One topic becomes more and more important in animal training: giving animals more choice and control during training and also in their normal life. I think the definition of “choice” is not so easy and changed a lot during the last 10 years. In the beginning of “force free training” or whatever we might call it, we or most people thought, that choice would be defined as “saying no without unpleasant consequences”.
As most people would have trained with negative reinforcement and positive punishment, that would mean without positive punishment. So back then a lot of people thought, that training with positive reinforcement means giving choice to the animal – and for many people this is still true today. But the deeper we dig, the more complex the topic becomes.
Yes, training with positive reinforcement generally means, the animal has more choice and therefore control than in working with traditional training. But just because it’s positive reinforcement training does not necessarily mean that the animal has choice.
By falling into the following traps you might unintentionally restrain the animal’s choice:
- working with high value reinforcers,
- working with high value reinforcers,
- a lack of alternative activities or free ressources available during training
- negative punishment / extinction due to too high criteria or to withholding the click; mixing quadrants (often without recognizing it), which leads the animal into a cognitive dilemma and can negatively affect the training,
- working with (food) deprivation (some trainers still follow the mantra “ressources need to be earned” to create “obedience”). This can also mean willingly choosing to keep a social animal in isolation,
- and even the time of day can have a huge impact, if the animal gets its meals on a regular schedule.
And this is just the the first level of “providing choice” and a broader definition. There is a lot of space between “absence of unpleasant consequences” and “having more than one pleasant option to choose from”. Such as choosing between different options, without creating a disadvantage for the animal (e.g. choosing “no” under the same conditions as “yes”).
So even when we are training with a cooperation signal or a “start button”, this needs to be taken into consideration. Yes, it is a good idea for some tasks to give the animal a way to say “yes” to the behavior, for example by adding a start button. But as long as we don’t reinforce the “no” the same way we would reinforce the “yes”, there is still a risk that the animal only takes “yes” because it’s a chain of behaviors and just one option will be reinforced. This is just one example of how difficult it can be.
And there are still some questions I need to think about, such as: “Does choice need a conscious, willing participation, especially when we think about classical conditioning?” Or do we take choice away, when we are just conditioning to a stimulus? Well, I think this is more of an ethical question, where everybody needs to make up their own mind . After all, we can’t always choose freely and sometimes we can’t even control what we need to train, because some things are just necessary (however, we can choose HOW to train). In any case, it makes sense to think this through and to become more aware of the implications for our training, regardless of how we decide in the end.
So I think, especially with regards to the experience of the trainer/handler, it’s just like the “least intrusive, minimally aversive”-agreement, we should know our options and provide as much choice and control as possible, but at the same time know, that it’s not always possible under all circumstances.
This is just a quick sum-up of my opinion on this topic. Let me know, what you think and if you want to read more about it!