Today’s Birthday honoree, BF Skinner, outlined the four quadrants of operant conditioning that dictate so much about how we, and dogs, behave. The better we are at understanding how our own behaviors have consequences that fall into one of these quadrants, the better we can figure out how to influence our pets’ future behaviors by controlling the consequences of their actions.
As we discussed earlier today, in Skinner’s vocabulary “positive” means adding something, “negative” means taking something away, “punishment” means lessening the chance that a behavior will happen again, and “reinforcement” means increasing the chance it’ll happen again. Here are the four quadrants of punishment/ reinforcement (aka operant conditioning) as explained through some of my favorite Little Golden books:
Positive reinforcement: The behavior (mixing and making colors by splashing) is reinforced because it results in pretty colors. Most humans, and kittens, would agree that pretty colors are rewarding. Chances that they’ll repeat this behavior are quite good.
The kids make music, which is fun–music is an addition to their lives, and therefore positively reinforcing. This is actually a self-reinforcing behavior: It feels good and sounds good (to them, at least), so they keep doing it. The fact that making music is a reward onto itself is why we live in a world of karaoke machines and garage bands. In dog world, barking would be an example of a self-reinforcing behavior.
Laddie is being scolded–mildly, but still. The reprimanding is the thing being added to this situation that makes it a positive punishment. Bruce’s tongue lashing will decrease the chance that, when next given the opportunity, Laddie will again stare at him lovingly and refrain from telling him his fly his open.
In her future adventures, Goldilocks’ behavior (lying in a stranger’s bed) is less likely to occur again because it has been met with a positive punishment: A bear leaning over her. On page two of our text, running is negatively reinforced: The behavior is encouraged because its result is the removal of the bad thing (baby bear).
This might look like an innocent scene, but the Poky Little Puppy is being threatened with negative punishment: Something good will be taken away (dessert) if a behavior occurs (digging). This will decrease the likelihood of future digging. Too bad his reading level is not exactly up to the challenge of a double negative command.
When the reindeer aren’t running fast enough, Santa shouts at them. If the behavior, flying faster, stops the annoying thing (getting called out by name–especially when your name is “Prancer”), than the reindeer are being negatively reinforced: Something unpleasant getting taken away in order to encourage a desired behavior (flying fast) This is how prong collars and choke chains work on dogs: The behavior of not pulling on the chain is reinforced because that’s what prevents the pain that otherwise would come from the prongs jabbing the dog in the neck, or the chain squeezing around the neck.
An unpleasant thing (getting shivery cold) is removed, but this is contingent on a behavior (hanging out with a magical snowman). The behavior is being negatively reinforced.
The behavior (playing) is being encouraged because partaking in the activity is the best way to avoid the unwanted removal of something (Alice’s head). Because she’s being negatively reinforced, next time Alice will be more likely to play. But only if she gets to keep her head. Otherwise, there won’t be much behavior to train at all.