What Does the Dog Think I Am Teaching? Introducing the Detection Dog Learning Framework
- Tom Moir
- Aug 1
- 5 min read
What Does the Dog Think I Am Teaching?
When I set up a training session, I know the principle I am trying to teach.
By principle, I mean the underlying understanding I want the dog to apply beyond that particular exercise.
The dog, however, has to infer (work out from the information available) a rule from the experience I provide. By rule, I simply mean the answer that has worked up to this point to:
What matters here, and what do I need to do to achieve the outcome (make reward happen)?
Ideally, the rule the dog learns reflects the principle I intended to teach.
It does not automatically follow that it will.
I know where the hide is, why I have put it there, what I am going to change and what I hope the dog will learn from it.
The dog knows none of that.
The plan exists in my head. From the dog’s point of view, each session is simply another experience from which it has to work out what matters.
That creates a problem for us as trainers, because what we intend to teach and what the dog actually learns are not always the same thing.
A dog may perform an exercise successfully because it has understood the principle behind it. Equally, it may have learned the sequence, the handler’s position, the familiar environment or how far it normally travels before searching properly.
We may not notice the difference until something changes.
Move the handler, change the environment, place the hide earlier in the search or remove one part of the familiar routine and the behaviour starts to fall apart.
We often respond by saying that the dog needs more training. Perhaps it does, but more training in what?
More repetitions of the same exercise may simply make the dog better at doing that exercise. If the dog has learned a different rule from the one we intended, we may also be strengthening the very thing that is causing the problem.
Before doing more, I think we need to ask:
What does the dog think I am teaching?
That question sits at the centre of the Detection Dog Learning Framework.
The lesson we provide
The dog experiences the whole training picture.
It experiences the environment, the order in which things happen, our position and movement, what earns reward and every pattern we repeatedly place around the task. It then uses that information to work out what leads to a successful outcome.
The rule it learns may be exactly the one we intended, but it may not be.
If a dog learns that searching only begins after it has travelled through the first part of an area, we need to look at the searches we have arranged.
If odour only appears to matter when the handler reaches a particular position, we need to consider how that position became part of the picture.
If the dog can complete a familiar exercise but cannot apply the same learning elsewhere, perhaps we taught the exercise rather than the principle behind it.
The dog has not misunderstood our lesson plan. It never knew the plan existed.
It has learned from the experience we provided, which means we are responsible for examining that experience when the learning is not what we expected.
Why a Detection Dog Learning Framework?
The Detection Dog Learning Framework looks at the relationship between the mission we are preparing for, the training we arrange and the rule the dog may learn from it.
It is not a new method or a set of exercises to follow. It can be used to examine any method, including our own preferred ways of training.
The framework asks us to look beyond whether an exercise worked and think about what learning may have produced the behaviour we saw.
It also changes where we begin.
Rather than starting with, “What exercise should I use?”, we start with:
What understanding am I trying to develop?
Once we know that, we can design an exercise that gives the dog the best chance of developing it.
Exercises are useful teaching arrangements, but they are not the destination. The aim is to develop learning that survives beyond the exercise and remains useful when the environment, handling or operational picture changes.
Successful performance still matters. It simply does not tell us everything we need to know.
Who is the framework for?
I wrote the framework for detection dog handlers, trainers and instructors who want to look beyond what the dog did and consider what it may have learned from the training that produced it.
It may also be useful to behaviour and cognition professionals interested in how learning history, context, generalisation and handler influence appear in practical detection work.
For organisations responsible for training, deploying or assessing detection dog teams, it provides a way to examine whether the training and assessment genuinely reflect the mission.
Are we preparing the dog for the task it may eventually face, or are we preparing it to pass an exercise that looks a little like that task?
You do not need to use a particular method to apply the framework. You only need to be willing to question whether the behaviour in front of you comes from the understanding you intended to develop.
Why I wrote it
The framework has grown from my work as a handler, trainer, instructor, assessor and adviser across military and commercial detection work and companion scentwork.
It has also grown from getting things wrong.
There have been dogs that did exactly what I expected in training but struggled when the picture changed. There have been exercises I repeated because the performance looked good, without stopping to ask whether I was developing anything transferable.
There have also been occasions when the dog found a perfectly reasonable way to solve the problem, just not the way I thought I had taught it.
Those experiences changed how I look at training.
I still care whether the dog succeeds, but I also want to understand why it succeeded.
What information was available? What patterns had I repeated? What rule might make the dog’s behaviour reasonable?
The Detection Dog Learning Framework brings together the questions and principles I now use to work through that process.
Download the framework for free
I am making the Detection Dog Learning Framework freely available.
It is not something I expect everyone to agree with. It is there to be read freely, questioned and tested.
If it causes us to look more carefully at the relationship between the mission, the training we provide and what the dog may actually be learning, then it has done its job.
When you have read it, I would be interested to know what this question makes you reconsider in your own training:
What does the dog think I am teaching?
Tom Moir
Detection Coach
Comments