Okay, I linked to this in the escape training thread but now at least one piece of it is bugging me...
Does that ring consistently with you all? I thought that, once a learning objective is taught, that reward didn't have to come everytime but it should probably come variably...or maybe I'm just getting mixed up. Either way, would be interested in comment and citation from you experts. If it's wrong, we should change it on Wikipedia.Drawbacks and limitations to operant conditioning
Skinner's construct of learning did not include what Nobel Prize winning biologist Konrad Lorenz termed "fixed action patterns," or reflexive, impulsive, or instinctive behaviors. These behaviors were said by Skinner and others to exist outside the parameters of operant conditioning.
In dog training, the use of the prey drive, particularly in training working dogs, detection dogs, etc., the stimulation of these fixed action patterns, relative to the dog's predatory instincts, are the key to producing very difficult yet consistent behaviors, and in most cases, do not involve operant, classical, or any other kind of conditioning.
The key to understanding this is that, according to the laws of operant conditioning, any behavior that is consistently rewarded, every single time, will be produced only intermittently and will not be reliable. However, in detection dogs, any correct behavior of indicating a "find," must always be rewarded with a tug toy or a ball throw. This is because the prey drive, once started, follows an inevitable sequence: the search, the eye-stalk, the chase, the grab-bite, the kill-bite. This is why dogs trained for detection work, through the prey drive, only work well if they are always reinforced, every single time they behave correctly, which breaks one of the laws of operant conditioning.
Some trainers are now using the prey drive to train pet dogs and find that they get far better results in the dogs' responses to training than when they only use the principles of operant conditioning, which according to Skinner, and his disciple Keller Breland (who invented clicker training), break down when strong instincts are at play.