Originally Posted by
Adam Palmer
The negative punishment comes into it because the dog doesn't get rewarded when not sitting. That's the basis of the training. Can be very stressful for a highly motivated dog.
If the reward is withheld or withdrawn, that is -P Adam, but you appeared to suggest that pos. reinforcement and -P literally go together, as +P and -R do, and that isn't really correct, because there is no
automatic relationship, in the same way as +p and -R.
It's often the case that pos. reinforcement is used and there is no neg. punishment at all, during training. IF the reward is never withheld, but always given (ie the dog is set up for success, which is the aim of positive training) there is no neg. punishment involved in the training.
(As previously stated, reward based trainers sometimes do use negative punishment, but it's usually set up carefully and is meant to instruct and help the dog to understand what is acceptable behaviour. They may also use extinction which can cause some frustration, but most dogs learn easily how to work/think for the reward, and offer different behaviours.Clicker training is a subset of OC and may use pos. reinforcement, extinction, neg. punishment and ofcourse the event marker to modify behaviour).
You rigntly place a lot of emphasis on Skinner and operant learning
. I guess then you have heard of Bob Bailey? BB worked with Skinner and learnt what he knew from him, and the Brelands, who did work very closely with Skinner, and between them they trained a
huge number of species of animals for many things, from war work to shows and displays. I've had an email conversation with him on this exact subject
BB: "Here is my take for those who spend an inordinate amount of time trying to parse the strict interpretation of the "quadrant" view of behavior:
Time spent on the endless discussions of into what quadrant an event or stimulus falls might better be spent on training the animal." Have to admit that made me laugh - he's quite right!
"Behavior of animals, especially of the "higher" animals is complex and dynamic. The neurologic processes underlying behavior are complex and dynamic. Simplifying behavior by division into actions and consequences can be useful and can further communication between students, practitioners, and scientists. Failing to recognize that these divisions of actions and consequences are simplifications can lead to hindering of communication and even misunderstanding between students, practitioners, and scientists.
The discussions of the thinking that goes on behind what an animal knows or does not know takes it out of the realm of OC, simply because those (Skinner,
et al) that study OC say it does. OC has to do with measurable behavioral events. In the future, when we can measure thinking, and what an animal is thinking, then OC may indeed study what is in the mind of the animal. Right now, we can't do that.
According to OC what we know of what the animal knows we discover by measuring behavior. This is one of the (more or less legitimate) arguments some use that OC is limiting what we can know about animals, and humans. Before condemning OC too much, the cognitivists have their problems too, creating constructs and processes that really don't explain how and why behavior is the way it is. In addition, the closer the cognitivists get to really dealing in the changing of behavior, the more it looks like OC."
What he is basically saying, is that we cannot measure if an animal is experiencing neg. reinforcement
unless the measurable behaviour itself is affected. This is the scientific way. I would agree with this. You can't say that a dog is experiencing neg. reinforcement if the owner has sausages in pocket and is just wandering around with them, and the dog knows the sausages are there, because that would be trying to mind read. I've seen shock collar people pretend that this is the case, and that the dog is experiencing -p, but I don't agree with that as it's not classic OC and it's not scientific. Perhaps in the future, as BB said, things may change, when/if we ever get to read minds.
In science, it's important to make up one's own mind based on the evidence, and not to rely on the interpretation of others, who may want to tweak it a little here and there for their own ends. If you use OC Adam, be true to the principles and don't be led astray by those who have a wee agenda, whatever your own methods.
Wys
x