by Andrew Perkins
Best Friends Training
In a week of dog walks, I probably see two or three different people trying to get their hyper-aroused dogs to hold a stay or a “watch-me” while an unfamiliar dog is walked past on the sidewalk. In some cases the dog holds the stay, and in others there is lunging and barking as the passing dog gets close. Even among those who hold the stay, most remain tense and white-knuckle their way through.
As a private trainer in the suburbs, clients struggling to manage their dogs’ reactivity on walks probably make up about 60% of my caseload, and I commonly hear that a previous trainer had suggested this same tactic to them.
In a world where clicker training and the positive reinforcement movement have helped a generation of dog trainers understand and apply the principles of operant conditioning like never before, I can’t help noticing among my peers and the dog-owning public a decline in the understanding and application of its counterpart, associative – or classical – conditioning.
Maybe your experience is different, and by reading and hearing some of the writers, thinkers and presenters on the subject, you’re among a more multi-faceted crowd. I can only judge from what I witness, and from my own observation, I thought a better understanding of employing associative conditioning in behaviour work was worth exploring in FORUM.
At its simplest, as we might explain it to our clients, operant conditioning is where the learner comes to act (or avoid acting) in order to get or avoid outcomes. Lying down makes a treat happen, turning left avoids bumping into a wall. Again at the client-explanation level, associative conditioning happens when the learner comes to feel about one thing the same way he or she feels about another thing, because of the association between the two. Doorbells, which have no intrinsic emotional value, become exciting because they predict arrivals, which are exciting. Usually the behaviour that results is more reflex than conscious act – in fact, once the resultant behaviour becomes conscious, you’ve probably crossed over the line to operant conditioning (perhaps reflex barking due to excitement made the intruder go away, so eventually barking becomes deliberate if the intruder’s departure was reinforcing).
In addressing problematic behaviour that is reflexive (or reflexive at its root), such as resource guarding, reactivity to other dogs, separation anxiety, or any of a variety of fearful behaviours, associative conditioning is gaining ground as a starting point, but I see plenty of evidence that it hasn’t penetrated public understanding or popular culture.
The operant approach I described above, where dogs are made to hold a stay despite the encroachment into their space of an unfamiliar dog, probably comes from a competitive obedience history, where performance of a skill was subjected to the pressure of distractions. While valuable in its own right, it doesn’t address the pet-owner’s basic objective; holding a stay isn’t actually their goal – walking calmly by is. In this scenario the stay is a tool for management that likely won’t reduce the source issue of arousal, because the association doesn’t change – the sight of an unfamiliar dog still results in its approach and eventual close proximity.
And that near-miss, a failure to address the emotional state that contributes to the behaviour we are trying to alter, happens with a wide range of other behaviour problems when they are approached from an operant-first perspective.
So let’s back up. With a basic understanding of how the brain works, we can devise an associative conditioning approach that addresses the real problem.
What the Brain Is Doing
It’s probably terribly reductive, but since I’m not a neuro-anything, the following description of what happens in the brain is how it was simplified for me by people with better neuro-stuff education.
The brain has a network of neurons that are either connected to other neurons or looking for a connection to transmit signals. When one neuron is stimulated, for example by the sight of another dog, it finds a connection with another neuron, perhaps one that produces arousal. Now when that first neuron is stimulated again (sight of another dog), it already has a bridge, called a synapse, to the neuron that starts the arousal, and each time the bridge is travelled, the neuron becomes more conductive by growing more conductive coating (myelin). Thus the bridge becomes easier and faster to travel next time.
The trick of associative conditioning is to tickle the first neuron so gently that the impulse doesn’t immediately start down its usual bridge, then provide a new outcome calculated to form a new bridge to a different neuron that creates a different emotional result (again: reductive but sufficiently true). Then that process is repeated until the new neural pathway is more conductive than the old one, and the level of stimulation can be gradually increased, continuing with the new associated outcome, until at full stimulation it has replaced the old neural path, which withers from disuse.
In human psychology, this exercise is often called exposure therapy. In dog training, it mostly gets referred to as counterconditioning. I say “often” and “mostly” because both words are used in both environments.
Control the Intensity of the Exposure
How we achieve the trick of the gentle tickle is our first challenge in applying the therapy. With many behaviour triggers, proximity is the simplest controllable factor. You see this in Grisha Stewart’s Behavior Adjustment Training (BAT), where proximity is managed at first, and the subject dog gets control over movements towards and away from fearful or arousing stimuli. With sound sensitivity, volume is the managed factor. With other triggers, it may be about breaking apart the components of the stimulus to create new associations with each – so a dog who barks at animals on TV might be presented with sounds without video at increments, while in separate exercises the dog might be shown video without sound, perhaps starting with Disney animated animals, or from a position farther from the TV. With separation anxiety and for an assortment of aroused behaviours, duration of the trigger is a factor that can be controlled. In many circumstances you’ll combine management of volume, duration, proximity, size, and who knows what other factors, all in increments that start below a behaviour-triggering threshold.
A New Conditioned Response
The second piece of the therapy is eliciting the new emotional response. Food is the easiest consequence to control in producing a desirable emotional response, but other controlled outcomes are valid, provided they elicit an emotion that serves the need. For example with a reactive dog, turning and/or walking away after brief exposures, with food delivered upon retreat, both changes the expectations and limits the duration of the exposures. Emma Parsons’ “Click to Calm” and Grisha Stewart’s “mark & move” techniques are examples. They also exemplify the overlap of associative and operant conditioning, since the new emotion triggers a new behavioural response – look for food – that is reinforced.
Building Tolerance
The third component of counterconditioning is the repetitive and incrementally increasing stimulation. Your clients need strict guidance on this part, because ambition can be their enemy. It helps if they understand why they’re doing what you’ve assigned. You can explain the whole neurons and synapses thing, which I find helps put owners in a clinical mood and allows them to feel less emotional about a process that can otherwise be so stressful they may avoid the work. Clients also need to understand that once therapy begins, any exposures above the point you’ve achieved in controlled exercises can result in a full-blown response of the old, unwanted sort, strengthening that old synapse and undermining or even undoing their progress.
The Trainer As Coach
Prescribing the steps, dictating the number of successful repetitions required to advance an increment of exposure, and providing detailed descriptions of the body language hints that are the earliest tip-offs to the desired and undesired emotional responses are the bare minimum. For bigger projects, it helps to provide a workbook for documenting exercises and observations, so that the progression is easily organized and analyzed. And as much as possible, your presence as a coach is the best-case offering. You’ll be better than your clients at noticing body language cues, more intuitive at changing gears when it’s called for, and able to stop proceedings when a B-plan should be concocted.
It starts with your mindset – your paradigm, if you like. Analyzing behaviour from an operant perspective is important, but by considering the dog’s emotional state and reflexive behaviour where you can also affect change, you might accelerate progress towards your operant objectives, and in some cases even eliminate the need for operant work in addressing a behaviour problem.
Andrew Perkins is a founding member and Life Member of CAPDT. He is an instructor at Dealing With Dogs in Oakville, Ontario, and offers private behaviour consultation as Best Friends Training (www.BestFriendsTraining.ca). He recently published a consumer-facing book, Best Friends: How to Make it Easy to Love Living With Your Pet Dog.
