You’ve probably heard it a hundred times – ‘dogs do what works’. I’m going to tweak this just a little and say ‘dogs do what makes sense to them’. If they get a treat for walking to a particular spot (perhaps a doorway) then, of course, it makes sense to walk towards the doorway; it’s an easy way of getting treats, right? Note: we can replace walking toward a doorway with any behaviour – the principle remains the same. But what if the consequence of walking toward the doorway is unpleasant? The behaviour will be ‘punished’ and the dog will not want to walk towards the doorway. Incidentally, one of the problems with using unpleasant consequences (apart from ethics) is that they may have consequences far beyond what we intended; for example, the dog may begin to avoid all doorways rather than the one we intended.
I don’t ever recommend unpleasant training techniques, but we can clearly see that pleasant consequences will make the behaviour more likely, and that unpleasant consequences will make the behaviour less likely. But what if approaching the doorway is associated (in the dog’s mind) with pleasant and unpleasant consequences simultaneously? What if the dog sometimes gets a treat for approaching the door and sometimes gets yelled at, perhaps for moving too slowly, or the door slams in their face? What the dog learns is that doorways are sometimes scary places. Evolution has equipped all animals with the desire to avoid scary things. So now we have a situation where, no matter how hard you try or how many tasty treats you tempt the dog with, they remain hesitant and untrusting of the scary doorway. This is the curse of the poisoned cue. These problems can be overcome, but it may take a very long time for the dog to ‘trust’ that doorway.
We see this over and over in animal training – especially where unpleasant consequences are often used. If we teach behaviours by using pleasant consequences for doing well, and unpleasant consequences for doing not so well, we are literally fighting against ourselves and making training more difficult to achieve. The behaviour will never be fluid, relaxed, or happy, because the cues have been poisoned. Note: cues may be anything which contribute to stimulating the animal to perform any behaviour.
Putting ethics aside, this is my main concern in regard to, so-called, balanced training. The trainers believe they are using all the tools, but I’m concerned that they are actually blocking the greatest tool there has ever been – pleasant consequences.