I push a button and a bell rings. Because of the timing I instantly make an association between the button and the bell. I’ll probably push the button again just to be sure, but when I hear the bell again I’m sure of the connection. And now, if you asked me to make the bell ring, I’d push the button with confidence. I don’t consider that conditioning, I consider that learning, and dogs have no problem learning. But there is a saying that fits what dogs are capable of understanding really well: Genetics loads the gun and environment pulls the trigger. In other words for a dog to understand and respond to something the wiring has to be there.
Brains are wired to connect things we sense to behaviors. In some animals it is as simple as being able to detect light being connected to a flight response. If they are connected to more than one response we have a choice. As connections get more complex we develop understanding. Pointing is a great example. Dogs are wired to understand pointing whereas other animals aren’t, and this ability makes training dogs much easier. Our evolving abstract thinking has taken this to a whole new level and we can create understanding without experiences but with language. Dogs genetics don’t allow them to do that and so for us to create complex behaviors in a dog, we have to construct them from what the dog is wired to understand.
But more than that, we have to understand the rules that govern learning in animals which are different from the rules that we learn to use as a result of our abilities. One of those rules is that dogs associations are very specific, both in what the dog senses and how the dog responds. When you first train a dog something simply changing your location can result in a need for some retraining even though you are doing everything identically. On the other hand asking a dog to change a learned response to a stimulus, like barking when the doorbell rings, requires conditioning techniques. Dogs can’t just relearn something the way we can. Another rule is that sequence matters. As adults we learn almost everything as a sequence in time, step 1 … step2 … step3. This makes little sense to dogs.
Conditioning strategies are necessary to retrain a behavioral response in a dog and therefore they are essential for animal behaviorists. But when I teach my dog properly nothing gets retrained. First I teach the most basic understanding by creating an association to body language, a cue or a prop. Then I may add on to it, or combine it with another behavior the dog knows. When I do this, the dog learns almost instantly and a great deal of progress can be made in a short period of time. Learning through conditioning requires repetition, takes a lot of time, and gives less effective results. So I do not shape a complex behavior in a way that refines how the dog reacts to a cue. I start with ‘be a splitter not a lumper’ and break complex behaviors down to a sequence of simple behaviors that the dog understands and develop the associations. Every cue or prop has one meaning and that meaning never changes. Then sequencing the associations in the proper order becomes critical if I want to train in a way that makes sense to the dog.
Here is an example. To train a tight turn on the ramp, I teach my dog to turn tight first, then I apply turning tight to turning on the ramp. This takes very little time and is very effective. If I train my dog to turn first, then try to tighten the turn, I have to resort to conditioning which requires a lot more effort and the results are often not as good because the dog is more likely to turn wide in competition. The difference is in the sequence used when teaching and combining the two behaviors. Adding a ramp to turning tight does not alter the dog’s understanding of turning tight and the dog learns quickly. But when you change a dog’s turn by making it tighter, from the dog’s perspective, you are not adding a behavior you are trying to alter the dog’s understanding of turning off the ramp. This means modifying an existing behavior, which requires conditioning strategies, and the better you understand the complexities of conditioning the better the result.
Another example. I teach a dog to hold an object until I say ‘drop’ when I play tug exchanging with my dog. I don’t teach my dog to retrieve a ball until after this is established and when I apply the ‘drop’ cue to retrieving the ball then doing an exchange the dog has no problem understanding it. But if I train a fetch first, then try to teach the dog to hold the ball while running after me and my tug, I have to resort to a lot of repetition and shaping and errors.
Another reason that conditioning is required has to do with the genetics of behavior. Inbreeding can result in a behavior, a reaction to a stimulus, being removed from a line of dogs. For example a good working Jack Russell cannot be distracted from its prey, its peripheral awareness has been turned into an over-riding focus. So if you want to be able to distract it from working you will have to condition the behavior. That response is no longer part of the dog’s natural repertoire, the dog no longer sees it as an option. It was bred out of the breed in order to create a better working dog
In summary, the reasons for conditioning behaviors discussed above are;
– needing to modify an existing behavior
– not breaking a behavior down to single associations
– not sequencing associations properly
– the behavior being bred out of the dog
To an extent far beyond that of any other animal, understanding people was a major focus of the evolution of the dog in the wild, and as a result they now intuitively learn from us in ways that are unavailable to the others.