Behavior Analysis: Is That All There Is?

We’ve read articles on several blogs recently that state applied behavior analysis is the basis for animal training. Here’s just one example:

“Animal training, when performed by professionals who have a background in behavior and learning is based on the science of Applied Behavior Analysis” (D. Jacobs, www.fearfuldogs.wordpress.com)

Fact is, animal training is based on much more than applied behavior analysis.  The “science of Applied Behavior Analysis” is a distinctly different field of study than the field of ethology and the study of animal behavior.

Simply put, behavior analysis is the application of operant conditioning to real world behavior issues in people and animals.  There’s no doubt that operant conditioning, which is the focus of applied behavior analysis, is an extremely important tool in training and behavior modification.  But animal behavior is influenced by much more than reinforcement and punishment delivered by the trainer/handler/owner.  Ignoring those other influences also ignores the essence of who animals are and can result in a failure to appreciate each animal as the individual she is. It doesn’t take into account the species typical behavior of the animal nor does it consider the behavioral needs of different species or individuals. Ultimately, it can lead to training failures.

That isn’t just our opinion.  Realization that animals don’t always respond as they’ve been conditioned/trained to do should come as no surprise to anyone.  And one of the reasons for that is because of the inherent tendencies, propensities, or preparedness animals have to behavior in certain ways.  In other words their species typical behaviors must be considered.  And that knowledge comes from ethology, not from behavior analysis.

Every trainer should read and take to heart the classic paper “The Misbehavior of Organisms” by Breland and Breland.  

The paper describes a number of examples in which a variety of species of animals, from raccoons to pigs, had been operantly conditioned to perform certain behavior sequences. The authors describe how the animals they had trained simply failed to behave in the way they had been conditioned do to.  They describe this as “a clear and utter failure of conditioning theory”.

And why did their operant procedures fail?  Because of the influence of species typical behaviors and behavior propensities;  what they labeled the animals’ “natural instincts” (this paper was written in 1961 when the term instinct was still commonly used by ethologists).

One example is pigs (who the authors describe as one of the most trainable species they work with) that had been trained to pick up a wooden coin, carry it about 6 feet, and deposit it in a piggy bank (a box).  Over time, the pigs instead would drop the coin, push it around on the floor, pick it up again, toss it in the air, push it around again and fail to deposit it in the box, despite consistent high value reinforcement if they did so.

It’s clear that the pigs were engaging in the species typical behavior of rooting, a pattern that is normally part of finding and consuming food.  Performing the rooting behavior was clearly more important to the pigs than the external reinforcement being given.  Knowing that requires an understanding of the natural behavior of pigs.  Read the paper for more examples.

It’s also well established that conditioned emotional responses are strong influencers of behavior.  While the field may more recently be broadening its boundaries, behavior analysis and the experimental analysis of behavior – with their focus on operant conditioning – have traditionally pretty much ignored the influences of classical conditioning on behavior.

It should be obvious that if one is going to train dogs, or any other species – one should know about the normal behaviors of the species being worked with.  It should go without saying that as Breland and Breland found out more than 50 years ago, the communication signals and systems, social behaviors, food gathering behaviors, agonistic behaviors, comfort seeking behaviors, courtship and sex and care of the young, are all crucial to understanding the species we work with and to successful training.

How often do we complain about people not recognizing signs of fear, stress, and anxiety in dogs and cats?  Knowledge about body postures and all the other behaviors listed above doesn’t come from the field of behavior analysis – it comes from ethology.

How often are behavior problems caused in part because the animal’s needs are not being met?  A cat is not using the litterbox because there are an insufficient number of litterboxes containing litter that is too deep, not clean or too coarse of a texture.  Rewards and punishment are not very effective at getting cats to use litterboxes if the characteristics of the box don’t take into account the normal elimination behavior and preferences of cats.

What about dogs in the same family who aren’t getting along because one dog is continuing to threaten the other even in the face of submissive and avoidance behaviors from the victim?  Understanding how social relationships among dogs work is going to be a crucial part of solving the problem.

We can’t think of a field that needs to be more interdisciplinary than that of training and working with behavior problems in companion animals.  Is a behavior analytical approach helpful and valuable?  Of course.  But not at the expense of ignoring the ethology and behavioral biology of the species.  That’s like saying all veterinarians need to know is what medication to give or what surgery to perform without any consideration of the physiology and anatomy AND behavior of the animals they’re treating.

If you want more of an interdisciplinary approach in your work with pet behavior and training issues, we think you’ll find a home with our Behavior Education Network community. Visit Members.BehaviorEducationNetwork.com to see what our members have been learning in recent months, and consider joining us.

Pig photo courtesy of JohnHarveyPhoto.com