How to train animals (and use your powers responsibly)

10-minute read

Obviously this isn’t an instructional how-to for controlling animals with your mind, but there are hundreds of years of study on animal training and we now have it down to a fine art. Maybe not so fine in some areas. If you’ve listened to A Dolphin Pod you’ll know what I mean.

One of the pioneers of animal behaviour was a 20th century zoologist called Konrad Lorenz. He did a lot of work on imprinting, which is the process by which baby birds learn to identify their mother. He divided a clutch of goose eggs into two groups. One he left with the mother, and the other he incubated and made sure he was the first moving objects the chicks saw.

The goslings that met Lorenz first followed him everywhere, and when the babies were all jumbled up, they immediately ran to either Lorenz or the mother goose, depending on which they’d imprinted upon.

To be clear, this isn’t an example of training. Imprinting is an evolved strategy to help young birds understand whose example to follow, both figuratively and literally. Lorenz just learned how to manipulate it.

At the dawn of animal behaviour science there were a lot of interesting discoveries that laid the foundation for training to become its own science. Ivan Pavlov worked with dogs in the early 1900s. He was measuring their salivation in response to red meat, but noticed that the dogs would begin to drool when they heard the footsteps of the assistant that brought in the food.

He discovered that animals can be conditioned to do things, and soon he was able to make his subjects dribble with the sound of a bell that he’d associated with the arrival of their meals. He was able to modify their behaviour, albeit in a very small and weird way.

The next big advance came from B. F. Skinner in the 1920s. He figured if an animal was given a reward, it might be more likely to repeat a behaviour. He tested his theory with a box he named after himself.

A hungry rat would be placed inside the Skinner box, which contained a lever that would release a pellet of food. At first, the rats would scramble around inside the box, and eventually come into contact with the lever. The kibble would appear and the rat would stop to eat it. After a few trials, the rats ran straight to the lever and chowed down without delay every time.

I can’t let this moment pass without mentioning John Watson, whose research was never completed because it crossed ethical boundaries. And it wasn’t on animals, but humans. Settle in.

It’s more commonly known as the Little Albert experiment. He’d noticed that kids didn’t seem to like loud noises, and he wondered if he could harness that fear and transfer it to another object.

He was then able to obtain a three-month-old boy from a nearby hospital. The baby was described as ‘emotionally stable’ at the start of the study. He was given access to all sorts of items and animals including a white rat, cotton wool and newspaper. The infant, who was codenamed Albert, didn’t have a fearful response to any of the things he was exposed to.

But in the next stage, Watson put Albert on a mattress on a laboratory floor and gave him a white rat to play with. Then he would make a loud sound behind Albert’s back by hitting a steel bar with a hammer every time the boy touched the rat.

After a while, Albert would scream and cry whenever he was presented with the rat. Watson had successfully instilled a phobia in a child. I hope he was proud of himself. But it didn’t stop there. Albert soon became just as afraid of the other small white objects in the study, cotton wool included.

This experiment would be against the law today, and for good reason. Plus, it’s not even good science. He only used one subject and no control subjects. And Albert had to be removed from the experiment because he was getting too upset. I just love the way it’s worded on the Wikipedia page, check it out.

Albert was about seven years old at the end of the experiment, and he reportedly left the hospital shortly thereafter. Though Watson had discussed what might be done to remove Albert's conditioned fears, he chose not to attempt such desensitisation with Albert, and it is thought likely that the infant's fear of furry things continued post-experimentally.

Seven years of phobia inception?! It’s unreal. That poor kid. But all of these studies, and a whole pile more, gave us exactly what we needed to craft the craft of animal training.

I’m going to use dogs as my main example here. I think you’re more likely to need to train a dog than a dolphin, especially if I succeed in my mission to make dolphin captivity a thing of the past worldwide.

The very first step in a training session is to get the animal to give you what we call a good OR. That stands for orientation response, and it really just means that the animal is paying attention. You want it to be in front of you, facing you, looking right at you. Having something rewarding on your person will help with this, like treats or toys.

There are two main forces when it comes to animal training - reinforcement and punishment. Both of these can be positive or negative and serve to increase or decrease the frequency of a behaviour.

Positive reinforcement - Giving a reward when the animal performs a requested behaviour.
Example - Asking your dog to sit, then giving them a treat when they do.

Negative reinforcement - Removing a negative stimulus when the animal performs a requested behaviour.
Example - Asking your dog to sit while wearing a shock collar, only turning off the current when the dog sits. Don’t do this one.

Positive punishment - Applying a negative stimulus when the animal performs a behaviour that has not been requested.
Example - Stomping on your dogs back paws if it jumps up at you. Don’t do this one either. Believe it or not, that was a legitimate piece of advice you’d hear at dog training classes in the 70s and 80s.

Negative punishment - Withholding a positive stimulus until the animal performs a requested behaviour.
Example - Holding a toy up then asking your dog to sit, only giving or throwing the toy when the dog sits.

My personal training approach is a nice blend of positive reinforcement and negative punishment. I give lots of treats and rewards, and use very passive punishment that doesn’t cause my pupper too much distress.

You can use these to help you establish the first stage of getting a solid OR. If your animal immediately focuses on you, you can use positive reinforcement by giving a treat.

You could negatively reinforce by making a weird or high pitched sound until your dog focuses on you. Stop making the noise when you get the OR. Then hand over a treat and positively reinforce because your dog did the right thing.

The most positive punishment I would ever deliver is a verbal ‘no’ signal. I say ‘eh’, but like a cockney gangster, all low and growly à la Danny Dyer. My dog knows that when I make that sound he needs to obey me. I only use this in emergency situations, like walking off-lead near busy traffic or if he’s about to eat something that will make him ill.

If you really can’t get your dog to pay attention to you, it might be worth considering using a very small amount of positive punishment to make sure your dog listens to you. Please never, ever hurt an animal or try and really scare it. Not only is it barbaric, but it’s wholly unnecessary.

Dogs are in a state of arrested development. That means mentally speaking, they’re puppies their whole lives. If we human owners can act like a mother dog, your pet will always know where they stand in the family and will want to do what you say. Ugh, I’m starting to sound like a 21st century Konrad Lorenz.

A mother wolf will growl and snarl to communicate with her pups. If there’s danger present, it’s better for her to startle them into submission than for them to be torn apart by a predator. An occasional guttural grunt can get your dog’s attention, and if you positively reinforce their OR they’ll be far more likely to do it again, just like Skinner found.

But once the dog’s focused, how does one get started with actual training? Dogs don’t speak English, how does asking an untrained animal to sit down make them any more likely to sit down?

It doesn’t. Well observed. But we know enough about natural dog behaviour to work around that. A lot of training is waiting for a behaviour to occur naturally, then rewarding the animal for it. If a dog is hanging around with nothing to do, it’s highly likely to sit down by itself.

You can then associate that behaviour with a cue and give the dog a reward. If you repeat that process enough times, soon the dog buns will be hitting the floor the moment you ask. You can train an animal to do almost any natural behaviour on command with this method.

If you want more control, you can start target training. You could use an implement such as a ball affixed to the end of a pole, like dolphin trainers do, but it’s much easier to use your hand.

Touch your dog’s nose, then give them a treat. Do that a few times, and then hold your hand a few centimetres away. If your dog reaches out and boops your hand, give them a treat. Keep it up, and you’ll be able to get your dog following your fingertips and open up doors to new behaviours that would be unlikely to occur naturally.

There’s another concept to add in to the process now. Training should be precise, and we have to make it clear to the animal exactly which behaviour is being rewarded. To achieve that, we make a consistent sound every time the animal nails a behaviour that the animal learns means ‘yes’. We call it a bridge, and annoyingly we use the bridge as a noun and a verb.

Some dog trainers go full SeaWorld and use whistles, but I personally eschew that for the sanity of the other dog walkers in my town. You could also use a clicker. That’s a little hand-held squeezy button that makes a clicking sound. I’ve had good success with clicker training with animals of all kinds, but when it comes to dogs I prefer to give a verbal bridge. I can forget a clicker, but I can’t forget my vocal cords.

When my dog does what I ask of him, I simply say ‘good boy’ and give him a treat as quickly as I can afterwards. I always use the same tone of voice for my bridge since that’s what he recognises when I speak, not the actual words.

Training a bridge of any kind is very simple. You get a good OR, you bridge, you treat, you repeat. Associate the bridge with a reward and the animal will soon learn that hearing it means food, toys or cuddles are due.

These kind of training sessions are such good bonding opportunities too. Setting out 10 minutes in the evening to call my dog a good boy and feed him kibble makes us both happy. And all the while, he’s learning that when I call him a good boy, he’s being a good boy. It will keep him motivated to keep being a good boy.

The only way to make any kind of training stick is to be consistent. When I run my training seminars I always tell people the same thing. Training is simple, not easy. The rules are very straightforward, we just have to be on the ball enough to implement them perfectly every time.

Even when your dog has learned a new behaviour, you have to keep practicing it. Maintenance of learned behaviour is essential to make sure it’s fresh in the dog’s mind. There’s no point training a dog to wait on command then hoping they still remember six months later in an unexpected emergency. You have to practice.

Dog training goes way beyond owner and pet relationships, it’s become essential to countless groups of people. Dogs assist disabled people, sniff out drugs and police evidence and diagnose medical conditions in humans.

There are even sniffer rats working in the world. African giant pouched rats were trained to scour landmine fields across Africa and Asia, and have successfully removed the danger from the small communities that were once plagued by accidental explosion injuries. Rats also help detect contraband at border controls around the world, looking for items like weapons or illegally smuggled endangered animal parts.

Rat training really comes into its own in the laboratory. 111 million mice and rats are used in American biomedical research every year. They make up more than 99% of all US lab animals. Sadly, a grand proportion of these are used for tumour research rather than behavioural studies like running through mazes.

There are plenty of other species entrenched in science, some of whom are trained extensively to participate in studies. But the bulk of training these animals get is practical nuts and bolts stuff to make day-to-day life in captivity less stressful.

It’s the same with zoo animals. They have to be desensitised to scary stimuli like crowds and veterinary treatment. In essence, they have to be trained to tolerate humans. Zookeepers have to train animals to participate in medical treatment, accept being moved between enclosures (or sections of enclosures) and for general mental stimulation.

Captivity can’t compare to the wild for most animals. Many zoo and aquarium exhibits are very bored because there isn’t enough to do in their enclosure. Having daily training sessions helps keep them entertained, and strengthens the bond between animal and human (or weakens the bond between an animal and its natural instincts, it depends how you look at it).

But there are plenty of animal establishments that train their animals for shows and public interactions. This is where the ethical question about using animals for entertainment purposes comes up proper. Plus, it’s difficult to judge whether rigorous training is in an animal’s best interest when their owners profit from their performances. It means they’re motivated to tell people the animals are happy. If people thought they weren’t happy, they would stop buying tickets.

Usually getting to have an up-close animal encounter at a zoo or aquarium will cost you extra. That’s on top of the entrance fee you already paid, of course. That means the facility has an incentive to keep up potentially stressful training and exposure to unfamiliar people, who sometimes behave in unusual ways. As a former zookeeper that oversaw many public interactions, I’ve seen it all. Animals being dropped, thrown, hurt, stolen, the lot. Y’all need to have a good look at yourselves. In fairness, my participation is just as bad if not worse. Nobody wins here.

Probably the most lucrative one-on-one animal experiences on the market is a dolphin interaction. They’re so profitable that many of the big new facilities don’t bother doing traditional shows, but instead have the dolphins split up individually working with people directly from dawn to dusk.

Every single aspect of these animals lives is under human control. Every scrap of fish they get is in exchange for something. The never-ending conveyer belt of customers make the people at the top get very rich, but I’m afraid I fail to see how the dolphins benefit in any way.

If you haven’t heard my podcast, please listen when you get a chance. There’s a lot of information about what captive dolphins are subjected to on a daily basis. It’s terrible.

It’s important to remember that training can wander over the line into outright abuse, like in the case of Jose Luis Barbaro. He was the subject of the documentary ‘The Last Dolphin King’, and committed suicide after a video surfaced of him verbally and physically abusing his dolphins in the name of putting on an entertaining show.

He was using positive punishment, applying a negative stimulus when the animal makes a mistake. That’s the wrong way to train, and when it comes to dolphins even the right way is still a convoluted perversion of nature. You can quote me on that.

I see terrible training and infliction of punishment with my own eyes when I’m out walking my dog. It’s usually another owner who clearly doesn’t understand natural dog behaviour. They expect total obedience, like a robot rather than a living being with its own mind.

The animal gets punished for stopping to relieve itself, sniff something or trying to play with another dog. You don’t need to exercise the ultimate control. It’s much better to let your dog be a dog. Let it sniff, pee and play. That’s what they need.

Just like dolphins need room to swim, make new friends and catch themselves fresh fish. That’s what they’ve evolved to do, and just because they can be trained to do spectacular tricks in captivity doesn’t mean they should. Just like John Watson shouldn’t have set out to traumatise a toddler. We all agree that was bad.

Dolphins face some of the most rigorous training in the business and the proceeds of their labour end up in the pockets of billionaires. It’s about time we all agreed that’s bad too.