I read this piece by Gary Wilkes today and thought it was important enough to share. Most dogs are brought to shelters and then die there because of what they do. Teaching them what not to do is what most owners want and what many dogs need to learn quickly and efficently. I would love nothing more than to throw treats at dogs all day long, but as that old song goes, 'sometimes love ain't enough'.
Adherence to a flawed ideology resembles nothing so much as abject
stupidity…GW
The Assumption of Imagined Harm
In the war over ethical training
techniques there is a boogeyman – imagined harm. Trainers that pander to
exclusively “positive” methods use this boogeyman to suppress logical and open
discussion of the topic. That is because their perspective has no rational
basis and cannot become paramount unless they suppress logical criticism. Their
primary tool is to propose that any use of aversive control is dangerous and
will lead to some imagined harm. That is obviously an irrational statement. A
leash and collar inhibits free movement and compels the dog to hold an
arbitrary distance from the handler. Not only is this not automatically
harmful, all trainers, vets, shelter workers and pet owners use leashes and
collars – even the anti-punishment ideologues. This begs the question of why
someone would propose a wide-sweeping claim that the most casual observation
contradicts. The answer may surprise you. They do this because it allows them
to create the fantasy that they are ethically superior while silencing anyone
who would question their statements. i.e. The anti-punishers promote imagined
harm in order to win the argument.
An
Example: False dichotomies
The tools for creating imagined harm
are often removing any real-world context and creating arbitrary and unrealistic
dichotomies. Here’s an example from the ASPCA’s webs pages on training.
“Some
training methods use punishment, like leash corrections and scolding, to
discourage dogs from doing everything except what you want them to do. Other
methods cut right to the chase and focus on teaching dogs what you do want them
to do. While both tactics can work, the latter is usually the more effective
approach, and it’s also much more enjoyable for you and your dog.”
The first sentence is a prime example
of removal of context and a false dichotomy. They have created a straw man who
uses only punishment, all the time. The two types of punishment they offer are
leash corrections and scolding – relatively mild forms of punishment. If leash
corrections and scolding are somehow harmful then all dogs in the US have been
harmed because they wear collars and inevitably hit the end of a leash, if only
by accident. This leads to several obvious questions. Why would using such
moderate punishments automatically imply that one uses only punishment?
(The stress on the neck from veterinary technicians and shelter workers
attempting to subdue fractious dogs often exceeds anything that a trainer would
do, even our imagined “all punishment” trainer.) Why would leash corrections
and scolding be able to stop every behavior other than acceptable behaviors? If
they are capable of that power and all dogs experience these things (apparently
to no purpose) why wouldn’t a rational person use scolding and leash
corrections as a part of their training protocol? If the unacceptable behaviors
stop, why would they straw man continue to punish the dog?
The second sentence leads us to a
variation on the false premise of the first sentence. This suggests that
teaching the dog what to do is the more direct means of achieving a trained
dog. What is completely missing is an understanding that reinforcement and
punishment have opposite effects. One effect decreases behavior and one
increases behavior. If the goal is to stop the dog from jumping on guests,
positive reinforcement does not “cut to the chase.” It delays the solution by
using the wrong tool. This is only logical. Punishment stops or reduces
behavior immediately and positive reinforcement, by definition, cannot stop
anything. EG: “Cut to the chase” means skipping the extraneous scenes of a
western movie to get to the action – meaning the important part of the movie.
The important part of a behavior
procedure is to get to the solution in a timely fashion. If you delay the
solution by using the wrong tool you leave the animal in jeopardy. Pet owners
do not have forever or always to solve problems like jumping up on kids or
destroying furniture. They need solutions that occur rapidly, safely and
without complex procedures that barely blunt the dog’s behavior. So, no, there
is no logical evidence that teaching dogs what to do is at all more effective.
On the contrary – most dogs die because of what they do. If you can stop the
unacceptable behavior, they live. Whether they have been taught to roll over or
fetch has no bearing in their survival. It’s what they do that kills them, not
what they don’t know how to do. Promoting the concept that we just need to
teach them new behaviors completely ignores the context. As for positive
methods being more enjoyable for you and your dog, consider what it’s like to
constantly have to use treats to bribe an animal into obedience – and still
having it fail routinely when a more powerful motivation intrudes – like a cat
running quickly through the yard. The “positive’ methodology promotes a
process that is pleasant to do, but leaves consequences that are far from
pleasant and may be lethal. That possible result is conveniently removed from
the context.
Exaggeration
– The Number 1 Tool
Another common tool of the
anti-punishment ideologue is to exaggerate wildly and assume that any use of
aversive control causes horrible “side effects.” They never talk
about intended, beneficial primary effects like saving a dog’s life by applying
a controlled, temporary procedure that includes unpleasantness but insures a
long life. For instance, this is an inert, menthol inhaler. It
looks a bit like
a lipstick tube. I recently used one to stop a 90 pound Chesapeake from jumping
and knocking down a four year old little boy. When the dog jumped up, I put the
inhaler to his nose. I repeated it until I couldn’t get him to jump up –
clearly a punishment procedure. Then I gave the inhaler to the little boy – who
chased the dog around unsuccessfully for a couple of days and then gave up the
game. The result was a dog that was cautious about running willy-nilly through
the house or jumping on the child. Ask yourself how this use of punishment
could result in the following “side effects.” Again from the ASPCA…
“Alternatively, you could grab your dog’s leash and jerk her
to the ground every time she jumps up to greet people, and you’d most likely
get the same effect in the end—no more jumping up. But consider the possible
fallout:
- Your
dog might decide that people are scary since she gets hurt whenever she
tries to greet them—and she might try to drive them away by growling or
barking the next time they approach.
- Your
dog might decide that YOU are scary since you hurt her whenever she tries
to greet people.”
As you can see, the author exaggerated and described an
imagined, specific procedure when the topic was supposed to be about the
general behavioral effect called punishment. To make sure that the scenario
would be horrific to the average pet owner, she included the words “jerk”
“hurt” and “scary” to imply pain, damage, fear and suffering. (Again, jerking a
dog by the neck is a standard practice in shelters – including the ASPCA
shelter in Manhattan.) There are several reasons why this is ludicrous and
dishonest. By scientific definition the presentation of a stimulus that causes
a behavior to stop is “positive punishment.” Therefore, by definition, I
plainly punished the Chesapeake. However, nothing I did hurt the dog. The dog
wasn’t even wearing a leash. There was no harmful fallout. The dog was not
frightened by any aspect of the training. The only emotional reaction you could
use would be “caution.” Why doesn’t the SPCA offer a caution about specific
dangers of specific procedures rather than lumping all punishment into the
category of abuse? My use of the Vicks inhaler benefited the dog, the child and
the dog’s owner. How can this use of punishment cause harm? Of course, it
can’t. To get you to obey them, the ASPCA has to scare you. That creates an
ironic hypocrisy – the ASPCA claims that scaring a dog is abusive but scaring people to
force compliance with their ideology is not.
No context, no
analysis of results:
When you read this stuff you will find that there is never a discussion of the
full context of the need for behavioral control. 7-8 out of ten dogs in this
country will not see their first birthday. Shelters see about 20% of the
overall walking-dead and kill 80% of the ones they get. The reason most dogs
are taken to shelters is because they do things that families and individuals cannot
live with. If the behaviors can be stopped, they live, if not, not.
In the example of the dog being jerked
to stop it from jumping, who cares about that if the behavior disappears?
(Before you jump to conclusions, all vet hospitals, shelters and boarding
kennels use “slip lead” collars that constrict the neck when tightened. Poll
any dozen vets and ask if they have ever seen a neck injury they can attribute
to a choke chain or other slip collar. I did about a year ago just to make sure
my information was correct. Of a dozen vets, two ER vets, none of them had ever
treated a dog for a neck injury from a collar of any kind. ) While the
description is loaded with exaggerated dangers it doesn’t tell you the likely
result of the dog doesn’t stop jumping on people – death. If it was proven that
jerking a dog by the neck would prevent it from getting killed, would you
refuse to do it? (The ASPCA pretends that teaching an alternate behavior will
end the jumping, but that is simply nonsense. Teaching you French doesn’t stop
you from speaking English. Meaning, positive reinforcement cannot stop
behaviors. At best it adds to the dog’s repertoire. The old behavior may be
less likely to happen but it isn’t blocked from returning. Research by Ivan
Pavlov confirms reality – old car thieves may go straight, but if they ever
need to steal a car again, they still know how to do it. That means that if you
can get a dog to sit instead of jumping the dog will likely return to jumping
when he stops getting treats for sitting. In almost all cases, that does not
save the dog’s life.)
To retain their pets and have a happy
home, dog owners need to stop unacceptable behaviors once and for all. They
cannot spend a fortune and many hours of each day controlling their dog. Most
of the behaviors that need stopping are innocuous but deadly. Like walking too
close to a rattle snake. The actual behavior is innocuous but the result can be
catastrophic. Regularly knocking down a small child is no different. (Oddly,
anti-punishment people do not oppose using electric shock collars to teach dogs
to avoid rattle snakes but would never countenance using the same collar to
teach a dog to not knock down children – even though the outcome for the dog is
identical but the odds of dying from a rattle snake are miniscule by
comparison.) This selective acceptance of punishment is mindless and
hypocritical…but, then, their world-view is mindless and hypocritical. They
claim to love animals – yet attack methods that could save lives. That harm
isn’t imagined – it’s plain to see and smell at a landfill near you.
