Often trainers are telling me that when you train with positive reinforcement you give animals a choice. While at first I would say yes definitely but it is actually not the case. Teaching choices is a complicated task especially the yes and no answer. Why is this?
Yes or No?
When I ask an animal to preform a specific behaviour and the animal doesn’t preform the behaviour this basically means the animal said no to me. The outcome might be that the animal receives an LRS (Least Reinforcing Scenario) or a DRI (Differential Reinforcement of Incompatible behaviour) or DRO (Differential Reinforcement of Other behaviour). Realistically we are actually ignoring the animals answer by giving the animal to be successful in the behaviour once again.
Read this again, “to be successful in the behaviour once again”, when the animal says no isn’t the animal successful anyway? Our focus is that a yes is successful and a no is not. Therefore we won’t reinforce a no and we do reinforce a yes. This means that we create a forced choice to the animal.
Reinforcement Choices
Teaching the concept of consent in regarding to yes and no is pretty difficult. What is easies to think about is teaching the animal the choice of reinforcement. You responded correctly to the signal now which reinforcement do you want to have? That’s an easy and simple choice to make for the animal.
We can build this to a variety of different options. It becomes difficult when we want the animal to make a specific behaviour preference. For example do you want to preform this or that behaviour. Or do you want this trainer or that trainer.
They all have actually one specific thing in common which is predictability. For the animal to choose a specific option we have to be predictable in the outcome of what we present to be a choice for the animal. If the animal chooses a ball to play with I shouldn’t first pet it to give the bal. Because that’s not what the animal choose to have. They asked the ball and not the ball and petting.
Working with many teams around the world at various facilities this is what makes this consent training difficult. The predictability and interpretation of the trainer after a choice is made. When there is a slight difference you bring your animal in doubt about the choice they made and therefore it might not be a proper choice anymore. When we move it even further we are actually able to ask the animal a choice where they understand the predictable outcome regardless if this is an uncomfortable outcome.
Blood Sample or No Blood Sample
You might have seen the video before but for example this animal who is trained for a blood sample. You can see the dog is in the small trey. The trey is meant for the dog to understand that when you step into this object we will be doing a blood sample. The blood sample might be uncomfortable but the reinforcer over takes the discomfort and the animal will therefore choose to participate. Well at least it seems to be this way. For us when the animal doesn’t step into the trey the animal says no. There is not a better way for us to know that the discomfort might be the reason why a no is present. The difficulty now is as explained before what to do from here. Because shouldn’t we be respectful to the choice the animal makes at that moment in time? When we give an LRS, DRO, DRI or a TO (Time Out) aren’t we very unrespectful to the animals choice?
Dolphin Interactions
Whenever a person tells me training is choice and control I believe it actually is not. It is a complicated system which needs to be understood well. When I was visiting the IMATA conference back in the day I remember a specific presentation about a dolphin who showed aggressive behaviour when doing interactions. With the system of LRSes, Reinforcement timing etc they didn’t make much progression and had mayor set backs. They decided to do it differently. They trained this specific dolphin when you touch this target at the other side of your facility everybody will leave the water and the interaction stops.
They gave the animal a sense of control. I’ve learned control is inherently a primary reinforcer. So what happened to this dolphin? The sessions actually progressed without the occurrence of aggressive behaviour. This is a great example where the trainers decided to let the dolphin decide when enough was enough. Remember that this specific target predictably meant “everybody should leave the water”.
Choice and control is not as easy as it seems. It is difficult and has to be thought about very well.
You can find a great presentation about the animal choosing from a couple different reinforcers on our Patreon.
0 Comments