The Balance of Reinforcement

Life is about balance I mean when one part in your life to high the other parts wont have the same meaning anymore. We see this when friends of family pass away or even our animals. The balance will be gone for a little bit. Balance is important, I discovered that when having a great job working with my dream animals that when your private life is not superb that it doesn’t really matter. Because you won’t feel 100% happy anyway. The same for the opposite side. Your private life is amazing but your job isn’t working well you won’t be happy neither. In many parts of living our life there are moment where we should think about this balance. To be honest the hardest ones are relationships.

In training session, it happens easily that we forget to generalize behaviours we train. It is as important to train the animal on signal as of signal, just for you to know that the animal really understands your signal. For example, if its more reinforcing for an animal to go in a transport box then it is to stay out of it, the balance of reinforcement might be leaning towards the transport box more than not going into it. We maybe should ask ourselves if generalization worked out well here?


This is a training session with one of Kolmårdens Zoo’s Hornbills. She is still getting trained to be comfortable in a box while the door closes. It is important to connect high value reinforcement with being in the box itself. The balance is higher when she goes in the box then staying outside in a later stage this will be balanced out again.

Where it becomes more difficult is that we oversee reinforcers for the animals we work with. I know stories where us trainers are to focussed on food reinforcement and forget the other reinforcers what might be sometimes in higher value. You might wonder what that would be… one day on my run in the forest, I passed a Eurasian Jackdaw eating a Grass Snake.Interesting to see from a distance, the fact that the bird was smacking the snake on a rock took my interest because of the tool use. When I walked by and had a closer look the Jackdaw left the snake and flew away to look at me from a distance. The question I asked myself right after was if the food source is so important and people would say is the best reinforcer why didn’t the Jackdaw stay instead? You se this type of behaviour a lot in animals who prefer to run away like antelope species.

When we talk about observation we mainly talk about the moment when us trainers observe the animals but do we ever think about the observation the animals have to us?

I’m wondering sometimes, when we become too predictable the animals start do work against us (or are they just bored because they know what’s coming?) trainers call this “the animal plays with me”. When we see, this happening the variation we give to the animal, what some of us never would do might be more reinforcing then that fish, meat or piece of appel we give after every behaviour. You know that dance when we become insecure and try to get animals back? Animals are observant as well. I believe that I don’t have to reinforce every bridge. Depending on how you see reinforcement we might reinforce every bridge without us knowing it due to the observation the animal has. From out the standpoint of reinforcing every bridge yes or no… what I feel is an ongoing discussion because some say yes some say no. I’m going to explain you my experience and what has worked very well with the animals I had the privilege to work with.

When a bridge is properly established I do not feed every bridge anymore. When I would reinforce every bridge with a reinforcer all the time the difference I want to make when the animal reaches the goal I want in the session might be very minimal. When I do not reinforce after every bridge what is strategically chosen within my session and observation. I’m able to make a bigger point to the animal when they reach the exact goal in my session. Obviously, you need to know your animal and their history. I find it working so well that I actually get the animal to learn with me or I’m shaping behaviours the animals has a higher pace to understand what I want in the steps taken at that point.

Successive approximation plans should be small, reinforcement should be planned right, the trainer should know the animal and the history, and the trainer should understand if the reinforcement given to the animal is actually reinforcing to the animal. To be honest I reinforced animals with jackpots what didn’t work at the end at all.

Reinforcing every bridge is more than thinking about yes or no. There is a lot more coming to it, but I believe that the balance of reinforcement should be considered more than the bridge discussion itself.

One day and this is 2 years ago I would say we were training the fur seals here at Kolmårdens Zoo. After a successful session we let all animals out what we do after every session, to our surprise we saw that we forgot a target. The target was lying outside the house and we couldn’t reach the target ourselves. At this point our young male came back and looked at us from a short distance he came to us and was standing at the door looking at us. At this point I decided to see if I could get him to give me the target. The only problem I had was that I didn’t have any fish with me. I had to think what I would reinforce him with and the only real option I had was clapping and praising but I had never really consciously established those types of reinforcers to him, or did I? I mean I might have without knowing it. After 3 minutes, I had my target back. It was interesting because would it be that the animal just likes to learn what might be reinforcing for him anyway? If it would be the case when we should reinforce every behaviour with a well-established reinforce why did he keep on going in participating in a 3 minutes session where no other reinforcers were given after the bridge?

If our balance of reinforcement is planned strategically and the whole team works in the same kind of structure then it might be that the variation is higher in reinforcement then the actual primary reinforcer. There is a lot to it not to forget to mention the species you work with. Killer whales will understand this variation very well so do dolphins but pinnipeds need some more time to get into that but its certainly not impossible.


Think about the balance of reinforcement once in a while. Training our lions is another balance of reinforcement story. We do not ask this behaviour often, it’s not because we are lazy or do not have any time. It’s because the less we ask the bigger the difference, the better the response. When we would ask it more often we should reinforce more often. We want to have a strong response, there for we do it less and can give them more variation in the amount of reinforcement and what kind of reinforcement we have for them. This only works with being unpredictable and thinking about the balance of reinforcement.


Think about it when animals do not want to come inside, or when animals only use a part of their exhibit. Maybe the balance of reinforcement is not right.


Peter Giljam

“Thinking Outside the Zoo”


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.