 |
|
|
|
Use of punishment in dog training
Applying operant
training principles
|
Morgan Spector on
positive punishment*
*from Clicker Training for Obedience, Sunshine Books,
Inc. 1999 |
"The need for compulsion arises from trainer error"
Page 42
"I offer for your consideration this simple rule: If at any time
in your training, you must resort to force of any kind to achieve the
behavior you want, you--and not the dog!--are doing something wrong.
Be honest and fair. Take a break, sit down, and figure it out
before starting to work again." Page 42. |
|
|
|
- About positive punishment (P+):
In behavioral terms, positive punishment is something
that is added to an animal's environment (positive) that suppresses that
animal's behavior (punishment). Another way to express this change in
the animal's behavioral repertoire is to say that a behavior is less likely
to occur when this something is introduced to the animal's environment or
perception. Nothing about this definition requires that whatever is added is
considered to be unpleasant or aversive to that animal. However, to
make an animal stop doing what it is doing by adding something usually
requires that the animal wants to avoid whatever that is. In dog
training, positive punishers are typically called "corrections" or
"compulsion" and involve the use of force, discomfort, a startle reaction,
or even pain and fear.
|
- About negative reinforcement (R-):
In behavioral terms, negative reinforcement is
something that is taken away (negative) to strengthen a behavior
(reinforcement), or in other words make that behavior more likely to occur.
In traditional dog training, typically positive punishments are taken away
to strengthen an alternative behavior to the one that the trainer was
suppressing with use of positive punishment. Although we can think of
other ways to apply negative reinforcement in animal training, in dog
training typically P+ and R- are paired.
|
- Alternatives to negative
reinforcement and positive punishment:
There are other ways than P+ to make a behavior diminish or
go way, specifically negative punishment and extinction.
There are other ways to strengthen alternative behaviors
(i.e., other than those you want to go away), namely
positive reinforcement. Actions resulting in negative
punishment, positive reinforcement, and extinction focus on
the parts of the dog's environment that it wants, that it
seeks out, and that it finds pleasurable. These latter
three elements (R+, P-, E) do not involve the use of
unpleasantries, aversives, fear , or pain, from the dog's
perspective.
|
- Which means of suppressing
or strengthening behavior are respectively "better"?
We can define "better" in comparing two methods to reach the
same end (strengthen or suppress behavior) in several ways,
including:
|
- Change in an animal's behavior occurs
more quickly (in less time training)
- Change in an animal's behavior is more
persistent, that is, the new behavior is maintained with
less reinforcement or less punishment.
- The method affects only the targeted
behavior and does not affect other behaviors
- The method does not cause fear, pain or
other suffering
|
By 1999, in fact in the early 1990's, dog trainers such as
Morgan Spector, and many others, began to question dog training
methods that employed positive punishment and negative
reinforcement. In 1993, the Association of Pet Dog
Trainers was founded in large part to bring positive
reinforcement into the forefront of dog training and replace
negative reinforcement and positive punishment. The personal
stories of these trainers are often moving and poignant as they
confessed the moment they realized that they did not want the
relationship that they had with their dogs under the traditional
methods, a moment that came with confessing the harm they
realized they'd done to their dogs. In these moments of
change, so-called "crossover" trainers recognized and regretted
the unfairness of the traditional methods that focussed on what
the dog was doing "wrong" and not what it did "right", and many
admitted the pain and fear that they had caused their dogs in
training.
These personal epiphanies (which many included in their books
and other publications) are all well and good, but are these
sentiments just that, irrational feelings of anthropomorphic
projection? What did a more objective, impersonal science
have to say about use of positive punishment and negative
reinforcement compared to positive reinforcement training? Do
dogs feel pain? Do they learn behaviors faster with R+
than R-? Are behaviors strengthened with positive
reinforcement more reliable than those of negative
reinforcement? |
|
Under construction |
|
|
|
|
|
©
Catherine Toft 2007
|
|