
People tend to think that “positive reinforcement” means giving the dog a treat as a reward for doing the right behavior, and this is basically correct. However, in the psychological jargon the word “positive” is not being used to mean “good” but rather “to add or apply” – like a plus sign in mathematics. The popular understanding of “negative reinforcement” is doing something that the dog dislikes – something aversive. But in the psychological jargon “negative” is not being used to mean “bad” but rather means “to take away or withhold” – like a minus sign.
These two terms are part of a group of four which are used by psychologists to discuss different aspects of learning theory. The confusion really becomes apparent when one hears the next two terms, namely “negative punishment” and “positive punishment” To the average person, this sounds absolutely ridiculous – how can a punishment possibly be positive ?! And doesn’t “negative punishment” sound like something truly dreadful?
To understand what these terms are actually supposed to mean, let us start at the beginning, with an explanation of each word. I will then put each term into a chart, and include an example of the concept as used in training, a Key Word (something to help jog your memory regarding the dog’s response) and a suggested alternative term that I consider to be more sensible and easily understood by all.
POSITIVE REINFORCEMENT To add or apply something good which strengthens the target behavior. Example: Giving the dog a treat when he sits will strengthen the behavior of sitting. Key Word: “Yum!” Suggested Alternative Term: Reward | |
POSITIVE PUNISHMENT To apply something aversive to eliminate the target behavior. Example: Squirting the dog with a water squirt bottle when he jumps on you will eliminate the behavior of jumping on you. Key Word: “Yuck!” Suggested Alternative Term: Punishment | |
Are you by now even more confused than you were at the beginning? When one tries to understand these terms (especially, I think, “negative reinforcement”) within the psychological jargon, one begins to wonder if whoever invented them was deliberately trying to confuse! The actual concepts, however, are clear and simple, and we need words to discuss them which illuminate our ideas, rather than muddying them.
“Positive reinforcement” and “negative punishment” are two sides of the same coin. If the dog sits she gets the treat; if she doesn’t sit she doesn’t get the treat (how much simpler to think of this as “reward” and “withheld reward”). This system of training is referred to as “Positive Reinforcement Training,” and is the ideal way to train.
Every “negative reinforcement” starts with a “positive punishment” being applied. The instant of jerking the choke chain tight is a”positive punishment,” the continued pressure and its eventual removal is a “negative reinforcement.” As you can see, “negative reinforcement” is the most difficult of all to use correctly, and should rarely (if ever) be used in training. This is the basis of old fashioned choke chain training, which consisted almost entirely of “positive punishment” and “negative reinforcement.”
If you are still feeling confused by this murky language, (and even if you aren’t) my advice is this: let’s ditch this unclear, counterproductive jargon and start talking about “reward based training,” and use the terms “reward,” “withheld reward,” “punish,” and, rarely, “continuous aversive” to discuss our training methods!