Beginner’s Dolphin Training Manual
Contents
Introduction
Classical Conditioning
Background
Application
Strengths and Limitations
Operant Conditioning
Background
Application
Strengths and Limitations
Observational Learning
Background
Application
Strengths and Limitations
References
Introduction:
Firstly, congratulations on officially joining our SeaWorld family. Owning this manual signifies your involvement in the dolphin training division. Chances are you’ve seen one of our dolphin shows in the past, and although our well-trained dolphins are the stars of the performance, none of it would be possible without our very capable trainers. This manual will teach you, the trainer, how to professionally handle and train these beautiful creatures with care. You will be walked through the three main methods of conditioning and training dolphins, being classical conditioning, operant conditioning, and observational learning. Each method will be thoroughly explained, starting with background information, followed by an in-depth guide for application with dolphins, and concluding with the effectiveness of the method, in terms of strengths and limitations. In case anything confuses you, don’t hesitate to approach any of our experienced trainers for assistance. We hope you enjoy this inspirational journey with our dolphins, and that your transition is as smooth as possible.
Cale Dobson, Head of Dolphin Training Division
Classical Conditioning
Background:
Classical Conditioning is a type of learning wherein the trainer pairs two individual stimuli, resulting in the dolphin forming an association between the stimuli. For example, A toddler feels joy when eating lollies. Every time the toddler is given a bath, he is given a lolly afterwards. 3 years later, he finds that he feels joy when having a bath, despite not receiving a lolly at the end of his bath. In this instance, classical conditioning has been applied to make the toddler feel joy when having baths. Various experiments have been conducted that prove the results that pairing an unconditioned stimulus that triggers a natural response with a neutral stimulus that has no initial involvement with that natural response will cause the neutral stimulus to trigger the natural response as an involuntary action or reflex. Some key theorists include Pavlov and Watson.
Pavlov’s experiment revolved around the salivation of a dog. The dog’s natural response was to salivate when given meat powder. Pavlov then rang a bell each time the same dog was given meat powder to eat over a certain period of time. After a while, Pavlov did not provide any meat powder, and only rang a bell. To his content, the dog salivated just as much as he would have if there was actually meat powder present. His conclusion was that classical conditioning was applicable to animals.
Watson’s experiment, on the other hand, was applied to a child (Little Albert) instead of a dog. While this is by no means ethically considerate or an acceptable practice nowadays, it did provide a great deal of information for the foundation of classical conditioning. In this experiment, it was observed that Little Albert displayed fear (through crying) when a loud, scary noise was produced. Little Albert was then placed in a room with a small white mouse, and every time he made contact with the mouse, the loud, scary noise was emitted resulting in Little Albert crying again and again. After a number of times, Little Albert was not only scared of the small white mouse, but any small white object that resembled the mouse such as white fuzzy slippers or a white rabbit (Stimulus Generalisation). His conclusion was that classical conditioning was also applicable to humans.
Now that you have a grasp of the concept of classic conditioning, an example application of this for teaching dolphins is provided. Firstly, follow the guide step-by-step and ensure its mastery, and then try applying your own variations of the guide based on the example given. Good luck!
Application:
Using Classical Conditioning to teach dolphins to associate the sound of a whistle to blowing air out of their blowholes, resulting in water being sprayed upwards like a jet.
This trick is essentially just the dolphin exhaling, using the blowholes located on the top of their bodies which directly link to their lungs. Since dolphins are mammals and not fish, they can’t actually breathe in water as they don’t have gills. When surfacing to breathe, dolphins exhale with a strong jet-like force from the top of their blowholes, resulting in the water looking like it’s coming out of their blowholes. In this instance, you will be conditioning dolphins to exhale when you blow your whistle.
Required items:
whistle
Instructions:
Have your whistle in hand, ready to blow into it at a moments notice.
Simply watch the target dolphin, not giving any signals as it swims around.
Eventually, the dolphin will exhale, spouting jet-like water upwards from its strong force. Immediately blow your whistle when this happens.
Repeat this every time the dolphins exhales and shoots water upwards. Different dolphins take different amounts of time to create the association, so patience is key.
If done correctly, blowing your whistle will make the dolphin involuntarily exhale, shooting water upwards from its blowhole.
Make sure to repeat this frequently, making sure the dolphin’s association does not diminish.
Strengths and Limitations:
Strengths:
Classical Conditioning has one of the highest success rates in both humans and animals, which mean dolphins can be conditioned effectively, as it will be an involuntary response. Also, many experimental researches can back it up, meaning its a reliable theory of learning.
Limitations:
Limitations of Classical Conditioning include that the response being conditioned must be a pre-existing reflex. It can’t be conditioned if the subject isn’t already capable of exhibiting the behaviour. Also, if the two stimuli aren’t associated in quick succession, an external stimulus may influence the association and result in a wrong association.
Operant Conditioning:
Background:
Operant Conditioning is a type of learning where an association between behaviour and consequence is formed by the dolphin. It should be noted that in operant conditioning, the response of the subject must be a conscious behaviour, requiring thought and a resulting action. Basically the theory is that good behaviour is rewarded with a positive consequence, which encourages the continuation of that good behaviour. Bad behaviour is punished with a negative consequence, which discourages the behaviour, eventually causing it to diminish. There are four types of consequences, two being reinforcements and the other two being punishments.
Positive Reinforcement – Introducing something that the subject likes after correctly performing a behaviour, which increases the likelihood of the subject repeating the behaviour. For example, a child cleans his room without being asked. His parents provide the positive reinforcement letting him watch the television for 30 extra minutes. This results in the child cleaning his room more often, with hopes of gaining more television time.
Negative Reinforcement – Removing something that the subject does not like after correctly performing a behaviour, which increases the likelihood of the subject repeating the behaviour. For example, a child cleans his room without being asked. His parents reward him by removing the child’s “grounded” status. This results in the child cleaning his room more often, with the knowledge that his parents may reward him in some form if he does so.
Positive Punishment – Introducing something the subject does not like after incorrectly performing a behaviour or performing and undesirable behaviour, which decreases the likelihood of the subject repeating the behaviour. For example, a child refuses to eat his vegetables. His parents punish him by “grounding” him. This eventually results in the child being more vigilant in eating his vegetables, from the fear of being punished through being grounded again.
Negative Punishment – Removing something the subject likes after incorrectly performing a behaviour performing an undesirable behaviour, which decreases the likelihood of the subject repeating the behaviour. For example, if a child refuses to eat his vegetables, his parents would punish him by taking away 30 minutes of television time. This results in the child choosing to eat his vegetables more often, in fear of receiving another punishment.
Some key theorists of operant conditioning include Thorndike and Skinner.
Thorndike’s experiment involved placing a cat inside a puzzle box, which had a lever that opened the box when pressed from the inside. A positive reinforcement (scraps of fish) was placed just outside the box for the cat to see, but not reach. The cat in the box curiously explored around until it stumbled upon the lever which opened the box, and it claimed its reward of fish scraps. Thorndike then placed the cat back inside the same box, and repeated the procedure. After a escaping the box a couple of times, the cat learned that pressing the lever will open up the box, and was faster at escaping each time to obtain the fish scraps. Due to this experiment, Thorndike is now regarded as the forefather of operant conditioning. His conclusion was that animals learned responses through experiencing consequences, repeating responses that were rewarded, and stopping responses that were punished. This is now regarded as the Law of Effect.
In Skinner’s experiment, he designed a puzzle box. It had a lever and signal lights on one side of the box, and an electric grid as a floor. When the red signal light lit up, food would come out of a dispenser given that the rat pulled the lever. When the blue signal light lit up, water would be dispensed if the rat pulled the lever. However, if the lever wasn’t touched while the lights were lit, the rat would be shocked through the electric grid. This box is now referred to as the operant chamber, or Skinner’s box. After a certain period of time spent in the box, the rat figured out when to pull the lever to both obtain food and water, and avoid getting electrocuted. Skinner’s conclusion resulted in the term “reinforcement” being included into the Law of Effect. Behaviour that is reinforced is strengthened, and behaviour that is not reinforced is weakened.
*DOLPHINS SHOULD NEVER BE GIVEN A CONSEQUENCE THAT WILL HAVE A NEGATIVE EFFECT ON THEM IN THE PROCESS OF TRAINING UNDER ANY CIRCUMSTANCE, AS IT IS UNETHICAL AND HARMFUL TO THE DOLPHINS.
Application:
Using Operant Conditioning to teach dolphins to associate the raising of a hand to being fed a treat (positive reinforcement) and swimming toward the hand.
The conditioning of the pairing of a raised hand and being given a positive reinforcement (treat) is essential for all dolphins, and should be trained as soon as possible. This will be useful for the training/ shaping of future tricks, particularly when there will be a delay between the completion of a trick/ behaviour and the rewarding of a positive reinforcement. For example, a dolphin performs a successful flip on the opposite side of the pool. Raising your hand will signal the dolphin that a treat is about to be given, and it will swim toward you to receive its reward.
Required items:
bucket of treats (generally fish)
Instructions:
Whilst training dolphins, it is standard to give them a treat (positive reinforcement) after successfully performing a behaviour expected of them.
Before giving them the treat, wave it around just out of reach to grab the dolphin’s attention. Raise your opposite hand simultaneously as you feed them the treat.
Repeat this consistently and immediately whenever a treat is given, as the dolphin may associate being fed with some other stimuli if the hand being raised is too delayed or if the dolphin is looking elsewhere.
A good way to apply this/ check if the dolphin is properly conditioned is to raise your hand from across the pool. If conditioned properly, the dolphin will swim toward you, expecting a treat.
If the dolphin does not do this, repeat step 3 until success.
This behaviour is expected to be practiced frequently, as it can be used in conjunction with many things such as calling dolphins to feed them a treat after a successful behaviour, or simply calling a dolphin from across the pool.
Strengths and Limitations:
Strengths:
Strengths of operant conditioning include that it is a very straightforward concept, making it easy to use. It is a scientific theory, meaning that the theory was proven through facts from observations and experiments, making it a reliable method for conditioning dolphins.
Limitations:
Limitations of operant conditioning include that it doesn’t always work. The subject being conditioned may fake a change in behaviour, in order to get the reward. In reality, there is no genuine change in its behaviour at all. Another limitation is it only addresses specifically good and bad, nothing in between. Also, there is the chance the if rewards were stopped being given, the subject will stop exhibiting the expected behaviour.
Observational Learning:
Background:
Observational Learning is a type of learning that requires the dolphin to watch the behaviour of another dolphin, note whether the consequences of the dolphin’s behaviour are positive or negative, and decide whether to imitate the behaviour or not. For example, a child might be on a swing, and not swinging very much. He then watches the boy next to him, who seems to be going much higher than him and having more fun. He takes note that he pushes upwards of the floor when swinging backwards, and keeps his legs straight a raised when swinging forward. He then decides to imitate the boy’s actions, as he wants to swing higher and have more fun as well. Some key theorists of observational learning include Bandura and Joslin, Fletcher, and Emlen, followed by Cook and Mineka.
Bandura’s experiment involved set groups of children watching adults beating up a bobo doll either live, via movie, or via cartoon. The adult was either rewarded, punished, or given no consequence whatsoever. Some children from each group were offered a positive reinforcement if they were to act aggressively. The results from this experiment showed that children who watched the adult receive a punishment showed less likelihood to behave aggressively. However, most of the children who were offered a positive reinforcement to act aggressively did exactly that, even if the adult they had watched was punished. Also, boys were generally more aggressive than girls. From his study, Bandura concluded that children don’t just blindly mimic adults, they decide based on the fact if the adult was given a reward or punishment. He also concluded that despite not immediately imitating a model, a child may still have learned the behaviour of a model, and will perform the same behaviour at a later time ,when a reward was offered for their efforts.
Joslin, Fletcher, & Emlen’s experiment compared the response of 31 laboratory-raised rhesus monkeys and 27 wild rhesus monkeys to snakes. It was found that the lab-raised monkeys had no response whatsoever other than curiosity to the snakes, while the wild monkeys displayed fear, which led to the conclusion that the fear of snakes is not naturally imbedded in rhesus monkeys, but learned. This was then further experimented nearly 20 years after by Cook and Mineka, whose experiment was to see if rhesus monkeys could observationally learn from other rhesus monkeys. Laboratory-raised rhesus monkeys were made to observe the behaviour of wild rhesus monkeys being exposed to snakes (behind a glass), which was fear. Prior to observing the wild monkeys, the lab-raised monkeys had no fear when exposed to snakes. However, the lab-raised monkeys exhibited fear when exposed to the snakes after observing the wild monkey’s behaviours. The degree of fear that the lab-raised monkeys displayed was closely-related to the wild demonstrator monkey’s fear, meaning the more fear displayed by the demonstrator, the more fear the lab-monkeys exhibited. The conclusion is that observational learning is applicable to rhesus monkeys, and the degree of behaviour of the observant is correlated to the subject being observed.
Application:
Using Observational Learning to teach dolphins to learn a backflip by watching another experienced dolphin perform the backflip.
Observational Learning is one of the most useful methods of conditioning a dolphin to perform various tricks. It’s self explanatory; all that’s required is the novice dolphin to be observing the experienced dolphin as you instruct it to perform a backflip. This can be applied to almost any trick that an experienced dolphin already knows, making it one of the most useful and convenient methods of conditioning. Babies watching mothers in particular show some of the highest success rates, as the child wants to be like its parent/ parent figure.
Required items:
experienced dolphin who already knows how to do a backflip when you rotate your arm.
Instructions:
Make sure both dolphins, experienced and novice, are in the performance pool.
Assuming the experienced dolphin has already learned this trick, rotate you arm in a wide circle, which will make it leap out of the water and do a backflip. Raise your hand to call it over, and give it a treat.
Repeat step 2 a couple times, while the novice dolphin is interested. You will have to do this over a couple of days depending on how quickly the novice dolphin picks up the trick.
When training with the novice dolphin, rotate you arm. If successful, the dolphin will imitate the experienced dolphin, and leap out of the water and perform a backflip. If unsuccessful, repeat steps 2 and 3 until success.
Strengths and Limitations:
Strengths:
The strengths of observational learning is that making mistakes can be avoided. Simply by watching other dolphins make mistakes/ get hurt/ do well, the subject will have a better grasp of the consequences that follow behaviours without actually having to attempt the exact same actions. This is useful as getting injured can be avoided just by studying other more experienced/brave dolphins. Another strength is that with a model, learning of techniques etc. for dolphins is much faster as they have a real life example right in front of them rather than trying to guess what to do with humans, due to the communication barrier at times.
Limitations:
Limitations of observational learning include the chance of a bad model. Bad models equal poor behaviour, which is exactly what we do not want novice dolphins to mimic by accident. In this situation, it’s a possibility that both the bad model and the influenced novice will have to be removed from the current habitat. Also, motivation to learn may also be diminished and work the opposite way, as the better model may discourage the novice dolphin to even try learning, as they feel inferior to the already experienced veteran.
References
6 Advantages and Disadvantages of Classical Conditioning. (2017). ConnectUS. Retrieved 13 March 2017, from http://connectusfund.org/6-advantages-and-disadvantages-of-classical-conditioning
A comparison of the responses to snakes of lab- and wild-reared rhesus monkeys. (2017). Sciencedirect.com. Retrieved 13 March 2017, from http://www.sciencedirect.com/science/article/pii/0003347264900235
Benefits Of Observational Learning. (2017). Benefits Of. Retrieved 13 March 2017, from http://benefitof.net/benefits-of-observational-learning/
Classical Conditioning | Simply Psychology. (2017). Simplypsychology.org. Retrieved 13 March 2017, from http://www.simplypsychology.org/classical-conditioning.html
Entertainment, S. (2017). Seaworld.org. Retrieved 13 March 2017, from https://seaworld.org/animal-info/animal-infobooks/animal-training/animal-behavior-and-learning/
ISTC 301: Integrating Instructional Tech / Strengths and Weaknesses of Operant Conditioning. (2017). Integratingtech301.pbworks.com. Retrieved 13 March 2017, from http://integratingtech301.pbworks.com/w/page/20021638/Strengths%20and%20Weaknesses%20of%20Operant%20Conditioning
Koopman, S. (2017). The Monkey and the Snake. Bird Brains and Monkey Minds. Retrieved 14 March 2017, from https://animalcogblog.wordpress.com/2014/01/29/the-monkey-and-the-snake/
National Aquarium, Dolphin Training Fact Sheet (1st ed.). Baltimore.
S, C. (2017). Selective associations in the observational conditioning of fear in rhesus monkeys. – PubMed – NCBI. Ncbi.nlm.nih.gov. Retrieved 13 March 2017, from https://www.ncbi.nlm.nih.gov/pubmed/2230660
The Disadvantages of Observational Learning | Synonym. (2017). Classroom.synonym.com. Retrieved 13 March 2017, from http://classroom.synonym.com/disadvantages-observational-learning-8571259.html