Proven Strategies

Why We Don’t (or Rarely) Use Punishment

Why We Don’t (or Rarely) Use Punishment

To first understand why we don’t use punishment, we need to understand what punishment is. Behavior is like a binary code. Just as incredibly complex computer programs, like video games or websites, are built from a binary code of 0s and 1s, all behavior, even incredibly complex behavior, is built from a binary code of Reinforcement and Punishment. Every consequence, to every behavior, is either a reinforcing consequence, or a punishing consequence.

Reinforcement = any consequence that causes a behavior to increase.
Punishment = any consequence that causes a behavior to decrease.

These two consequences can be delivered in only one of two ways: they are either delivered Positively or Negatively. Many people hear the words “positive” and “negative” and think that they mean “good” and “bad”, but remember that this is the scientific assessment of behavior, not a moral one. So:

Positive = Added to the environment.
Negative = Removed from the environment.

You may have seen the Consequence Quadrant that is commonly used to explain behavior consequences. It looks like this:

 

Positive

Negative

Reinforcement

Increasing a behavior by adding something to the environment.
(Example: Giving a dog a treat for sitting on cue.)

Increasing a behavior by removing something from the environment. (Also called “Avoidance Learning”. Example: Pushing on a dog’s hips so the dog will sit.)

Punishment

Decreasing a behavior by
adding something to the
environment.
(Example: Yelling at a dog for
barking.)

Decreasing a behavior by removing something from the environment. (Example: Removing a toy that two dogs are fighting over.)

In the course of studying behavior over the past century or so, behavioral scientists have come up with a Consequence Hierarchy. This is the list of the best, most effective, least intrusive method of training down to the least effective, most intrusive method. The hierarchy looks like this:


1st: Positive Reinforcement
2nd: Negative Punishment (only if Positive Reinforcement isn’t available)
3rd: Negative Reinforcement (only if Negative Punishment isn’t available)
4th: Positive Punishment (only if Negative Reinforcement isn’t available)

This hierarchy is why the scientific method of training is frequently called “Positive Reinforcement Training”: it is our first line of response, the one we choose, go to, and use the most. There are indeed times when other consequences may be indicated, but Positive Punishment is the last resort and very rarely necessary. What most people call “punishment” and what we call punishment here is actually “Positive Punishment”.

Another important thing to remember is that only the learner decides what is reinforcing and what is punishing. When we say that all consequences are either punishing or reinforcing, that means that everything we do to attempt to modify behavior is perceived by the learner as either reinforcement or punishment, regardless of what we intend it to be. This means that consequences that we think are just “interrupting” or “distracting” are still a form of punishment if the dog perceives it to be so. Alternately, some dogs may perceive “distractions” as reinforcing. You can determine how the dog perceives your chosen consequence by asking yourself a simple question: after receiving this consequence, did the dog’s behavior continue or increase? Or did the dog’s behavior decrease or cease altogether?

“So what’s so bad about decreasing an undesirable behavior?” Nothing. Through positive reinforcement we can indeed decrease undesirable behavior by increasing a more desirable, incompatible behavior. But there are many reasons that we almost never use Positive Punishment:

  1. Positive punishment creates a behavior vacuum. The fundamental goal of training is to teach the learner what TO do rather than telling the learner what NOT to do. By suppressing an undesired behavior, a dog may learn not to do something you don’t want them to do, but it doesn’t give them any information about what they should be doing instead. In fact, on the exceedingly rare occasions that a positive punishment may be necessary, it should as often as possible be followed by an opportunity for positive reinforcement in order to teach the dog what they should be doing instead.
  2. Positive punishment looks better from the outside than it feels on the inside. The reason that many celebrity animal trainers use punishment methods on TV and in demonstrations is because they look very flashy and appealing: animals are flailing, behaving aggressively, etc. Then, after the trainer punishes the animal, they are quiet and compliant. That looks very impressive to an audience. What isn’t explained is that the animals aren’t quiet and compliant because they are calm and happy; they are frequently exhausted, afraid, and defeated. The realization by a learner that they have no control over their environment and can’t do anything to change it is called “learned helplessness,” and that is what we are seeing at the end of these punishment-based training methods. Learned helplessness creates a variety of mental and emotional problems, which will be discussed later.
  3. Positive punishment does not change the underlying cause for the behavior. Most behaviors arise from an emotional state in an animal. Many undesirable behaviors are caused by fear, anxiety, boredom, or the desire to protect something or someone they love. Suppressing those behaviors will not do anything to change the dog’s underlying emotional state; the dog will continue to live with those emotions.
  4. Positive punishment can cause long-term damage. Even though punishment does frequently appear to have immediate results, it rarely provides a long-term solution. Because the underlying emotional state doesn’t change, the cause of the initial behavior will frequently bubble up to the surface as another, different behavioral symptom sometimes months or even years later. And even if punishment does successfully eradicate a certain behavior, the learned helplessness it causes creates a whole new set of problems. When a human or nonhuman animal experiences learned helplessness, they can become depressed, anxious, aggressive, resentful, and/or shut down. They can lash out at others and/or become self-destructive. They can develop maladaptive coping mechanisms like compulsive repetitive behaviors or superstitious rituals. So in the attempt to correct one problem behavior, a whole host of others can arise. These problems are rarely recognized as being caused by punishment methods. Instead, they are attributed to the learner’s intrinsic nature. People interpret them as being because the learner is “stubborn” or “mean” or “crazy” or “hostile” or “rebellious” or “standoffish”, etc., rather than recognizing them as symptoms of learned helplessness.
  5. Positive punishment frequently teaches the wrong lesson. Since punishment is by its nature unpleasant, a learner will frequently make associations between the unpleasant experience and anyone or anything that happened to be in their environment at the time. So, when people punish a dog for pottying in the house, or getting on furniture, or fencing, or barking/growling, they may learn to not exhibit those behaviors when their trainer is present out of a fear of punishment, but they will not learn to not do those behaviors when their trainer is not around. They just learn how to avoid getting punished by their trainer.
  6. Positive punishment generalizes fear and anxiety. Because the learner frequently learns the wrong lesson when punished, they don’t understand exactly what they are being punished for. This creates a sense of fear and anxiety, because in their minds, they can’t tell when their trainer is going to be “nice” or “mean”, so they tend to live their lives not knowing when, where, or why the next punishment will occur.
  7. Positive punishment can do serious damage to the dog-trainer relationship. As discussed above, learners frequently make an association between the unpleasant experience of being punished and the trainer who is doing the punishing. Think of trust like a bank account: every time you reinforce a dog, it’s like making a deposit to your dog’s trust account, and every time you punish a dog, it’s like making a withdrawal from their account. If you have made lots and lots of deposits, you can get away with making a withdrawal here and there, but if you don’t have a lot of trust in that account and you make even one withdrawal, that can put you in the red, and you will have lost that dog’s trust.
  8. Positive punishment is negatively reinforcing for the trainer. When a dog’s behavior is frustrating to the trainer, positively punishing the dog can feel cathartic. It is, effectively, getting to take out your frustration on the source of your frustration. Punishment also frequently yields immediate results: you yell/clap your hands/snap fingers/spray citronella/whatever, and the dog immediately stops what it is doing. In other words, when a trainer positively punishes the dog, the unpleasant behaviors go away which is a negative reinforcer for the trainer (see the Consequence Quadrant above). The problem with this is that, as we know, negative reinforcement will cause a behavior to increase. That means that the trainer is more likely to select positive punishment methods in the future when faced with a training choice, and in the heat of the moment may reflexively select a positive punisher rather than a positive reinforcer.
  9. Positive punishment requires escalation. Oftentimes a learner can get used to mild punishments that are used regularly, and it no longer becomes punishing for them. This requires the trainer to escalate to a more severe version of the punishment, which the learner will also get used to. The continually increasing escalation of punishment is often how abuse begins. Good people with good intentions trying to train their dogs don’t notice how extreme their punishment methods have become because they arrived there gradually, one baby step at a time.
  10. Positive punishment is easily misused. There are times when positive punishment may be merited, but in order to achieve this in a way that is effective, ethical, and empowering for the learner requires a great deal of skill and knowledge that most people do not possess. If it is not done well, it can do irreparable damage, as described above. Conversely, if positive reinforcement is not done well, it can always be retrained later with minimal damage to the learner’s behavioral, mental, and emotional health. This is another powerful reason to choose positive reinforcement methods over positive punishment.

Here at Best Friends, the ONLY reason to use positive punishment (in the form of citronella, yelling, collar grabs, time out, etc.) is in a true emergency where dogs or people are in danger. If you have any questions about how to handle a specific situation, ask the trainer for your area.

Download the PDF