Review of “7 ways to get behaviour:Finding the optimal technique in animal training“

Read original article here, by Karolina Westlund, PhD

General

I really enjoyed this article and like the style and demeanor of the writer.  Even when I disagree with her (often), I LIKE her.    I have many areas of disagreement and many of agreement.  Details follow.

Many roads to Rome

True that there are ‘many’ (might say ‘infinite’) ways to get to ‘Rome’.  Also true that many trainers act as if there are not, and that they are teaching THE way.

Applications of tools she does not endorse

Does not appear knowledgeable about tools she does not use.  Never tried a prong collar, for example, and takes Grisha Stewart’s word for how they are.

Rigid way to look at tools, techniques, and punishment.  Tools are either desirable or undesirable.

I see it differently.  Tools can be appropriate for particular situations and applications but not desirable for all.  For example, my training is done with verbal explanation, targeting, capture, and modelling, but enforced with punishment if the animal errs in a potentially dangerous way during proofing or real life performance.  

Am I worried about ‘fallout’?  Well….  First, there rarely is fallout.  Second, there is likely to be catastrophic fallout, like the dog’s death, if he does not correct this behavior.  If I have not convinced him during very voluntary, gentle, and encouraging teaching, then at some point, he needs to decide he ‘wants’ to do it the way I request.  Period (for these few health and safety related behaviors).

So, when I teach, all things are optional and voluntary.  I don’t even use any lead or collar except as an emergency back up when outdoors or in public.  Once learned, I classify behaviors two ways:  optional and enforced.

What is ‘enforced’:  all behaviors critical to health and safety.  You DON’T:  bite the vet, the judge, a child, threaten others, and you DO stop, drop it, leave it and wait.  

Recall:  I like it to be strong, but I still want the dog to override a recall if needed.  For example, let’s say I did not realize a door failed and a dog got out.  I see the dog across the road.  The road is clear.  I recall the dog.  The dog sees a car suddenly approaching.  I want him to wait.  I don’t want him to mindlessly, obsessively, cross the road just because I said to.  Anyway, as long as he will stop, drop it or leave it, and wait, I can go to him to bring him back safely.

EVERYTHING else is voluntary.  And, it all normally works this way:  I get what I train for.  If I am currently in a strong training relationship with an animal, I get close to 95% compliance, or better.  When I come home from a three week vacation, I get the ‘look’ (“You never write, you never call, and now you want me to work!).  It normally takes about three days to ease back into a strong training relationship.  Any more than that, and I start looking for medical reasons.  All things being equal, we all love to learn and to be amongst friends.

Does a learning animal ever get physically punished by the trainer?  If the dogs starts a potentially dangerous behavior in a real life situation that he has not yet been trained for, then I am likely to punish that behavior if I deem it in the best interests of general health and safety.  If a dog tries to lunge into traffic to chase a car, and he has not fully learned about how to act in this situation, that behavior is likely to be punished.  If the dog threatens anyone else, or tries to instigate trouble with another dog or person, or jump on them:  punishment is likely.  In those cases, the need to responsibly manage an irresponsible dog in public outweighs teaching preferences.  Note:  these situations RARELY happen in the training of exotic animals.  In my experience with exotics, they are born into a well developed training program and priorities like health and safety, are almost the first things taught.  The animals are not taken into situations they have not been taught and proofed in.  However, with pets, people are with them 24/7 and pet animals are much more likely to find themselves in situations they are not equipped to cope, from an education perspective.  But they still must cope safely.  So for example, young rowdy dog develops a medical condition before he is fully trained in vet compliance.  Still cannot let him bite the vet, the dogs in the waiting room, etc.

So, in my experience, punishment is a tool of enforcement, not teaching.  I am aware that this is not true of all trainers or all training systems.  But it can be true.

Ways to ‘get’ behaviors

I love the explanations and photo essay, however I was concerned by the condition of her model, species of which was undeterminable.  Cryptozoology experts:  pig, cow, dog?  Let’s say, holstein pig.  Anyway, as benevolent as Karolina’s training perspectives are, her training assistant looks profoundly depressed!  🙂  Go see the article to share my laugh.  I think this part of the article is brilliant, clear, and she is inspiring me to use the tool of automatically sequenced videos to teach.  Brilliant.

Pressure and Release

Karolina is apparently not aware of current best practices amongst trainers who use this tool, and in particular, the work of Chad Mackin.  Chad is incredibly gentle and subtle in the use of pressure and release, and he teaches, internationally, to an enthusiastic following of trainers with a similar mind set.  Here is an example of Chad’s work:  

The article is missing discussion of yet another elephant in the room:  why don’t we give more useful information to animals during training???  If we just tell them, up front, what we want, they are often willing to simply comply.  Or, they often perform in a new, untrained situation, as if they were already trained for that situation.  Read more discussion of this here and here.

This, the practice of working cognitively with animals, is where I now spend most of my time, having exhausted my patience with simple operant conditioning over a career spanning over 40 years as a professional trainer working for zoos, aquariums, private clients, and now, as a trainer of trainers.

Also missing:  Discussion of the importance of stress-coping skills in the preparation of all animals for safe, benevolent training and optimal learning and compliance.  A calm, relaxed, coping dog is a different animal than a hyperaroused dog trying to use any and all environmental stimuli in order to ramp himself up.  Chad does talk about this, here.  I work extensively teaching trainers how to solve these issues, via “Perception Modification,”  A protocol that I developed over years as an exotic animal trainer, where animals can die on the spot from hyperarousal, and coping skills are essential assets.  I have been teaching this since 1990, formulated it in 2000, published various articles from then on, and am still working on the manual.  This is the most difficult subject I have ever tried to teach.  It is also the most important, and the most powerful.  Suffice it to say, in my estimation, it is unethical to teach an animal general behaviors before teaching him how to cope with stresses and manage his emotions.  In real life, most behavioral issues stem from stress and coping failures,  not from lack of learning behaviors.  In other words, ‘sit’ is inconsequential in comparison to ‘remain calm and self possessed and responsible in public’.  Talk about the elephant in the room!

Analysis of training techniques

Karolina gives a neat little table with different colors of happy faces.  I love that table, except… it is not accurate in my experience.  For example, targeting is rated as less precise than shaping.  What???  And also as having less resistance to extinction.  I don’t agree, but then I don’t want rote responses, so I don’t necessarily want something to be resistant to extinction.  I want it to be rigorous in spite of environmental challenge.  There is a difference.

In short, the table might be accurate if you don’t use the Intermediate Bridge, Name & Explain, and Conditioned Relaxation – nawwwww, targeting is more precise than shaping, no matter how you slice it.  So, once again, Karolina writes well in her area of experience, but extrapolates conclusions into areas where she is apparently not experienced.  She is not considering important tools that can modulate the other techniques.

I hope Karolina will explore further, because I will be curious to see how she expands her perspective and explains these other powerful options.  I bet she will be brilliant.  I look forward to that.