It's Simple, But It's Not Easy

The reality of my work is that I get to play with dogs for a living. I drink coffee, talk about dogs and throw “tennis balls”.

That doesn't mean there is nothing underneath dog training. Quite the opposite. The concepts are often simple, and people are sometimes amazed when I break something down and they suddenly see the simplicity of the task ahead of them and what that task is building towards.

Making something simple is not the same as making it simplistic. That simplicity comes from understanding what matters, isolating it and presenting it in a way that both the dog and the handler can understand. In my writing, I sometimes need to get technical because some of the people reading come from scientific, academic or behaviour backgrounds. If I want to explain properly why I think something is happening, sometimes I need to speak that language.

That language has to be learned. It certainly wasn't natural to me. I remember The Dog's Mind by Bruce Fogle being prescribed reading and struggling to get much beyond the opening pages because, frankly, I had very little idea what was being said.

Over the years, I delved deeper. Properly understanding classical conditioning was a turning point. Understanding operant conditioning was another. More recently, thinking much more about cognition and asking what the dog has actually learned has stepped up my training and troubleshooting again.

That knowledge matters because it helps me explain why something works rather than simply knowing that it does. More importantly, when something stops working, it gives me somewhere sensible to look. But years of actually training dogs have done something else as well: they have allowed me to strip a lot of that complication back out again.

I can go technical when I need to. I just don't necessarily need to carry all of the terminology onto the training field with me. It's simple, but it's not easy. The difficult bit is looking at the dog in front of you, working out what matters at that moment and changing the right thing.

Two languages describing the same world

There can appear to be a gap between the scientific world and the practical one. One may have the terminology and the research. The other may have years of experience applying, adapting and sometimes discovering the same principles without necessarily knowing their formal names.

It also amazes me how often experienced practitioners arrive at similar points in their careers without ever having spoken to or worked with one another. We may describe our training differently and use different exercises, but when we strip those differences away, the dog often tells us that the learning and the destination are much the same. Our experiences simply took us there by different routes.

That convergence is worth paying attention to. The dog is the common denominator. Different practitioners observe, adapt and refine their work until the dog leads them towards a similar answer, even if they give that answer a different name.

I don't see those worlds as opponents. Each has something the other can use. Science can give the practitioner a more accurate explanation of what may be happening, while practical experience forces that explanation to meet the untidy reality of an individual dog, in a particular environment, on a particular day. The dog then provides the evidence.

That is where I want my writing to sit. I want to understand the science accurately enough to explain what I am doing, but plainly enough that someone can use it with the dog standing in front of them.

Making something understandable is not the same as dumbing it down. Equally, giving something a scientific name doesn't automatically make our understanding of it deeper. Terminology should help us examine what happened; it shouldn't allow us to assume that it happened.

Premack and tidying your room

Until yesterday, the term Premack Principle was alien to me. Once it was explained, I understood the basic relationship: a more probable or valued behaviour can be used to reinforce a less probable behaviour. My immediate thought was, is that not just, “Tidy your room, then you can go out and play”?

That isn't to dismiss the Premack Principle. Giving the relationship a proper description allows it to be studied, discussed and applied more precisely. It also encourages us to look more carefully at which activity this individual actually values in this particular situation. But the formal description shouldn't make the practical relationship harder to see.

I had understood and used versions of that relationship for years; I simply didn't know its scientific name. That doesn't mean the science has nothing to offer me. Learning the language gives me access to other people's work. It allows me to examine my own thinking, describe it more accurately and discover where the everyday explanation stops being sufficient.

Knowing the term isn't the same as being able to apply the principle skilfully either. The practitioner may understand the relationship without knowing its name, while someone else may know the name without having the practical experience to recognise when, why or how to use it. The useful conversation begins when we bring the two together.

The dog in front of you

However sophisticated our explanation becomes, it still has to return to the dog in front of us.

What did the dog in front of you do?

We have a habit of attaching labels to dogs:

• The nervous dog.

• The low-drive dog.

• The environmentally sensitive dog.

• The dog that lacks commitment.

• The dog that isn't particularly interested in food.

Those observations may have been absolutely correct when we made them, but when was that? Two years ago? Six months ago? Last week? We are teaching these dogs, so surely that teaching should be changing them.

If my teaching is effective, my description of the dog should have to change. If I have been training a dog for two years and I am still seeing and training the dog that stood in front of me two years ago, I have to ask a fairly uncomfortable question:

How effective has my teaching actually been?

A label can be useful because it helps describe what I am seeing. It becomes less useful when I start training the label instead of reassessing the dog. Observation has to keep updating the picture:

• Maybe this dog was nervous. Is it nervous now?

• Maybe food was low value. Is it now?

• Maybe the dog did lack commitment to odour. Does it still?

Don't power-wash the whole house

That also changes how I troubleshoot. I use an analogy in my talks: if a dog has an accident in the house, I don't get the power washer out and scrub the house from top to bottom. I clean the bit that needs cleaning and work out why it happened. Training should be much the same.

• If the trained final response becomes sloppy, perhaps the trained final response needs some work.

• If the nasal search becomes loose, perhaps I revisit an exercise that allows me to isolate and sharpen nasal search.

• If source localisation becomes poor, perhaps that is where I need to look.

We don't automatically need to go all the way back to the beginning every time something slips, but sometimes we do. Sometimes the thing we can see going wrong isn't where the problem started.

A flat search might be a search problem, or perhaps the dog simply isn't very motivated to search anymore. A slow indication might be an indication problem, or perhaps finding odour no longer predicts anything the dog particularly cares about.

I start with what I can see, isolate it and then work backwards as far as the evidence takes me. I don't automatically return to the beginning, but neither do I stubbornly refuse to go there.

The embarrassingly simple question

Sometimes the evidence takes me all the way back to the reward. Then I arrive at a question that is almost embarrassingly simple:

Did what just happened make this dog want to do that again?

We can talk about primary reinforcers, secondary reinforcers, conditioned reinforcers and reinforcement schedules. These are legitimate concepts, and understanding them can make us better trainers. But the terminology cannot answer that question on its own.

I can hand a dog a piece of food and let it swallow it, or I can make that same piece of food move. The dog chases it, misses it, chases again, captures it and eats it. It is the same food but a very different experience.

Likewise, I can offer a tug toy to a dog that thinks it is the greatest thing ever invented, or offer exactly the same toy to another dog that couldn't care less. The category doesn't tell me whether the reward event worked. The dog does.

Is the reward enough for the dog to do what it just did to access it again?

Why this exercise, with this dog, today?

The same thinking applies to exercises. An exercise is something we deliberately arrange to teach the dog a rule. The broader aim is for the dog to understand the principle and apply it beyond that particular exercise.

Once the dog understands, the exercise doesn't necessarily go in the bin forever. Athletes don't reach the top of their sport and stop training. They practise, maintain and sharpen individual components. If something deteriorates, they work on the bit that needs attention, and sometimes that will mean returning to a foundation.

The important question is why I am doing this exercise with this dog today:

• Am I teaching something?

• Am I confirming something?

• Am I maintaining or sharpening something?

• Am I trying to diagnose something that has slipped?

• Am I simply doing the exercise because that is what my training programme says comes next?

• Or, if I am completely honest, am I doing it because I enjoy it?

There is nothing wrong with enjoying training or enjoying watching a dog perform an exercise well. The problem comes when my enjoyment keeps us inside the exercise after it has served its purpose.

Pots are an exercise to teach odour and indication but are they the final job?

I did this with Focus Walls. I enjoyed the exercise, I enjoyed watching the dog work and I got my own little dopamine hit from seeing it done well. The dog wasn't the only one being reinforced.

But the wall was never the mission. It was an exercise intended to teach intense nasal search. I became stuck on performing the exercise and failed to deliberately carry that lesson into the search area. The exercise looked good, but that didn't mean I had transferred the principle.

That is another reason to keep asking what an exercise is for. Sometimes we continue because the dog needs it. Sometimes we continue because we like the picture it produces.

Theories help me understand, terminology helps me describe and exercises help me teach, but none of them replaces observing their effect on the dog standing in front of me.

Perhaps simplicity in training isn't about knowing less. Perhaps it is what happens after enough years of learning, training, getting things wrong, fixing them and learning some more: you become better at stripping away what doesn't matter and identifying the thing that does.

That is why I love troubleshooting workshops. A dog and handler can arrive with something that feels like one large, complicated problem. I watch the dog and handler together because the answer may lie in the dog, the handler or, as is often the case, somewhere in the interaction between them. We break the performance into its component parts and identify the bit that actually needs attention. Often, the moment everything changes is when the handler suddenly sees how simple the next task really is and, just as importantly, understands what that task is building towards.

My job isn't to make the answer sound clever. It is to use everything I have learned to help the handler see what matters, change the right part of the picture, give the dog a problem it can understand and then observe what changes in the team.

I still drink coffee, talk about dogs and throw “tennis balls” for a living. There is just rather more going on behind the tennis ball than there used to be.

Tom Moir

Coffee Drinker, Dog Watcher, Ball Thrower