Hi,
When you are presented with a training puzzle, the constructional approach suggests to start with the identification of your goal, followed by an assessment of your current position in relation to that goal.
My goal was a saddling routine during which Blondie remained calm, relaxed and engaged from approaching with a saddle until the saddle was girthed up for mounting.
You have seen Blondie’s behaviour before training in email 16.
The training environment “paddock” was selected because Blondie was already calm and relaxed there and I could build the desired behaviour from a successful starting point.
The constructional approach further suggests that building the skill, the lack of which constitutes the problem, is essential.
My goal was to build active participation in “putting on the saddle”. If Blondie acquired the skill of cuing the next step in the saddling routine, the escape and avoidance behaviours had no function and would not occur.
To get from my starting point to the goal behaviour, I needed to rebuild the saddle in a way that maintained the calm and relaxed behaviours throughout the process and giving her the skills to communicate the next step.
I applied a stimulus change procedure during which I cued Blondie to target an object with a specific body part and after each successful trial, I changed the object slightly. I did not repeat the same object a certain number of times, instead I kept changing something small every time. But the change was so small, that the movement cycle was maintained.
Blondie’s movement cycle was: approaches target, makes contact with target, consumes the treat, approaches the target, etc.
My movement cycle was: present target, click and feed, move away.
Dr. Jesús Rosales Ruiz and Mary Hunter call this an interconnected chain between teacher and learner.
I present target —> Blondie approaches and makes contact —> I click —> Blondie orients to me —> I offer the treat —> Blondie eats the treat —> I move away and present the target.
Alexandra Kurland calls this Cue communication.
I call it a (behavioural) dialogue, individual A cues the behaviour of individual B, which cues a behaviour of individual A, and so on.
My aim in the shaping process is that this communication continues fluently and in the same rhythm while I change the target to a small degree (only as much as the rest stays the same).
I present target variation A—> Blondie approaches and makes contact —> I click —> Blondie orients to me —> I offer the treat —> Blondie eats the treat —> I move away and present the target variation B —> Blondie approaches and makes contact —> I click —> Blondie orients to me —> I offer the treat —> Blondie eats the treat —> I move away and present the target variation C, etc.
I vary continuously until I have a saddle. Magic ✨
Before I show you the first step, let me explain what you’ll see.
Initially, I thought I could use a long band to simulate the girth, but that was way too difficult for Blondie. She expressed that as getting tense, ears back, walking away. Instead, I started with a rolled up leg wrap. In addition, I placed another roll and her fetch toy in the training area to offer an alternative way to obtain reinforcement, so she was not coerced into the interaction with the new object.
For more info on this aspect listen to Dr Joe Layng’s talk about Degrees of Freedom, e.g. on the Equiosity podcast. It’s a fascinating discussion about how you can still coerce an individual despite using a positive reinforcement procedure.
You will also notice how I strategically change the body part requested to target the object. I start by asking the higher neck, then base of the neck, then shoulder and then thorax or girth area. Starting from the safest to the more challenging is something a learned from the Feldenkrais method. I maintain that same pattern as I begin to slowly unwrap the bandage – always maintaining the dialogue. Horses often appreciate this predictability.
Ok, I think you have the crucial information. Here’s the first clip.
You see that it is a very simple behaviour. The magic lies in the decision process before asking for that simple behaviour:
Identify an environment where you already get something you can reinforce Find your starting pointBuild your new behaviour in clean loops.
We’ll continue building the saddle in the next email.
Happy training,
Michaela