One Thing Leads To Another

I’ve always loved comic strips. I even blog about them sporadically over at Daily Comics Review with a few other fans and artists. Of course comics strip often involve dogs and training, but sometimes the most appropriate training comics don’t even have dogs in them.

The strip above demonstrates classical conditioning: Janis associates baby powder with playing with her baby, while Arlo associates it with a distastful chore.

This is exactly how classical conditioning works, a stimulus — in this case an odor — evokes a emotional memory. This is powerful stuff! The physical mechanisms that classical conditioning utilize are different from other kinds of learning. The associations are very durable – they take a long time to be forgotten, if ever. The associations also seem to involve parts of the brain involved with emotion, although this is still under study. But we have all experienced a smell or a small snippet of music that years later can bring back a strong emotional response.

As a matter of fact, behavior people often refer to a conditioned emotional response (CER.) Not just because "pavlovian response" has fewer syllables and doesn’t sound as cool, but because that is exactly what is happening. Janis is happy – she is thinking of a beautiful, special, and intimate time with her son. (He’s in his 20s now, and almost done with college.) Arlo is disgusted, thinking of dirty diapers.

While which areas of the brain are involved may still be unresolved, we do know that conditioned emotional responses are not strengthened or weakened that way “operant” behaviors are: they can not be reinforced or punished. Comforting someone when they are afraid doesn’t make them more afraid. Punching someone for being angry doesn’t make them less angry.

Conditioned emotional responses can go extinct, in other words, the association can be lost. But this takes time — how long really depends on the individual and the situation — and usually requires that the situation be avoided completely for a long time. It’s also possible for the association to spontaneously recover – come back even after it appeared to be extinct. Pavlov encountered this back with his famous dogs, when he did the first intense research into this kind of learning.

Rather than attempting extinction, we usually work to change the association. Pair the stimulus that triggers the response, starting with low intensity and gradually increasing it, with something pleasant.

So in Arlo’s case, we could pair increasingly large concentrations of baby powder with mugs of beer, until he started to associate the baby powder with something more pleasant than dirty diapers.

What do you think? What good and bad associations do you have?