs as horse trainers we would not have antibiotics, radiographs or cancer drugs.
There are many trainers who don't use any PR
in their training that are ethical and humane in their methods and produce quiet, responsive horses. I look up to and respect many of them and use some of their techniques myself. But to state that PR
doesn't work, can't work or shouldn't even be tried is simply silly when there is ample evidence it does work and many horse owners find it hugely improves their training and their relationship with their horse. And in the end its those pressure cues, pressure equipment and their (mis)use that leads to the injuries due to the bucking, shying and bolting- not one too many carrots.